Distributed Collaborative Practices in Resource Restricted Settings. By Nima Herman Shidende

Distributed Collaborative Practices in Resource Restricted Settings. By Nima Herman Shidende
Distributed Collaborative Practices in Resource Restricted
Settings.
Ethnographic studies from the Tanzanian Primary Healthcare Information System
By Nima Herman Shidende
A thesis submitted in partial fulfillment of the requirement of the degree of
Doctor of Philosophy (PhD)
Department of Informatics,
Faculty of Mathematics and Natural Sciences,
University of Oslo, Norway
August 2015
© Nima Herman Shidende, 2015
Series of dissertations submitted to the
Faculty of Mathematics and Natural Sciences, University of Oslo
No. 1695
ISSN 1501-7710
All rights reserved. No part of this publication may be
reproduced or transmitted, in any form or by any means, without permission.
Cover: Hanne Baadsgaard Utigard.
Print production: John Grieg AS, Bergen.
Produced in co-operation with Akademika Publishing.
The thesis is produced by Akademika Publishing merely in connection with the
thesis defence. Kindly direct all inquiries regarding the thesis to the copyright
holder or the unit which grants the doctorate.
TABLE OF CONTENTS
LIST OF FIGURES.............................................................................................................................III
LIST OF TABLES...............................................................................................................................IV
ABBREVIATIONS AND ACRONYMS USED................................................................................. V
ACKNOWLEDGEMENTS................................................................................................................VI
ABSTRACT ....................................................................................................................................... VII
CHAPTER 1: INTRODUCTION ........................................................................................................ 1
1.1 Research Motivation and Problem Area: ...................................................................................... 1
1.1.1 Research Background: Improving Maternal and Child Healthcare Services in Tanzania ..... 2
1.1.2 Theoretical Motivation: .......................................................................................................... 6
1.1.3 Personal Motivation.............................................................................................................. 10
1.2 Research Objectives, Questions and Expected Contributions..................................................... 10
1.3 Research Approach...................................................................................................................... 11
1.4 Thesis Organization..................................................................................................................... 12
1.5 Brief Summary of Research Papers............................................................................................. 12
CHAPTER 2: RESEARCH SETTING ............................................................................................. 15
2.1 Tanzania’s Country Profile.......................................................................................................... 15
2.1.1 PMTCT program .................................................................................................................. 21
2.1.2 Immunization........................................................................................................................ 22
2.2 Health Administrative Structure.................................................................................................. 23
2.3 Healthcare System of Tanzania................................................................................................... 25
2.4 Official Information Systems that Support Maternal and Child Health Services ....................... 27
CHAPTER 3: RELATED RESEARCH............................................................................................ 30
3.1 Review of HIS Literature in Developing Countries .................................................................... 30
3.1.1 Health Managerial Orientation in the HIS Literature........................................................... 30
3.1.2 Patient-Care Oriented HIS literature .................................................................................... 33
3.2 Collaborative Work ..................................................................................................................... 37
3.2.1 Collaboration in Facilitating Clients’ Trajectories ............................................................... 39
3.2.2 Challenges of Collaborative Work in facilitation of Clients’ Trajectories........................... 42
3.3 Distributed Participatory Design in Patient-Care Software......................................................... 45
CHAPTER 4: THEORETICAL FRAMEWORK............................................................................ 50
4.1 Activity Theory ........................................................................................................................... 51
4.2 Relevant Activity Theory’s Concepts for Analysis..................................................................... 54
4.2.1 Activity System .................................................................................................................... 55
4.2.2 The Concept of Mediation.................................................................................................... 56
4.2.3 The Concept of Contradiction, Disturbance and Mediation Breakdown.............................. 57
CHAPTER 5: RESEARCH APPROACH ........................................................................................ 60
5.1 Research Strategy ........................................................................................................................ 60
5.2 Research Design .......................................................................................................................... 62
5.3 Data Collection............................................................................................................................ 67
I
5.3.1 Primary HealthCare Related Fieldwork ............................................................................... 67
5.3.2 DHIS Tracker Implementation ............................................................................................. 75
5.4 Field notes and Data Analysis ..................................................................................................... 78
5.5 Ethical Considerations................................................................................................................. 81
5.6 Methodological Reflection .......................................................................................................... 82
CHAPTER 6: ANALYSIS OF THE KEY FINDINGS.................................................................... 85
6.1 Collaboration in Facilitating Care Trajectories across Health Programs and Facilities.............. 86
6.1.1 Activity Systems and Mediation .......................................................................................... 86
6.1.2 Disruptions within Care Trajectories and their Resolutions................................................. 90
6.2 Collaboration in Ensuring that Clients Complete their Care Trajectories................................... 94
6.2.1 Activity Systems and Mediation .......................................................................................... 94
6.2.2 Disruptions and their Resolutions ........................................................................................ 99
6.3 Collaborative Work in the Design and Development of the DHIS Tracker.............................. 103
6.3.1 Activity Systems and Mediation ........................................................................................ 103
6.3.2 Disruptions and their Resolutions ...................................................................................... 106
6.4 Summary ................................................................................................................................... 109
CHAPTER 7: DISCUSSION AND CONTRIBUTION ................................................................. 110
7.1 HIS in Developing Countries .................................................................................................... 110
7.2 Research on Collaborative Practice........................................................................................... 112
7.2.1 Actors involved in Facilitating Care Trajectories............................................................... 112
7.2.2 Mediation of Interactions between Distributed Actors....................................................... 113
7.2.3 Understanding of the Impact of Context on Distributed Collaboration ............................. 114
7.3 Participatory Design .................................................................................................................. 116
7.4 Activity Theory Literature......................................................................................................... 120
7.5 Design and Implementation Implications.................................................................................. 121
7.5.1 Primary Healthcare Information Systems........................................................................... 121
7.5.2 DHIS Tracker ..................................................................................................................... 123
7.6 Concluding Remarks ................................................................................................................. 124
References .......................................................................................................................................... 126
Appendices ......................................................................................................................................... 142
II
LIST OF FIGURES
Figure 1: Collaborative Work Practices within Primary Health Care Settings ....................................... 8
Figure 2: Tanzania map showing country boarders, health administrative zones ................................. 16
Figure 3: Dodoma Region Map ........................................................................................................... 20
Figure 4: Illustration of collaboration between MOHSW and PMO-RALG structures ....................... 24
Figure 5: Illustration of maternal and child health services in Tanzania............................................... 27
Figure 6: The first generation of activity theory ................................................................................... 51
Figure 7: The hierarchical structure of the activity system ................................................................... 52
Figure 8: The structure of human activity system ................................................................................. 53
Figure 9: Two interacting activity systems ........................................................................................... 53
Figure 10: Network of interacting activity systems............................................................................... 54
Figure 11: Observation of child health services provision .................................................................... 69
Figure 12: Some illustration from the observation sessions concerning defaulter tracing practices..... 70
Figure 13: Group interviews illustrations.............................................................................................. 74
Figure 14: Illustration of observation session conducted during the DHIS Tracker fieldwork............. 77
Figure 15: Three interacting activity system in the PMTCT services................................................... 87
Figure 16: The network of activity systems involved in the PMTCT Care........................................... 89
Figure 17: Resolution of contradiction between CAS and HCAS resulting in DTAS .......................... 93
Figure 18: The network of activity systems involved in defaulter tracing practices............................. 97
Figure 19: The network of activity systems involved in DHIS Tracker Design ................................. 104
Figure 20: Sample DHIS Tracker screen............................................................................................. 105
Figure 21: Standard DHIS tracker system with changed features....................................................... 107
III
LIST OF TABLES
Table 1: Overview of health and social welfare indicators in Dodoma Region ................................... 20
Table 2: Overview of health and social welfare indicators in the two districts .................................... 21
Table 3: Overview of study facilities showing ownership and type features ........................................ 64
Table 4: Summary of interviews at facility and community levels ....................................................... 74
Table 5: Summary of interviews at district, regional and zone levels ................................................... 75
Table 6: Summary of the Key Findings .............................................................................................. 109
IV
ABBREVIATIONS AND ACRONYMS USED
ARV ······················· Anti-Retroviral
CHMT ···················· Council Health Management Team
CHW······················· Community Health Worker
CSCW ····················· Computer Supported Cooperative Work
DHIS······················· District Health Information System
DHMT····················· District Health Management Team
DMO······················· District Medical Officer
EMR ······················· Electronic Medical Records
EPR ························ Electronic Patient Records
FOSS ······················ Free and Open Source Software
HIS ························ Health Information System
HISP ······················· Health Information Systems Program
HIV/AIDS················· Human Immuno-deficiency virus/Acquired Immuno-deficiency Syndrome
HMIS ······················ Health Management Information System
ICT························· Information and Communication Technologies
IS ··························· Information System
IT··························· Information Technology
MCH······················· Maternal and Child Healthcare
MOHSW ················· Ministry of Health and Social Welfare
NBS ······················· National Bureau of Statistics
NGO ······················ Non Governmental Organization
OSS ························ Open Source Software
PBS ························ Practice Based Studies
PD·························· Participatory Design
PHC························ Primary Health Care
PLWHIV ·················· People Living with HIV
PMORALG ··············· Prime Minister’s Office-Regional Administration and Local Government
PMTCT···················· Prevention of Mother To Child Transmission of HIV
RCH ······················· Reproductive and Child Healthcare
RHMT ····················· Regional Health Management Team
RMO ······················ Regional Medical Officer
TB·························· Tuberculosis
UNICEF ··················· United Nations Children’s Fund
VHW ······················ Village Health Worker
WHO ······················ World Health Organization
V
ACKNOWLEDGEMENTS
There are many people and organizations whose support and encouragement have made this work
possible. First and foremost, my special and sincere thanks go to my first supervisor, Margunn
Aanestad, for unconditional support, intellectual guidance and patience during this prolonged journey.
Thanks a lot Margunn for insightful intellectual directions and discussions, some of which I plan to
use in my future research. I also extend my gratitude for your belief and facilitation for me to
undertake my PhD. To my second supervisor, Christina Mortberg; I am very thankful for your
intellectual guidance and critiques/comments throughout my study journey. And my third supervisor,
Faraja Igira; you have inspired me with your academic guidance, tolerance, and Tanzanian HIS
perspective.
I have also received insightful comments and discussion that have contributed to my
research/dissertation from Miria Grisot, Bjørn Erik Mørk, Jens Kaasbøll, Tone Bratteig, and Sundeep
Sahay. I also thank Dodoma region bureau of health, district medical offices and primary health
facilities that I visited for your wonderful cooperation you provided me during the fieldwork. My
special thanks also go to my colleagues for wonderful discussion both academic and personal:
Caroline, Marlen, Hanne Cecilie, Zufan, Chipo, Zeferino, Tiwonge, Edem, Selamawit, Yahya, Ranga,
Anne, Lise, Pandula, Roshan, Esther, and Anna.
My sincere gratitude goes to my lifelong education benefactors that helped me reach this stage. I am
very grateful to my mother Veneranda Tembo and my late father Herman Mashinagwi Shidende for
the financial and emotional sacrifices made to make sure that I would become a scholar in some day. I
extend my special thanks to the Tanzanian government that provided me with quality primary,
secondary and bachelor degree education at lower cost which enabled me to reach this stage. Big
thanks to the Norwegian government through the quota scheme for financing both my master and
PHD studies. I also extend my gratitude to the HISP project for financing research costs and also for
personal financing during last year of my study. Thanks also to my employer, the University of
Dodoma, for granting me the prolonged study leave, financial support and research facilitation that
allowed me to accomplish my study.
I owe a lot to my family. Special thank my beloved husband, Lazaro Linjano, for encouragement and
loving support during this journey which started since my master studies. I also thank you for the
emotional support you gave me when I felt that the PhD journey was taking too long. Big thanks to my
children. Innocent, you especially as you have endured my absence for longer than your siblings.
Nahele, you made me sad whenever I called home as you would ask when I would be coming back.
Thanks a lot Juliana and Samuel for coping with an absent mom while you were still very young; it
made me sad the first semester when you could did not instantly recognize me when I arrived home
from Oslo. I extend my heartfelt thanks to Yuda, Martha, Constancia, and Stamili for taking care of
my children whenever we needed your help. I also thank my mother-in-law, Winfrida, Linus, and
Edna for their support during this journey. My special thanks go to Deogratias, Luka, Paul, Peter,
Rehema, Loveness, and Lucy for their support.
Last but not least, I thank the Holy Spirit for wisdom, understanding, love, inspirations, health and
encouragement. I also thank my favorite saints for their intercession; Blessed Virgin Mary, Joseph (the
husband of Mary), Maria Magdalena, Michael the Archangel, Don Bosco, Dominic Salvio, Jude
Thaddaeus, and Antony of Padua. My special appreciation also goes to African and Filipino prayer
community groups from St Olav catholic parish for wonderful moments I had with you which
simplified and made enjoyable my life in Oslo.
VI
ABSTRACT
The thesis is positioned within work practice studies with the aim of contributing to the understanding
of work practices in facilitating care trajectories in resource-restricted settings. My research has a
practical motivation stemming from primary healthcare services. The primary healthcare service is an
important health care component in the management of common diseases facing society since it is a
client’s first entry point to the healthcare system in any country. While the HIV-AIDS disease is a
major public health problem globally, Sub-Saharan Africa is the world’s most affected region. Thus,
primary healthcare services play crucial role in the management of HIV care and physical wellbeing of
citizens. The problem of clients’ attrition in HIV-related programs has been observed in many
countries like Tanzania, South Africa, Ethiopia, Cambodia, and Kenya. The literature recommends
that, in order to support client retention, a well-designed health information system, and preferably a
computerized one, needs to be in place. Unfortunately, failure in computerized patient-care
information systems is common due to the mismatch between the existing practices and the design of
the computerized information system. To minimize failure of information systems, researchers
recommend that the understanding of practitioners’ work practices – for example, how they utilize
information artifacts and handle the challenges – can be valuable to information systems development
teams. However, there is a dearth of HIS research which provides insights into the utilization of
designed and/or implemented computerized patient-oriented information systems to support healthcare
provision in the developing countries’ settings. I argue that the HIS research field has put more
emphasis on managerial aspect of HIS and less on the understanding operational primary healthcare
information systems, which is what this study aims at contributing to.
This study has been conducted in Dodoma, Tanzania under the Health Information System Program
which is based at the Department of Informatics of the University of Oslo. The study employed a
qualitative research strategy and has adopted an interpretive approach for data analysis. Research data
was collected in two districts (one rural, one urban) for the duration of 14 months. Some empirical
fieldwork was also conducted at the Department of Informatics of the University of Oslo. The actual
empirical findings were mainly obtained through multi-site ethnographic studies conducted in multiple
health programs, 11 primary healthcare facilities and catchment population. Also, participatory design
approaches were used to inform the design and implementation of the software for monitoring clients’
care trajectories. Activity theory has been the major conceptual lens through which the empirical data
have been analyzed.
This research extends the understanding of primary healthcare information systems by providing a rich
analysis of health workers’ practices related to the use of IS tools to support coordination of clients’
trajectories. This research also contributes to the understanding of the characteristics of the distributed
collaboration in healthcare services in resource-restricted settings by showing that there are more
actors in different health programs, community, and health administration levels. I have argued that it
is important that we pay attention to the role played by implementation mediators and the challenges
they face in the design-after-design practices in order to develop appropriate systems to support
healthcare. Moreover, this dissertation adds to the existing knowledge regarding the impact of context
in the distributed practices by highlighting the role played by other forms of contextual conditions (e.g.
disease specific, clients’ decisions, and rural/urban settings) in distributed collaboration. The thesis
concludes by calling more attention to context-sensitive design, implementation, and use of primary
healthcare information systems in resource-constrained settings.
VII
CHAPTER 1: INTRODUCTION
This chapter introduces the empirical problem and the research area addressed by this thesis,
as well as the conceptual perspective and research approach adopted. The expected
contributions and their relevance are also introduced.
1.1 Research Motivation and Problem Area:
This research addresses the issue of how health information systems can be designed and
implemented to support healthcare services provision in developing countries. By developing
countries I mean the countries that have a low gross national income per capital per year. 1
Consequently, developing countries are constrained by resources such as staff, finance,
transport infrastructure, technology and skills.
Although there have been improvements in health worldwide, there are still a number of
public health challenges, for example, diseases (e.g. cardiovascular diseases, infectious
diseases, and HIV-AIDs) which are leading causes of death globally in contemporary society.
While HIV-AIDS is a major public health problem globally, Sub-Saharan Africa is the
world’s most affected region. For example, 69% of HIV-AIDS global cases are found in the
region, which has only 10% of the world’s population 2. For developing regions like SubSaharan Africa, HIV-AIDS is also a social-economic and development problem since it adds
more challenges to already resource-constrained settings, by increasing expenditure on health
services and further reducing the available labor workforce for economic activities,
productivity and income of families and individuals.
Subsequently, Sub-Sahara African countries have responded to the HIV/AIDS pandemic in
many ways. For example, interventions in HIV diagnosis, treatment, and care (e.g. provision
of Anti-Retroviral (ARV), Prevention of Mother-to-Child Transmission (PMTCT)) and
prevention and management of opportunistic infections like tuberculosis, viral hepatitis and
mental health. The PMTCT program offers a range of healthcare services that reduce the risks
of mother-to-child transmission of HIV during pregnancy, delivery and breastfeeding periods.
The specific services includes HIV testing and counseling for pregnant women (and their
partners), delivery of ARV prophylaxis or treatment, safer delivery practices, counseling and
1
2
http://www.isi-web.org/component/content/article/5-root/root/81-developing
www.unaids.org
1
support related to breast-feeding of the infant, as well as treatment for mothers and children
living with HIV.
Primary health care (PHC) service is an important healthcare component in the management
of health services in any country (Van Bemmel and Musen, 1997) because, it is a first entry
point for clients to the healthcare system in any country (Lippeveld and Sauerborn, 2000;
Ludwick and Doucette, 2009), and thus offers a point where HIV cases can be diagnosed and
care provided. PHC services are aimed at health promotion, prevention and curing of diseases
without admitting the patients to the health facilities, or, in some cases, PHC services can
offer basic in-patient care (Gilson et al., 1994; Smit and Venter, 2009). Moreover, by its
nature, PHC systems are geared towards addressing the health problems facing a certain
community by providing health promotion, prevention, rehabilitation and cure (WHO, 1978),
and thus make an important contribution to physical wellbeing. It is common for PHC units to
offer health services related to prevention and control of endemic diseases, appropriate
treatment of common diseases and injuries, provision of essential drugs, education concerning
prevailing diseases and health problems, and maternal and child health care (WHO, 1988).
The PHC also offers care for chronic disease management, such as for HIV/AIDS,
tuberculosis, mental illness and diabetes.
It is recognized that well-designed health information systems are crucial tools for improving
the quality of care (Lippeveld and Sauerborn, 2000). Similarly, it is acknowledged that PHC
information systems play essential roles in the PHC system since they facilitate many
functions: to identify people who need such services, to determine what their needs are and
their
whereabouts;
to
identify
problems
in
services
provision
or
clients’
compliance/adherence; to monitor progress; and to show whether the program is having any
effect, including impact on health status (Husein et al., 1993). In this thesis, I define Primary
Healthcare Information System (PHC-IS) as the collection of data, information artifacts,
people, and the information-related procedures that are used to support primary healthcare
services.
1.1.1 Research Background: Improving Maternal and Child Healthcare Services in
Tanzania
This research is motivated by the current efforts undertaken by the Tanzanian government to
improve maternal and child healthcare services in general and specifically for the PMTCT,
and how information systems can be designed to support the services. In the past, child
2
mortality rates in Tanzania have been high due to pregnancy and childbirth complications,
newborn illnesses (e.g. malaria), childhood infection, HIV/AIDS pandemic, and malnutrition
(MOHSW, 2008). However, Tanzania has been keen to reduce the mortality rates and thus
has incorporated several interventions proposed by the international organizations (MOHSW,
2008). For example, Tanzania has committed to achieving the Millennium Development
Goals (MDG) set globally by the United Nation organizations, of which goal number 4 aims
to reduce child mortality; goal number 5 aims to improve maternal health; and goal number 6
aims to combat HIV/AIDS, malaria and other diseases.
The Ministry of Health has initiated several interventions to realize the MDG goals, including
immunization of children against vaccine-preventable diseases and incorporating the PMTCT
program in all related Maternal and Child Healthcare (MCH) services in the country.
Moreover, the actual treatment of HIV for pregnant women and their born children is done in
collaboration with other clinics such as HIV/AIDS- and tuberculosis clinics.
There has been a tremendous improvement in the reduction of both infant and child mortality
rates. For example, the report released by the UNICEF shows that Tanzania has achieved
reduction of under-five child mortality3 from 155 to 38 deaths per 1000 births between 1995
and 2013. Nevertheless, among prevailing challenges in the provision of MCH services in
general and in HIV related services is the problem of client attrition. For example, the survey
conducted in the year 2010 shows that only 75% of children age 12–23 months were fully
immunized, a modest increase relative to the proportion reported in the 2004–05 survey
(71%) and the 1999 survey (68%) (NBS and ICF, 2011). Also, official documents from the
Tanzanian government’s HIV/AIDS program indicate that not all children born to pregnant
women with HIV have access to the PMTCT services.7 Moreover, the HIV care and treatment
is challenged by the low enrolment of children into HIV care and ART services and poor
retention of children under care and treatment (MOHSW, 2013, p.1).
There is also evidence in reports from a Non-Government Organization (NGO) that is an
implementing partner of HIV care, and from research articles. For example, the analysis of
patient data conducted by the NGO in the year 2013 indicated that retention of patients in
HIV care was low and trends showed continual decrease over time. 4 Also, research conducted
in the Mwanza region in the northern part of Tanzania reports on the differences between the
http://data.unicef.org/child-mortality/under-five accessed on February 21 2015 14.50
4
(www.tunajali.or.tz/career/factors-assosciated-plhiv-lost-treatment)
3
number of referred PMTCT clients from antenatal care and those who attended the HIV clinic
to receive care and treatment within the HIV/AIDS services (Watson-Jones et al., 2012). The
article’s findings indicate that only 45% of referred PMTCT clients from antenatal clinics
attended the HIV clinic. Similarly, other researchers have reported on the problem of client
retentions in the HIV program in other parts of Tanzania (see for example Somi et al., 2009;
Mwatawala et al., 2012; Layer et al, 2014).
The problem of client attrition is known as ‘defaulting’ in the public health discourse (e.g.
Ramadhani et al., 2007; Assefa et al., 2010; Bennett et al., 2012) and has consequences for
patients themselves, and it undermines the efforts put in by the government to reduce
morbidity and mortality rates. The policy document developed for the acceleration of
reduction of maternal, newborn and child deaths in the year 2008 points to several strategies
for countering client attrition in maternal and child healthcare, including raising the level of
national immunization coverage, increasing the percentage of HIV positive pregnant women
who receive antiretroviral treatment, and improving access to the PMTCT services for the
children born to HIV-positive pregnant women (MOHSW, 2008). At the operational level, the
efforts to counter client attrition in vaccination, PMTCT and HIV programs include reaching
every child and all districts. Also, the government recommends that improvement of retention
of clients in health programs can be achieved through improvement of the clients’ adherence
to medication and clinic visit schedules through community health workers visiting clients in
their homes (MOHSW, 2010). However, proper information is needed to aid the
understanding of health providers on what has happened before, the patient's HIV clinical
stage, what medical care, education, and psychosocial support have been provided on earlier
visits, and a summary of the patient's antiretroviral therapy (WHO 2004).
Health information systems (HIS) can be defined as the use of paper-based documentation or
computer-based tools to generate information for improved decision making at each level of
the health system (Sauerborn and Lippeveld, 2000). HIS tasks for generating information
include processes of data capturing, transmission, storing, retrieving, manipulating and use of
information both for healthcare services provision and planning and management. As such,
HIS can be broadly categorized into patient-care information systems and Health
Management Information Systems (HMIS). Patient-care information systems contain detailed
information about clients related to their demographics, diagnosis and treatment provided,
which is then used by health providers to coordinate processes of care services and to support
4
their collaboration (Berg, 1999b). Examples of patient-care information systems include
electronic medical records (EMR), clinical decision support systems, pharmacy information
systems, laboratory information systems, and radiology information systems (Van Bemmel
and Musen, 1997). Meanwhile, HMIS is a combination of statistical data obtained mainly
from five sources: censuses, population surveys, vital event registration systems (birth, deaths,
and migration); administrative systems (e.g. budgeting, health financing system, health
personnel system, health supply and logistic system); and data about health service provision
from various related activities from various sources (e.g. patient register book, tally sheets,
etc.) (Sapirie, 2000; Nsubuga, 2001). These data are used to derive information about the
health needs, health resources, costs, use of health services, and outcomes of use by the
population of a specified jurisdiction. Some sub-components of HMIS (e.g. vital registration
system, censuses, health surveys, and finance) are not administered directly by the health
sector (Sapirie, 2000; WHO, 2008).
Both HIS and public health literature also suggest that to support client retention, a welldesigned health information systems, and preferably a computerized one, needs to be in place
(see for example, Fraser et al., 2007; Thomson et al., 2011; DeRenzi et al., 2012; MacLeod et
al., 2012). For example, the use of information systems that keep records of different health
services offered, provide continual updates about treatment status and track all patients within
health programs, is recommended (Fraser et al., 2007; Ghadrshenas et al., 2013).
Unfortunately, the problem of failure of computerized information systems in developing
countries is not uncommon (e.g. Lyytinen and Robey, 1999; Heeks, 2002; Xue et al., 2005;
Pan et al., 2008). Such failures are attributed to a mismatch between the existing practices and
the design of the computerized information system (Heeks, 2002; 2008). Moreover, the design
and implementation of computerized patient-care information systems is a complex endeavor
that has faced challenges even in the developed world (e.g. Berg, 1999; Jones 2003; Harrison
et al, 2007; Safadi and Faraj, 2010; Cimino, 2013). These studies also indicate that one main
challenge facing the successful implementation of patient-care information systems is the
tension between existing work practices and the new technologies.
To minimize failure of information systems, researchers recommend that the understanding of
practitioners’ work practices for example, how they utilize information artifacts and handle
the challenges – can be valuable to information systems development teams (William et al.,
2005; Safad and Faraj, 2010). However, there is a dearth of HIS research that provides
5
insights into the utilization of designed and/or implemented computerized patient-oriented
information systems to support healthcare provision in the developing countries (Kaplan,
2006; Fraser et al., 2007; Lester et al., 2010). In particular, the evidence from the HIS
literature suggest that there is less emphasis on patient-care information systems than on
HMIS in the research field. Moreover, research on primary healthcare information systems
has tended to favor design of functionality and has not focused on understanding the work
practices within the PHC settings. For example, my literature review shows there is no
research that has explored the utilization of paper-based information systems with sufficient
detail to adequately inform the design of patient-care information systems. Hence, it is not
certain from the literature review how HIS can be designed, implemented and used to support
healthcare services, for example in relation to the problem of client retention within the
PMTCT services. Therefore, this research aims to increase the understanding of the use,
design and implementation of HIS to support maternal and child care services with the
specific focus on the challenges posed by the need to ensure clients’ retention.
I build this understanding of client retention practices from Practice Based Studies (PBS)
which offers the opportunity for understanding how real-time practices are carried out in the
workplace (Nicolini et al., 2003). Brown and Duguid (1991) argue that practice is central to
understanding work and success or failure of organizations can be determined by actual
practices and not formal practices, since reliance on formal description of work can blind an
organization to the valuable practices of its members. I draw on work practice studies that are
undertaken to understand collaborative work in order to direct attention to real work practices
and the use of technologies within work settings as well as in informing the design of
computerized information systems (e.g. Orr, 1996; Luff et al., 2000; Karasti, 2001a). More
specifically, I relate my dissertation to ethnographically-inspired workplace studies found
within the field of Computer-Supported Collaborative Work (CSCW). The aim of the CSCW
field is to understand the nature of cooperative work in its natural settings with the objective
of designing computer-based technologies to support work practices (Schmidt and Bannon,
1992). By definition, collaborative work is where different actors work together towards a
common goal.
1.1.2 Theoretical Motivation:
The main theoretical motivation of my dissertation is to strengthen the understanding of
collaborative work in distributed healthcare practices in resource-restricted settings. There is
6
currently a gap in the literature as concerns the understanding of these challenges. The HIS
literature has its emphasis on strengthening HIS in a resource-constrained context.
Regrettably, as I will show in chapter 3, the existing literature on HIS does not pay much
attention to studying the existing work practices of health service providers. The CSCW
literature, on the other hand, provides insights on the work practices of health workers, but
less so from resource-constrained contexts (Fitzpatrick and Ellingsen, 2013). There is also an
inclination in the CSCW literature towards studies of small-scale, co-located groups rather
than distributed collaboration (Monteiro et al., 2013). Therefore, as I will demonstrate in the
literature review, we lack knowledge about how HIS can be designed, implemented and used
to support the problem of client retention in the PMTCT services. I will now present my
argument for why my case requires the combination of the insights from HIS and CSCW and,
moreover, I will describe the complexity of my case that motivates the need to develop further
understanding.
Comprehensive intervention for PMTCT services involves many specialists working in
multiple clinics such as, maternity and child care, HIV/AIDS, Tuberculosis (TB) (in case
pregnant women develop TB infection), and for children up to 5 years of age (for
immunization services). Also, previous research acknowledges that client retention practices
require the efforts of both facility- and community-based approaches (Sapirie, 2000; Busza et
al., 2012). The community actors that can become involved are diverse: community health
workers, peer counselors, volunteers, mentor mothers, traditional chiefs and religious leaders
(Marcos et al., 2012). The proper functioning of healthcare services delivery also needs the
work of health administration actors to coordinate and provide planning and management
support (Lippeveld and Sauerborn, 2000). Furthermore, design and implementation of patientcare information systems requires collaboration between practitioners and Information
Technology (IT) professionals for the successful operation of support healthcare services
(Bandon et al., 2005). I have used insights from Tanzanian PMTCT care settings and the
literature review to illustrate the collaborative practices within PHC settings in Figure 1.
7
Primary Health Care Information
Systems Design and Implementation
Primary Health Care Provision Settings
Health Administration
Health providers in
health program 1
f
Health providers in
another facility
c
Health providers in
health program n
a
b
h
g
Software
development team
Other Software
implementers
and designers
d
Community members
e
Figure 1: Collaborative Work Practices within Primary Health Care Settings
Key:
a ······························ client (green)
b, c, d ························· health providers in different health programs and facilities (orange)
e ······························· community members (turquoise, pink, purple and navy blue)
f ······························· health administration staff
g ······························ implementers (blue)
h······························· other software professionals
The diagram shows that health providers (b, c, d) from different health programs collaborate
around a patient trajectory by offering different types of care, such as for MCH, HIV and TB.
The primary healthcare services also involve interaction with clients (a) who interact with
health providers, and moves in multiple health programs (b, c, d) to acquire healthcare
services. Moreover, community members (e), such as community health workers and
community-based organization, are needed to facilitate retention of clients in their care
trajectories. Primary healthcare service provision also requires the work of health
administration staff (f) that provide supervision and coordination support. The design and
development of PHC-IS needs the participation of different actors in order to develop
appropriate systems to support the work of the facilities. Health workers also participate in
8
information system design and implementation activities through training and use of the
system and in giving feedback. Software development organizations also have a range of
people with different professional roles, such as a project coordinator, developers, system
analysts, and implementers (h). Implementers (g) link health workers and other software
professionals.
The diagram indicates that the efforts to facilitate a client’s care trajectory require multiple
actors who could be in different locations (health facilities, home, and community). As
indicated in the diagram and the insights from the above discussion, the achievements of
Tanzania’s PMTCT services are the outcome of the collaboration of many actors who are
geographically and organizationally distributed, such as within health programs, health
facilities, at community and administrative levels in regions, and districts. As such, the
collaborative work that takes place in facilitating healthcare trajectory in PMTCT is
analogous to distributed collaborative systems which involve multiple, geographicallydispersed people (Star and Ruhleder, 1994; O’Day et al., 2001; Hinds and Mortensen, 2005).
Unfortunately, existing literature on CSCW in healthcare has not paid sufficient attention to
studying clients’ care trajectory in distributed and resource-constrained settings. Therefore,
CSCW studies are lacking that analyses and discusses collaborative work around patient
trajectory in multiple settings where healthcare services are offered (Fitzpatrick and Ellingsen,
2013). As a result, there is limited understanding of distributed collaborative practices in
resource-restricted settings and related to care trajectories distributed across multiple settings.
Primary health care is influenced by contextual factors such as geography, history, culture,
infrastructure, and the economic, social-cultural and political characteristics of the place
where it is based (WHO, 1988; Byrne and Sahay, 2007). The HIS and Information and
communication technologies for development literatures seldom explicitly theorize the role of
context, with some exceptions (e.g. Tiihonen et al., 2006; Tiihonen, 2011; Tongia and
Subrahmanian, 2006; Lund and Suutinen 2010; Sinnig et al., 2010; Pitula and Radhakrishnan
2011). The majority of these works address aspects of context that are relevant specifically to
design and implementation of ICT, such as how (computer) literacy and limited material
resources have a direct impact on the design and development activities. On the other hand,
Macaulay et al. (2000) claim that contextual features of the use situation may include
environmental, social, cultural, individual and historical factors affecting how people manage
their work practices.
9
Unfortunately, previous CSCW research has also paid little attention to studying the impact of
other contextual features on collaborative practices (Harrison and Dourish, 1996). Where the
CSCW research addresses context, it is either as spatial context or as the organizational
context that influences the work practices. Consequently, the literature offers limited insights
into broader contextual conditions and their impact which can be used to inform primary
healthcare services of the developing countries context.
1.1.3 Personal Motivation
My personal motivation for studying primary healthcare information systems stems from my
master’s studies at the University of Oslo. Maternal and child health services were my
empirical settings and I collected research data from PMTCT activities in antenatal care
services in two districts in Tanzania (see Shidende, 2005). The findings highlighted misfits
between information system design and health service provision work practices due to
unavailability of information tools, client confidentiality, and clients’ geographical
movement. The gap between information systems design and actual health services provision
in the settings led to adaptation of information tools which had further consequences, such as
data duplication and lack of information-sharing between related health services. The
experience I acquired during my masters’ study motivated me to continue researching
maternal and child healthcare information systems practices and it also influenced both the
choice of empirical settings and the research problem I wanted to focus on. In my master
studies, I spotted the problem but I did not analyze it systematically, for example in
associating PHC facilities’ adaptation practices and their socio-cultural-economic situation.
Neither did I provide future recommendations in the design and implementation of PMTCT
information system, which I aim to do with my PhD thesis.
1.2 Research Objectives, Questions and Expected Contributions
In the light of the issues discussed above, this research aims to strengthen the understanding
of distributed collaborative healthcare practices in resource-restricted settings and thereby use
the understanding gained to inform the design of development of HIS to support PHC. To
accomplish the main objective, the study has identified two research questions:
1. What are the characteristics of distributed collaborative work of the PHC settings?
2. How does context impact the distributed collaborative work of the PHC settings?
10
This thesis is intended to contribute to conceptual and practical knowledge for primary
healthcare information systems in two ways. First, the thesis contributes to work practice
studies by adding knowledge about distributed collaborative practices in resource-restricted
settings in care trajectory, as the area is currently under-researched (Fitzpatrick and Ellingsen,
2013). Second, these insights will provide a basis from which to facilitate the design of ICT
tools to improve healthcare services, such as retention of clients. The problem of client
retention in PMTCT or HIV programs is not unique to Tanzania, as it has been observed in
other countries like South Africa, Ethiopia, Cambodia, and Kenya (e.g. Fraser et al., 2007;
Rispel et al., 2009; Thai et al., 2009; Assefa et al., 2010; Thomson et al., 2011). Thus, the
insights from the study regarding the recommendation for the design and implementation of
information systems to support client retention can be used to improve data management
practices not only in Tanzania but also in other developing countries.
1.3 Research Approach
This study has been conducted in Tanzania under the Health Information System Program
(HISP) which has as its primary goal to enhance the information-use behavior of health
managers, planners and workers at different levels of the health sector in developing countries
(Braa et al., 2004). HISP is based at the Department of Informatics at the University of Oslo.
HISP has developed software known as DHIS (District Health Information Software), which
is used to manage data at various levels of the health sector in more than 47 countries across
four continents (www.dhis2.org). The DHIS Tracker is a module within the DHIS software
platform designed to follow up and report on individual program service deliveries to
beneficiaries, such as patient with tuberculosis, expectant mothers in antenatal care, children
who attend children’s clinics in maternal and child healthcare, or patients with HIV/AIDS.
There has been long-term collaboration between the HISP project and the Ministry of Health
in Tanzania in improving the health information system in the country. Moreover, the DHIS
system is the standard software used to process aggregate data within Tanzania’s HMIS.
This study employed a qualitative and interpretive (Walsham, 1995; 2006) research strategy
and empirical fieldwork was conducted over 14 months in Dodoma Region, covering the two
districts and 11 primary health facilities. Data was collected using ethnographic and
participatory design-inspired approaches (Hughes et al., 1994; Blomberg and Karasti, 2012;
2013). The main data collection approach was participant observation and interviews. The
informants in the study were doctors, nurses, health officers, community health workers and
11
community leaders who were involved in maternal and child healthcare. Semi-structured
interviews were used to back up participant observations, and were conducted from
community to zone levels. The study also obtained data by software system implementation
where the DHIS Tracker was piloted in two health facilities. Other sources of data include
various interactions with software developers of the DHIS Tracker module at the University
of Oslo and also by conducting interviews.
1.4 Thesis Organization
This thesis is divided into seven chapters. Chapter 2 is a description of the research setting,
where it provides an overview of the maternal and child healthcare system of Tanzania and its
information system practices. Chapter 3 presents the related literature while chapter 4 presents
the theoretical framework used to analyze the research findings. Chapter 5 presents my
research approach where I provide detailed descriptions of how I have conducted the study.
Chapter 6 presents an overall, repeated analysis of the research findings by using the
theoretical framework presented in chapter 4. Chapter 7 discusses the research findings as
well as contributions arising from the thesis and its implications. It also provides concluding
remarks arising from the study.
1.5 Brief Summary of Research Papers
The findings of this thesis have been reported in five individual papers that are appended in
the annexes. Taken as a whole, the papers present various angles for understanding
collaborative practices performed by multiple actors when they need to facilitate clients’
trajectories across health programs and facilities. The summaries of the papers are presented
below:
Paper I: Shidende, N.H., Grisot, M., Aanestad, M., and Igira, F. (2014). Coordination
challenges in Collaborative Practices in the Prevention of Mother-to-Child
Transmission of HIV in Tanzania, Journal of Health Informatics in Africa, 2(1).
This paper analyzes collaborative and coordinative work practices in the prevention of
mother-to-child transmission of HIV (PMTCT) in Tanzania. The empirical material was
gathered from 11 facilities, using observation and interviews at facility, community and
administrative levels in the two districts. The analysis of the empirical material was done
through the concepts of patient trajectory, coordination, and coordination mechanism. This
paper illuminates collaboration between b, c, and d in figure 1. The paper illuminates also the
12
role of clients (a in figure 1) in facilitating collaboration across multiple health programs and
multiple facilities. The concepts of coordination and coordination mechanism from the CSCW
field frame the study’s overall academic contribution.
Paper II: Shidende, N.H. (2014). Challenges in Implementing Patient-Centred Information
Systems in Tanzania: An Activity Theory Perspective. The Electronic Journal of
Information Systems in Developing Countries, 64.
This paper describes the defaulter tracing systems operating in maternal and child health
services. The empirical findings were gathered from 11 primary healthcare services in the two
districts, using observation and interviews at facility, community and administrative levels.
The findings of the paper were presented and analyzed, using activity theory’s concepts of
mediation, activity system and contradiction. This paper discusses collaboration practices
between facility and community levels in facilitating patient trajectories through clients’
retention.
Paper III: Shidende, N., Aanestad, M., and Mork, B.E. (2014). The Role of Information
Artifacts in Organizing: an Empirical Study from Tanzanian Health Care. Paper
presented at the Conference of Organizational Learning, Knowledge and
Capabilities, Oslo, Norway, 2014.
The aim of this paper is to contribute to the understanding of the role of objects in supporting
the organization of work practices. The empirical data were gathered from four facilities in
the vaccination program in the two districts by using observation and interview techniques at
facility, community and administrative levels. The empirical data were analyzed using the
concepts of situated work from PBS (Gherardi, 2009; Nicolini, 2009; 2013),
convergence/divergence (Star et al., 2003) and cognitive artifacts (Norman, 1993). This paper
enriches the understanding of the collaborative work across health facilities, administrative
and communities levels (between d, e, and f in figure 1) and the qualities of artifacts needed
in different facilities to supports the multiple practice supporting immunizing children.
Paper IV: Shidende, N.H., and Mörtberg, C. (2014). Re-visiting Design-After-Design:
Reflecting Implementation Mediators Connectedness in Distributed Participatory
Design Activities. In Proceedings of the 13th Participatory Design Conference:
Research Papers-Volume 1 (pp. 61-70). ACM.
13
This paper aims at an extended understanding of the design facilitators' role, here
implementation mediators, in participatory design practices. The main focus was on
understanding the use of the DHIS Tracker to support defaulter tracing practices in the
PMTCT and immunization health programs. The empirical material was gathered using
ethnographic and participatory design approaches during the implementation of the DHIS
Tracker system in the two facilities from the rural district in Tanzania and also from the
interaction with its global developers based at the University of Oslo. To analyze the
empirical case, the role of the implementation mediators or design facilitators and their related
participatory activities are conceptualized by using the notion of design-after-design. This
paper illuminates collaboration between health facilities’ providers, design facilitators and
global actors in the design of software to manage patient data (b, c, d, and g in the figure 1).
Paper V: Shidende, N.H., Aanestad, M., and Igira, F. (Forthcoming). The role of context in
the co-evolution of work and tools: a case from the primary health sector in Tanzania.
Under review for publication in the international, peer-reviewed IS Journal.
The main focus of the paper is to examine the role of context for the emergence of
information-related practices and tools in healthcare delivery. Empirical work was conducted
in four primary healthcare facilities in two districts in the PMTCT program. The facilities had
different geographical, socio-economic and institutional locations. Empirical materials were
gathered through participant observation, interviews and document analysis at health facility,
community and administrative levels. The findings of the paper were presented and analyzed,
using activity theory’s concepts of activity system, contradiction and mediation breakdown.
The paper elucidates collaborative practices across health programs and facilities,
communities and the artifacts needed in different facilities to guide multiple practices
supporting the PMTCT program (b, c, and e in figure 1). Moreover, the paper illuminates how
context can be conceptualized within collaborative practices by using the activity theory.
14
CHAPTER 2: RESEARCH SETTING
This chapter offers an overview of the Tanzanian primary healthcare with a focus on maternal
and child healthcare services. The chapter starts by introducing the country’s profile:
geography and demography, education and maternal and child health system. Afterwards, the
healthcare delivery system and its management system around primary healthcare will be
presented. The chapter also gives an overview of the Prevention of Mother-to-Child
Transmission of HIV (PMTCT) and Immunization programs in maternal and child health
services because that is where the empirical work was conducted. Lastly, I present the health
information system that is operating in Tanzania.
2.1 Tanzania’s Country Profile
The United Republic of Tanzania is the largest country in east Africa, covering 940,000
square kilometers; 60,000 of which are inland water. The United Republic of Tanzania is
made up of two countries, Zanzibar and Tanganyika, which is formally known as Tanzania
mainland. This study was done in Tanzania mainland. Geographically, Tanzania lies south of
equator and shares borders with eight countries: Kenya and Uganda to the North; Rwanda,
Burundi, the Democratic republic of Congo, and Zambia to the west; and Malawi and
Mozambique to the south. Administratively, mainland Tanzania is divided into 8 zones which
are lake, western, northern, central, eastern, southwest highlands, southern highlands and
southern zones. These zones are further divided into 25 regions which are subdivided into 169
administrative districts 5 . Figure 2 depicts the map of Tanzania showing country borders,
zones and regions.
The districts are further divided into divisions and wards. The smallest level of administration
is called a hamlet which consists of about 50 to 100 households. About five to ten hamlets
make up one village and three to six villages constitute a ward. Several wards (3–5) make a
division, while 3–7 divisions make a district. Tanzania is currently operating under a
multiparty democratic system of government, with the president and the national assembly
members being elected every five years. There are two official languages in the country,
Kiswahili and English.
5
www.nbs.go.tz
15
Tanzania is currently projected to have a population of approximately 47 million people:
48.7% males and 51.3% females. 6 The high population growth rate in Tanzania has been
brought about by high fertility and declining mortality levels. The life expectancy rates at
birth are 63 and 60 years for female and males respectively (NBS and OCGS, 2014). The
population of Tanzania is predominantly rural, 74% according to a 2010 projection (NBS and
ICF, 2011).
Figure 2: Tanzania map showing country boarders, health administrative zones
and regions (TACAIDS et al, 2013)
Tanzania is one of the least developed countries in the world with a GDP (Gross Domestic
Product) of $49.18 billion (2014 estimation). 7 Agriculture is the largest sector of the economy
in Tanzania and is the major means of economic gain for the majority of the population who
reside in the rural areas. The actual agricultural activities comprise crop growing, animal
husbandry, forestry, fishery, and hunting.
6
7
http://www.nbs.go.tz/sensa/new.html
http://data.worldbank.org/country/tanzania
16
Tanzania’s formal education system is divided into three levels: primary, secondary and
tertiary education. Primary education consists of pre-primary and 7 years of primary
education. It is universal and compulsory for all children from the age of seven years. While
primary education is compulsory in Tanzania, the school infrastructure does not have the
capacity to enroll more than 50% of eligible children. Consequently, the literacy rates in the
country have been declining, with a current figure being 71.8% (5 years and above) while in
the past, due to adult education campaigns, more than 90% of people over the age of 15 were
literate (NBS and OCGS, 2014). Moreover, there are illiteracy disparities between gender and
geographical settings. The illiteracy levels are higher for females in comparison to males,
where 27% of females have never attended school, while for males it was found to be 18%
(NBS and ICF, 2011). Likewise, urban residents are more likely than rural residents to have
attended school and remained in school for longer. There are gender differences between
geographical settings as well. For instance, literacy rates for women and men in urban areas
are 88% and 94%, respectively, compared with 66% and 78% in rural areas (NBS and ICF,
2011). The highest percentage of population who has never been to school is found in regions
of Tabora (42% for females and 34% for males) and Dodoma (40% for females and 33% for
males).
Tanzania faces health problems similar to most developing countries; in particular, malaria,
HIV/AIDS, tuberculosis and leprosy, malnutrition and micro-nutrient deficiencies, child
illnesses, accidents and non-communicable diseases (MOHSW, 2006, 2007b; TACAIDS et al.,
2013). Also, although children under five years of age constitute about 16% of the population,
they account for 50% of the total mortality burden for all ages and neonatal conditions
account for over 80% of deaths (MOHSW, 2008). Further, the maternal mortality and child
mortality rates are higher in comparison with the developed world. Every year about 157,000
children die before reaching their fifth birthday. The 2012 census’ preliminary results
revealed an infant mortality rate of 45 deaths per 1,000 live births (NBS and OCGS, 2014).
The 2012 census’ preliminary results on maternal mortality rate in Tanzania estimated 432
deaths per 100,000 live births. The higher mortality rates of under-five years’ deaths are
caused by infectious diseases such as malaria, measles, neonatal tetanus, pneumonia, diarrhea,
and dysentery.
In relation to the HIV-AIDS pandemic, Tanzania is among the countries that have been
mostly heavily affected. For example, it is estimated that approximately 2 million people have
17
been infected with the disease since the first cases in 1983 (MOHSW, 2013b). Currently, it is
estimated that 5.1% of adults of child-bearing age (15-49) in Tanzania are HIV positive.
However, there is a disparity between the geographical settings and rural HIV prevalence is
lower than that of urban areas, with 4.3% and 7.2% respectively (MOHSW, 2013b). The
highest prevalence of HIV cases is found in Njombe region, with 14.8%. There are also
gender differences where HIV prevalence is higher among women (6.2%) than men (3.8%),
and is even higher for women attending antenatal clinics (6.9%). Consequently, AIDS-related
mortality rates among children under five years of age are at an unacceptable level. It is
estimated that 200,000 children under 15 years of age are living with HIV. Factors that have
contributed to the HIV prevalence include multiple sex partners, mobility, transactional sex,
cross-generational sex, poor quality of transfused blood, and gender inequities accompanied
with poverty.
Mother-to-child transmission (MTCT) is the most common mode of HIV infection among
children. According to MOHSW (2008), 18% of all HIV new infections in Tanzania are due
to MTCT. In Tanzania, the prevalence of HIV among women who attend antenatal care is
6.9%, and currently the estimated percentage of infants born to HIV infected mothers, who
themselves become HIV infected, is around 25%. It is expected that appropriate interventions
can reduce the transmission rate to less than 5%. One of the operational targets to be achieved
by the Tanzanian Government by the year 2015 is to have PMTCT services provided to at
least 80% of pregnant women, their babies and families.
The government has adopted several global commitments and the respective plans of action
including MDGs, universal access to comprehensive prevention, treatment and care program
by 2010; and the global elimination of Mother-to-Child Transmission of HIV strategy, 2010–
2015. Strategies to achieve the targets set for 2015 include advocacy and resource
mobilization;
health
system
strengthening
and
capacity development;
community
mobilization; promotion of reproductive and child health behavioral change; and fostering
partnership and coordination with community and private sectors (MOHSW, 2008).
The Ministry of Health has initiated several interventions to realize the adopted plans in order
to reduce childhood morbidity and mortality. For example, The PMTCT programs aims to
prevent HIV infection through the MTCT mode. Another intervention put in place by the
government is the immunization of children against vaccine-preventable diseases (NBS and
ICF, 2011, p.143). Children are prevented from malarial infection by offering their mothers
18
mosquito nets bonus credit during pregnancy and after birth and anti-malarial drugs during
pregnancy. Also, vitamin A supplements are given to children under five years of age. Also,
the maternal and child health services are provided free of charge in all government-owned
healthcare service provision units.
However, there are many challenges at ground level which hamper the intervention put in
place. For example, the health sector lacks adequate human resources and medical supplies,
which in turn challenges the provision of primary healthcare (MOHSW, 2007b; Simba et al.,
2010; Kwesigabo, 2012). There are also problems stemming from the existing weak
relationship between facility- and community-based organizations observed in healthcare
services. It is also identified that low educational attainment among women of child-bearing
age in Tanzania has a strong effect on reproductive behavior, fertility, infant and child
mortality, morbidity, and attitudes related to family health and hygiene (NBS and ICF, 2011).
For example, the low literacy rates among women of child-bearing age have an impact on the
survival of children and the government initiatives put in place. This has been shown by the
survey conducted in the year 2010 which indicated that the infant mortality rates of children
born of mothers who not completed primary school and mothers who have secondary
education is 67% and 52% respectively (NBS and ICF, 2011, pp. 121–122).
The research reported and analyzed in this thesis was conducted in Dodoma region centrally
positioned in Tanzania. The region covers an area of 41,310 sq. km., equivalent to 5% of the
total area of Tanzania Mainland. The region has 7 districts which are subdivided into 29
divisions, 190 wards and 557 registered villages. Figure 3 shows the map of Dodoma region
with administrative district boundaries. Bantu-speaking people are the major inhabitants (75%)
of the region’s population. Most tribes, including the Gogo, Rangi, Nguu, Zigua, Kaguru, and
Sagara, practice circumcision for males and female genital mutilation. The other majority
ethnic groups are Indians, Arabs and Somalis, mainly found in urban and rural trading centers.
Dodoma’s current regional average GDP is 151,608,000 Tanzanian Shilling (TZS), which is
about a 3% annual contribution to the national GDP. The regional economy is almost entirely
dependent on agriculture and animal husbandry. The main staples grown in the region include
sorghum, bulrush millet and maize, while major cash crops are groundnuts, sunflower and, to
a lesser extent, castor, sesame and pulses. The region is among the top four in the country in
terms of livestock numbers. However, the HIV/AIDS pandemic has had an impact on
agricultural production and health in the Dodoma region and the whole country.
19
Figure 3: Dodoma Region Map (TACAIDS et al, 2013)
The region has a total of 8 hospitals, 36 health centers and 293 dispensaries providing health
services to the community of Dodoma. Table 1 summarizes the health and social welfare
indicators of the region.
Table 1: Overview of health and social welfare indicators in Dodoma Region (RMO, 2013)
Area (km2)
Population (2012 projection)
Socio-economic indicators
41,310
2,264,133
Literacy rate 67.5%, national figure is 71. 8% (5 years and
above)
Economic status: one of the poorest regions in Tanzania
Maternal mortality rate per 100,000 live births
Infant mortality
Under five mortality rate per 1,000 live births
HIV prevalence among pregnant mothers (%)
56/100,000 live births
20/1000 live births
39
2
Other Challenges
Shortage of skilled staff
20
The study was conducted in two districts in the region: Bahari and Bwawa 8. Table 2 provides
an overview of the socio-economic context in the two districts as indicated in the district
profiles.
Table 2: Overview of health and social welfare indicators in the two districts (DMO, 2013, a, b)
District
2
Area (km )
Rural/ urban
Population (2012 projection)
Socio-economic indicators
Bahari
Bwawa
8742
Rural
319,044
Literacy rate 60 % (declining)
2769
Urban
552,135
Literacy rate 84 %
Among
the
poorest
and
least
developed districts
Maternal mortality
33/100,000 live births
97/100,000 live births
(unreliable data because lack of
information from community level)
Infant mortality
51/1000 live births
51/1000 live births
(unreliable, lack of information from
community level)
HIV prevalence among
pregnant mothers (%)
Health staff shortage
Other challenges
2%
8%
49 %
Nomad and pastoralist communities,
35 %
Weak coordination among
i.e. migrating population
donor/partners driven activities.
The study involved visiting primary health facilities in the two districts offering maternal and
child health services of PMTCT and immunization of under-fives.
2.1.1 PMTCT program
The PMTCT program offers HIV testing and counseling for pregnant women (and their
partners) in antenatal care, delivery of ARV prophylaxis or treatment to prevent transmission,
safer delivery practices, counseling and support related to breastfeeding of the infant, as well
as treatment for mothers and children living with HIV. The Tanzanian ministry of health and
social welfare collaborates with international partners to implement the PMTCT program
across the country. The roll-out of the PMTCT services is financed by the Government of
Tanzania, as well as a United States of America’s government program (The president’s
emergency plan for AIDS relief), United nations agencies, the global fund to fight AIDS, TB
and malaria, and Deutsche Gesellschaft für Internationale Zusammenarbeit. 9
8
9
The names are anonymous, in order to protect the health workers’ confidentiality
www.pmtct.or.tz
21
The Tanzanian government aims to increase the percentage of HIV-positive pregnant women
receiving antiretroviral treatment, to reduce transmission of the disease from mother to child,
as well as to improve access to care and treatment to HIV clinics (MOHSW, 2008). However,
there are challenges related to reaching pregnant women, and to ensure that they are tested
and that they comply with the treatment. Also, since approximately 50% of births in rural
Tanzania take place at home unknown to the health facilities, the enrolment of exposed
children into the programme becomes difficult. Only about 30% of the HIV-exposed infants
have access to an early infant diagnosis service7. There are also challenges related to the
actual provision of the service due to the availability of trained personnel, testing kits and
medication. Only about 37% of the PMTCT centers provide early infant diagnosis. In total,
about 56% of the estimated number of HIV-exposed infants in the population receive
preventive treatment (ARV prophylaxis) while, within the PMTCT Program, the percentage is
73%.
2.1.2 Immunization
Child immunization prevents diseases such as diarrhea, pneumonia and tetanus infection.
According to the existing guidelines, to be considered fully vaccinated, a child is supposed to
receive the following vaccinations: one dose of BCG
11
10
, three doses of rotavirus,
12
pneumococcal (PCV13 ), DPT and the polio vaccine, and one dose of the measles vaccine.
BCG, which protects against tuberculosis, should be given at birth or at first clinic contact.
Rotavirus vaccine protects against rotavirus infection, a common cause of diarrhea and
sickness in children. 13 DPT protects against diphtheria, pertussis (whooping cough), and
tetanus. In Tanzania, DPT is no longer given to infants as a stand-alone vaccine, instead, it
has been combined with other antigens that protect against hepatitis B (DPT-Hb). DPT,
Rotavirus, PCV13 and polio vaccine guidelines require three vaccinations given at
approximately 6, 10, and 14 weeks of age. The measles vaccine is to be given at 9 months.
However in Tanzania, the MOHSW is still trying to improve immunization coverage since
not all children are immunized (NBS and ICF, 2011). For example, a 2010 health survey
showed that there is a decrease in vaccination coverage between the first and third doses of
DPT/DPT-HB/DPT-HB-Hib and polio, which are 8% and 12% respectively. Three percent of
10
Bacillus Calmette–Guérin
Pneumococcal conjugate vaccine
12
Diphtheria, Pertussis and Tetanus
13
http://www.nhs.uk/conditions/vaccinations/pages/rotavirus-vaccine.aspx
11
22
children have received no vaccinations at all. With the exception of measles, more than 80%
of the vaccinations were given by 12 months of age, as recommended. There is also a
significant variation by residence: 86% of urban children were fully immunized compared
with 73% of rural children.
2.2 Health Administrative Structure
The administrative structure of the Tanzanian health sector is organized into five levels:
national, zone, regional, district and facility levels. Administratively, the Ministry of Health
and Social Welfare (MOHSW) is the highest authority of health management in the country.
The MOHSW is headed by Chief Medical Officer. The national level is responsible for
formulation of health policy and developing guidelines to facilitate the implementation of
those policies. The main roles at zone and regional levels are to co-ordinate and support the
regions and districts in planning and implementing the services, as well as conducting
monitoring and evaluation. The regional level is headed by Regional Health Management
Teams (RHMTs) under the leadership of the Regional Medical Officer (RMO) who is in
charge of all health services in the region. RHMTs interpret health policies developed by the
MOHSW and monitor their implementation in the districts they supervise.
The district is the focal point for the implementation of the health policy and interventions. Its
major role is formulating, carrying out and monitoring health intervention packages and
quality assurance programs. The district level is managed by District Council Health
Management Teams (CHMTs) which are headed by a District Medical Officer (DMO) who is
in charge of all district health services. CHMTs are responsible for health services provision
at primary health facilities and district hospitals and eventually the district team will supervise
all health facilities within its jurisdiction.
At the health facility level, the health facility management team will implement the health
policy and interventions. All primary health facilities are headed by health facility governing
committees under the leadership of those in charges of the health facilities. Facility governing
committees include members from the community level. Beyond facilities, community
leaders are also responsible for preventive tasks where they mobilize, educate and sensitize
community members to receive facility heath services by encouraging households and
individuals to participate in and contribute to their own health and the health of their village.
23
However, the actual implementation of the developed health policy and guidelines for primary
healthcare services in the MOHSW is done with the support of other ministries, one of them
being the Prime Minister’s Office for Regional Administration and Local Government
(PMORALG). Figure 4 shows the collaboration between MOHSW and PMO-RALG
structures in the inauguration of new vaccines at regional level in Dodoma Region.
A
B
Figure 4: Illustration of collaboration between MoHSW and PMO-RALG structures
A: Children who will undergo immunization
B: Dodoma regional commissioner prepares to inaugurate new vaccines at the regional level. Other local
government officers together with health officers are witnessing the inauguration
One of the PMORALG’s broader functions includes facilitating local government authorities
to provide quality services. The local government authority structure is organized from the
zone, regional, district, ward, to community level leaderships. The zone levels’ main roles are
to co-ordinate and support the regions and districts in planning and implementing the services,
as well as conducting monitoring and evaluation. The regional levels are headed by regional
commissioners and Regional Administrative Secretaries (RAS), while the district levels are
headed by district commissioners and District Administrative Secretaries (DAS). The wards
are headed by ward executive officers while villages are headed by village executive officers.
The hamlets/streets (where the population lives) are headed by hamlet leaders.
The local government structures under PMORALG work in collaboration with their
counterparts (medical professionals) who work under the structure of the MOHSW to
facilitate provision of quality primary healthcare. The RMO is accountable to the MOHSW on
issues related to technical management, and to the PMORALG (RAS) on issues related to
health administration and management of health services. The CHMT follow planning and
management guidelines issued jointly by the MOHSW and PMO-RALG. The DMO is
accountable to the council director on administrative and managerial matters, and otherwise
24
reports to the RMO on technical matters. All village/community councils consist of health
committees which work closely with a nearby health facility. The communities have the
mandate to choose their own community health workers, who are the main linkage between
the community and the nearest health facility.
2.3 Healthcare System of Tanzania
Tanzania health service provision involves both public and private health facilities. Tanzania
healthcare is divided into primary, secondary and tertiary levels. Primary health facilities are
responsible for healthcare provision such as curative, preventive, promotive, rehabilitative and
palliative care. Secondary services are provided at district hospitals, where most obstetric
interventions including surgery and blood transfusion are available. The district hospital is
also the first-referral hospitals for primary care in maternal and child services. Tertiary
services are provided in regional and national hospitals. These hospitals normally undertake
curative and rehabilitative services, but they also undertake a limited amount of preventive
care and health promotion. Regional hospitals provide specialized care that requires skills and
competences not available at district hospitals, which makes them the second level of the
referral system. The third level of the referral system is the national level which receives
referred patients from the second level of the referral system. Patients are referred up as well
as down through these levels, though lower levels can be bypassed for higher levels in the
event of emergencies and accidents. This study was conducted at the primary healthcare
facilities.
Primary healthcare services are provided in communities, dispensaries and health centers
responsible for providing health services to the population. The community level forms the
foundation of the service delivery system. A community is defines as a group of people living
in the same geographical area and under the same administrative system who share (more or
less) common social and economic conditions (MOHSW, 2010, p.35). In communities,
primary healthcare services are delivered by non-health professionals with supervision from
professional staff at a nearby health facility. Examples of non-health professionals who
provide services include community health workers, family members and a non-government
organization (MOHSW, 2010). The services offered range from care for chronic diseases and
include curative, psychological, or hygiene care at clients’ homes, and promotion of health
practices to pregnant women and the under-fives in their hamlets. The service provision
related to preventive care involves also the community leaders who mobilize, educate and
25
sensitize community members to receive facility heath services. All primary healthcare
services at the community level are supported by referral from community to a nearby health
facility.
Dispensaries and health centers promote healthy practices and provide preventive care and
curative services and are usually responsible for providing health services to a certain
(geographically-delineated) population. The services are provided by medical professionals.
Dispensaries are expected to serve a population of 6,000–10,000 people. The rural
dispensaries are staffed by a clinical officer and clinical assistant with one or two helpers, and
in some cases a nurse/midwife is also provided. The clinical assistant receives three years
training in anatomy, physiology and hygiene with a good grounding in diagnostic methods
and treatment of common diseases. The dispensaries refer complicated cases to health centers
which is the first level of referral system.
Health centers are expected to serve a population of at least 50,000–80,000 people. The health
centre is run by a clinical officer, with a secondary school education, and a more advanced
education in diagnosis and treatment as well as training in minor surgery. The clinical officer
is usually assisted by one or two clinical assistant, a nurse/midwife with one or two maternal
and child health aids, a health aid and a health assistant. Health centers give priority to
preventive measures and hygiene but, in practice, they are extensively used for the treatment
of common diseases. Most health centers have a room for minor surgery and provide 20–30
beds for in-patients, including maternity cases. Health centers augment their coverage with
outreach (mobile) services and refer severe and complicated conditions to district hospitals.
Reproduction and Child Healthcare (RCH) is one of the healthcare components provided in
almost all primary health facilities. Services provided in the section are antenatal care,
delivery, postnatal care, family planning and under-fives clinics. Figure 5 shows pictures
elaborating examples of maternal and child health services in Tanzania. The RCH section is
usually headed by one nurse who reports to the nurse-in-charge of the facility.
Consequently, PMTCT services are integrated into antenatal care, delivery, postnatal, and
child health management, which cut across all services in reproductive and child health
services. However, comprehensive intervention for PMTCT services involves collaboration
with home-based care services, and other clinics such as HIV/AIDS clinics’ Care and
Treatment Centres (CTC), and TB clinics (in case pregnant women develop TB). In every
26
facility there is one health provider designated as HBC (Home Based Care) coordinator who
has the inventory of all community-based service providers.
A
B
Figure 5: Illustration of maternal and child health services in Tanzania
A: Delivery ward.
B: Antenatal and Children Care Services in a facility
2.4 Official Information Systems that Support Maternal and Child Health
Services
The Government of Tanzania considers Health Information System (HIS) as one of the
keystones of an efficient health system (MOHSW, 2007a). In healthcare system structure,
every level collects and reports data in both routine and non-routine modes to the next
immediate level. HIS operating in Tanzania is known as MTUHA, an acronym in Kiswahili,
and was established in 1989 (Rubona, 2001). HMIS consists of both paper-based and
computer-based processing. Data generation starts from primary healthcare level, then is
aggregated and transmitted to the higher levels. Every section of the primary care services
have patient registers which are maintained by the facility in order to document the service
delivered.
At the community level, community volunteers or community home-based care providers are
required to have HBC notebooks for recording their activities with patients receiving services
(MOHSW, 2010). Also, every hamlet/street leader has residents’ register which records
information of household members, such as their name, age, household leader, and status (e.g.
child, wife, relative) (MOHSW, 2007a). The household information collected by the
community leaders is also used in identifying and providing health needs of their communities
such as health education and mobilization as well as social support for orphans and other
needy individuals. The hamlet chairperson is also responsible for updating the registers with
27
support from the village executive officer. Community information is reported to the nearby
facility (MOHSW, 2010).
The patient records which exist at the community level also include those that are generated
during home-based services, such as for chronic diseases like cancer and HIV/AIDS
(MOHSW, 2010). For example, in the management of tuberculosis cases, the clients’
treatment supporters are responsible for filling out the cards of the patients they supervise in
order to elaborate on the observation of treatment in the home settings (MOHSW, 2006).
Clients 14 maintain care identification cards which contain some medical records such as
encounter summaries, next appointment and patient demographics.
Facility-based health records generate data/information that could be used for local health
service delivery. Examples of service delivery registers include outpatient, antenatal,
postnatal, family planning, dental care, delivery, communicable diseases, and children
booklets. There are also other registers used, such as an inventory (e.g. ledger for equipment,
PMTCT drug dispensing register), laboratory services (e.g. blood sample collection form,
tuberculosis test request) and referral forms (general purpose referral form, PMTCT
transfer/referral form, infant referral form HIV clinic).
The data flow process goes from the community to the national levels. The collected data
have to be aggregated in each section on a monthly, quarterly and annual basis. The data
reported are in the form of rates, ratios and absolute numbers, and each has a threshold and
target value on which assessment of performance is based. Community information could be
reported to the nearby facility (MOHSW, 2010). Health facilities produce reports using the
collected information from every section on a monthly, quarterly and annual basis, and
information flows upwards through the CHMT level. For example, all facilities were
supposed to report to the district level on monthly basis about immunization services, such as
data on how many children were immunized for a particular vaccine. In the PMTCT program,
reporting is also done monthly and involves information such as the total number of children
registered in the PMTCT service in the current month, the number of children confirmed as
HIV-positive during the first, second and third testings, and the number of children referred to
the HIV clinic. There are also requirements to report on administrative data in the form of the
finance management system, human resource management system, drugs management system,
equipment, buildings, vehicles and other aspects of infrastructure. Districts obtain their raw
14
The terms clients, pregnant women, children and patients have been used interchangeably
28
data from the health facility reports, and compile them into quarterly and annual reports which
are forwarded to the regional levels. The regional levels prepare reports which would be
forwarded to the national level.
Data management at primary care facilities level is faced with challenges such as inadequacy
of healthcare service records, inadequate information analysis and use and dissemination at
health facility level (MOHSW, 2007a; Mushi and Marahaj, 2013; MOHSW, 2013).
Moreover, the reporting activity may be perceived to be a burden to healthcare service
providers at lower levels.
The government recognizes that information and communications technology is instrumental
in strengthening health service delivery and monitoring (MOHSW, 2007c, 2013; Darcy et al.,
2015). Computers are used for managing data at administrative levels. The DHIS system is a
national standard in Tanzania for processing aggregate data at district, regional, and national
levels. However, the use of computers to support health service delivery at health facility
levels is still minimal and is based only in chronic care disease management with the aim of
supporting report preparation. For example, in HIV/AIDS clinics in Tanzania, computer
systems have been deployed but not for the point-of-care data processing but, rather, for
supporting report generation. Also, the National AIDS control program has designed various
software tools to help health facilities and districts to better manage and use their data, and
produce the reports required by it. Moreover, there are plans to extend the use of the DHIS to
the facility level. Consequently, the DHIS Tracker had been piloted in Kibaha region (Ngoma
et al., 2012; Ngoma, 2014).
This chapter has described the research setting of Tanzania’s primary healthcare sector, with a
particular emphasis on the services of PMTCT and vaccination of children. In the next chapter
I will describe the related research guiding the study.
29
CHAPTER 3: RELATED RESEARCH
This thesis addresses topics in mainly three areas: health information systems of the
developing countries, collaborative research and participatory design. Accordingly, the
chapter is organized into three main subsections. The first part focuses on reviewing the
literature on HIS in developing countries. This is followed by the literature review of
distributed collaborative work around facilitating a healthcare trajectory with the focus of
understanding the actors involved, and the information tools they use to facilitate their
collaboration. The third part focuses on the participatory design within distributed design of
software to manage patient data.
3.1 Review of HIS Literature in Developing Countries
In the introductory chapter, I pointed out that HIS’s broad categories are patient-care
information systems and Health Management Information Systems (HMIS). In this section I
review the HIS literature of the developing countries and demonstrate that the literature has
less focus on patient-care information systems of the primary healthcare settings than on
HMIS. I will also show that the literature of the patient-care information system is oriented to
the design of artifacts rather than understanding work practices around the utilization of
Information and Communication Technologies (ICT).
3.1.1 Health Managerial Orientation in the HIS Literature
The HIS literature of developing countries recognizes that health facilities need to collect
accurate and complete information in order to offer the services effectively, and managers
need timely and correct information in order to govern, plan and support the service provision
well (Bodart and Shrestha, 2000; Lippeveld and Sauerborn, 2000). However, the need for
information differs between health workers and managers. Health providers collect detailed
client-oriented data from each incident of illness or client encounter in primary health
facilities. Such client information is used by health providers to manage the work itself
(organize services and ensure that appropriate services are provided to individual clients). The
information is also needed to support coordination and collaboration among multiple
professionals in managing the trajectories of individual clients. This kind of information
mostly flows horizontally among health providers (e.g. in clinics where health services are
offered), with clients, or with other stakeholders (schools, religious organizations and
influential community leaders) (Cutts, 1991; Sauerborn, 2000; Chadambuka et al., 2012).
30
Information may also flow vertically between healthcare delivery units when clients are
referred from lower levels to higher levels (Lippeveld, 2000).
In contrast, health managers need aggregated data from client encounters and other internal
management activities within primary healthcare settings (e.g. disease statistics, finance
expenditures, medicine supply and usage, etc.). This information flows vertically, routinely at
certain intervals (daily, weekly, monthly, annually) from health facilities to the next level of
authority (e.g. district) and then between levels of the healthcare system (e.g. from regional to
the national level) (Shrestha and Bodart, 2000). This information from health facilities is used
by health managers to monitor the performance of various health programs, supply resources
(e.g. health providers, finance and medical supplies) and also for policy- and strategy-making.
Health managers can also share such information with other consumers at the same level of
the healthcare system (e.g. healthcare development partners) and other sectors relevant to
health (e.g. agriculture, education, housing, finance and planning) (Sauerborn, 2000).
The research also acknowledges the important role of peripheral health workers for HMIS, as
a well-working HMIS depends on timely, correct and comprehensive reporting from the
frontline workers to be able to generate an accurate picture of the disease burden, service
situation and resource needs. Yet, the common problem discussed in the HIS literature is the
problem of data quality which manifests in irrelevance of information gathered, poor use of
information, and lack of timely, correct and comprehensive reporting (see for example,
Sauerborn and Lippeveld, 2000; Chilundo et al., 2004; Moyo, 2005; Simba and Mwangu,
2005; Mphatswe et al., 2012; Mlambo et al., 2014; Ledikwe et al., 2014). There are two
opposing insights given by HIS researchers who explain the shortcomings with data quality
problems. The research conducted at the primary healthcare level explains that reasons for
poor information practices are related to the problems of a policy that restricts traditional birth
attendants to conduct deliveries (Kanjo, 2011), the use of non-standardized tools (Damtew
and Moges, 2013), the lack of tools (Igira, 2008a; Ngoma and Igira, 2012), and poor/weak
design and support of existing information systems (Heunis et al, 2011). In contrast, other HIS
researchers point how the problem of poor quality data is related to issues arising from health
workers, such as health workers being unaware of the importance of a health information
system, health staff shortages, high staff turnover, lack of information system skills in some of
them, and poor motivation among health workers (Lippeveld, 2000; Sauerborn and Lippeveld,
2000; Heywood and Rohde, 2001; Igira, 2008; WHO, 2008). The literature also mentions that
31
the problem could be related to health management information systems’ weak procedure and
practices, such as, multiplicity of reporting channels, lack of feedback from higher levels, and
lack of coordination between the data collection sites for different PHC programs at the grass
root (Chilundo et al., 2004; Farah, 2010; Mlambo et al., 2014; Sychareun et al., 2014).
While the importance of improving HIS to become the tool for improving healthcare has long
been identified in the literature (Sauerborn and Lippeveld 2000), there is less emphasis on
studying health providers’ work practices than on the managerial aspect within primary
healthcare information systems in the HIS research field. I will now demonstrate two areas
which are rarely discussed in the literature. For example, the tension between patient-related
and reporting-related health information systems has received little research attention. It is
acknowledged by researchers in the developed world that there could be conflicts between
primary and secondary uses of client information (Garfinkel and Bittner, 1967; Heath and
Luff, 1996; Berg and Goorman, 1999). The source of conflict is because the primary needs of
information (providing quality healthcare to individual clients) are shaped by the pragmatic
considerations of the context in which they are generated (Health and Luff, 1996; Berg and
Goorman, 1999). For instance, an experienced medical doctor that examines a routine case
may skip filling in several data fields in the system in order to attend to clients quickly, and,
while this is an appropriate decision and not misconduct, it means that data cannot therefore
be directly used for secondary purposes, such as by managers and researchers.
Also, HIS research that discusses the efforts of improving information systems has focused
more on the needs of higher levels (managers) of the healthcare hierarchy (national, regional
or district levels), and less on finding ways to improve information management at the
(facility) operational level. The research shows various efforts to improve HIS so that it can
become a tool for managing the health system (see for example Smith et al., 2007; Sæbø et
al., 2011; Kossi et al., 2012). Examples of such efforts are integration and standardization of
information systems (Nyella, 2009; Abdusamadovich, 2013; Poppe et al., 2013), training
health workers at the district and national levels (Kimaro, 2006; Ngoma, 2007; Ngoma et al.,
2008), and improving information quality through peer-review meetings (Adaletey et al.,
2015). There are also efforts at training health workers at facility and community levels in
order to improve information flow from lower to higher levels (Ngoma and Igira,
forthcoming; Mphatswe et al., 2012; Wagenaar et al., 2015). I argue that, although the
literature has long acknowledged that a well-designed information system will support the
32
needs of both healthcare deliverers and health managers (Sauerborn and Lippeveld, 2000), the
research on HIS seldom focuses on related operational needs at the primary healthcare level.
The focus on operational needs as one way to improve HIS for healthcare delivery services
and not just for managerial needs requires us to also look at health facility work practices. In
the next section, I review the patient-care-oriented HIS literature to demonstrate that it is
oriented more to technology design and less on understanding work practices of health
providers.
3.1.2 Patient-Care Oriented HIS literature
There is an accent on technology in the literature that deals with patient-care needs. A number
of studies have been undertaken to understand how ICT can be designed to support healthcare
service at the health facility levels. These studies have provided an understanding of what
functionality or information technologies could be implemented to support healthcare
delivery. Ngabo et al. (2012) illustrate the design and implementation of a mobile phonebased system that aimed at monitoring pregnancy and reducing bottlenecks in communication
associated with maternal and newborn deaths in Rwanda. The mobile phone-based system
used an SMS
15
-based platform in order to enable effective and real-time two-way
communication between community health workers at community level and other healthcare
levels. Likewise, DeRenzi et al. (2008) discuss how to facilitate management of child health
through the design of personal digital assistant software, which would help clinicians make
decisions and adhere to child treatment protocol in Tanzania. In a similar manner, Avilés et al.
(2007) describe the introduction and design of integrated computer-based systems intended to
support screening of dengue virus for children attending primary care in Nicaraguan settings.
Further, research with a focus on the support of patient care shows what benefits can be
achieved by using certain kind of digital IS, such as improving patient data security, graphic
visualization to aid clinic officers in making diagnoses, real-time analysis of data, and
improving flow/quality of information (Avilés et al., 2007; Luk et al., 2008; DeRenzi et al.,
2011; Hindemark, 2013; Kabuya et al., 2014). The benefits of the use of a digital system can
also be related to facilitating healthcare service delivery, such as improving oversight of
patient activity along the care trajectory (Hindemark, 2013), tracking blood samples between
laboratories (Avilés et al., 2007), and improving adherence to antiretroviral treatment in HIV
clinics (e.g. Fraser et al., 2007; Forster et al., 2008; Pop-Eleches et al., 2011; Thomson et al.,
15
Short Message Service
33
2011). It is also suggested that the use of computerized patient records can improve
information management and its quality within health facilities (Karara, 2013; Kabuya et al.,
2014).
My literature review shows that research at primary healthcare facility level has tended to
favor design of the functionality of the Information system and has focused less on gaining an
understanding of the work practices around the use and implementation of ICT to support
healthcare provision in the developing countries (Kaplan, 2006; Fraser et al., 2007; Lester et
al., 2010). Studies which evaluate the use of mobile phones to improve the adherence of
attendance by clients in the HIV/TB clinics (Lester et al., 2010) and the analysis of effects of
EMR deployments in antenatal clinics (Chawani, 2014a) are exceptions. Further, there are
few pilot studies which examine the effect of an information system in the healthcare services
in the setting of developing countries (see for example; Haberer et al., 2010; Hoffman et al.,
2010).
A focus on work practices can help researchers and designers gain understanding of the
processes of interaction between practitioners and their use of the artifacts in their
collaborative activities and can help to gain an understanding of the problems that workers
face (Blomberg et al., 1996; Karasti, 2001a). However, there is a dearth of HIS research
which provides insights on the utilization of patient-oriented information systems to support
healthcare provision in the developing countries (Kaplan, 2006; Fraser et al., 2007; Lester et
al., 2010). As a consequence, there is limited understanding on work practice regarding how
health providers interact among themselves and with information system artifacts to support
their collaborative care provision and how such understanding can be used to improve the
design of the patient-oriented information systems. For example, it is possible to see from the
above literature that there are limited insights on how health workers utilize patient-oriented
tools to support their collaboration in healthcare services. Hence, there is a need to develop a
more solid understanding of work practices. In particular, there are two thematic limitations
from the insights provided by the HIS literature related to patient-care information systems:
collaborative work practices around facilitation of care trajectory and the role context plays in
practitioners’ work practice.
Collaborative practices in the patient-care HIS literature: In the literature of health
information systems in developing countries, studies of collaboration in healthcare have
empirically investigated the use of telemedicine and the organization of software development
34
activities. For instance, telemedicine systems can be used to support collaborative practices by
facilitating communication between providers (sometimes with patients), the sharing of
medical information, and also educational purposes (Geissbuhler et al., 2003; Bath, 2006;
Bagayoko et al., 2008; Mars, 2010; Martinez-Alcala et al., 2013). Also, a number of studies
examine collaboration in the context of software development activities between different
software professionals who could be in different locations or with users and implementers
(Ijadunola, 1998; Korpela et al., 2002; Saugene, 2013; Chawani, 2014b; Saugene and
Kaasbøll, 2014). Hence, the above empirical studies indicate how technologies can be used to
support collaboration among health providers.
However, existing research of HIS in developing countries have paid less attention to the
understanding of collaborative practices in the context of healthcare services. In the review
conducted by Fitzpatrick and Ellingsen (2013), only three publications reported studies
conducted in Africa. In the first study, Cheng and colleagues assess residents’ attitudes toward
the use of handheld computers in the collection of HIV/AIDS survey data in Angola (Cheng
et al., 2008). They compare residents’ attitudes toward handheld computers with their
attitudes toward the standard mode of data collection: paper surveys. The authors conclude
that computerized data collection in Sub-Saharan Africa may lead to biased reports of
HIV/AIDS-related risk behaviors when compared to the traditional method of paper-andpencil surveys due to social-economic factors. In the second study, DeRenzi and colleagues
discuss how to facilitate management of child health through the design of software,
specifically a Personal Digital Assistant (PDA) system which would help clinicians make
decisions and adhere to a child treatment protocol (DeRenzi et al., 2008). During their study
in Tanzania, which focused on human-computer collaboration, some features from PDA
software had to be redesigned in order to match the observed human-PDA collaborative
practices. In the third study, Luk and colleagues describe how providers collaborate with the
use of telemedicine in Ghana and provide principles to inform the design of telemedicine
systems in developing regions (Luk et al., 2008). The remote system enabled doctors to enter
medical case information into a distributed repository and then request a consultation from a
pool of medical specialists. As illustrated in these three studies, in healthcare service
provision, coordination is important and can be improved by designing and implementing
digital information systems.
35
Impact of context on HIS: It is acknowledged that the understanding of context is important in
order to improve the design and implementation of HIS (Igira, 2008a; Avgerou, 2010;
Tiihonen et al., 2006; Tiihonen, 2009, 2011; Kanjo, 2012; Tswane et al., 2015). Accordingly,
some researchers propose a theoretical model/framework for understanding context during
design and development of IS. For example, Asangasi (2013) suggests that institutional theory
is useful for gaining an understanding of context. Tiihonen and associates suggest that the
LACASA (Levels of Analysis, Categories of Analysis, and Scopes of Analysis) model can be
used to aid identification of contextual features during information systems development
(Tiihonen et al., 2006; Tiihonen et al., 2010; Tiihonen, 2011). Nevertheless, the majority of
these works address aspects of context that are relevant specifically to design and
implementation of ICT, such as how (computer) literacy and limited material resources have a
direct impact on the design and development activities (Tongia and Subrahmanian, 2006;
Lund and Suutinen 2010; Sinnig et al., 2010; Pitula and Radhakrishnan 2011). On the other
hand, Macaulay et al. (2000) claim that contextual features of the use situation may include
environmental, social, cultural, individual and historical factors affecting how people manage
their work practices. Similarly, primary healthcare is influenced by contextual factors such as
geography,
history,
culture,
infrastructure,
economic,
social-cultural
and
political
characteristics of the place where it is based (WHO, 1988; Byrne and Sahay, 2007). Thus, we
need more understanding of the impact of work-oriented contextual conditions and how it can
be used to improve the design of primary healthcare information systems.
The HIS research gives a general impression that context impacts the use and implementation
of information systems supporting healthcare delivery. Lippeveld and Sauerborn (2000)
suggest that information systems developed in an urban area where most health services are
offered by the private sector, and the majority of the population are literate, would certainly
be different from the those where the majority of the population is illiterate and the health
services are predominantly government-managed (p.15). Also, Smit and Venter (2009),
reporting on the design and implementation of clinical health information systems for primary
healthcare clinics in rural South Africa, indicate that, after three months of system
implementation, the uptake of the system was challenged by a lack of computer literacy in
some health workers and resistance from patients who did not like the capturing of their
biometric data. Moreover, a higher workload within health facilities influenced the use of
designed child management software since health workers could bypass treatment protocols
they perceived to be time consuming (DeRenzi et al., 2008). Likewise, Dwolatzky et al.
36
(2006), outlining the design of global positioning system and PDA systems that could be used
to improve tracking of TB patients in South African settings, point out that as many clients
live in informal settlements, the absence of formal addresses that can be captured and
processed limit the usefulness of system to support tracking of clients.
The design practices can also be influenced by contextual condition. Puri et al. (2004)
describe how the process of design and development of a child health community-based
Information System in a South African country was shaped by aspects of socio-economic,
cultural and political situations. Ngoma et al. (2012) noted that health workers had to change
their information practices in order to fit the practices of clients who did not attend their
vaccination appointments as scheduled. Likewise, Chawani et al.’s (2014a) report from the
analysis of the antenatal work practices indicated that there were discrepancies between
protocol and actual work practice, the reasons being that some clients did not know their last
date of menstruation and, hence, were not able to provide information that was important for
calculating their gestation age (in weeks). In both these cases studied by Ngoma et al. (2012)
and Chawani et al. (2014a), the computer system had to be redesigned to fit the contextual
issues.
From practice-oriented studies, context of use plays a crucial role in organizing activities (Star
et al., 2003; Oborn et al., 2011). The reviewed HIS literature also provides a general
impression that there is interdependence between context of use and the design,
implementation and development of information systems. While it is acknowledged that a
deeper understanding of context can help identify what to change and how to intervene in the
real settings (Igira, 2008a; 2008b); it is not possible to see from the HIS literature how this
gained understanding of context can be used to inform the design of patient-care information
systems in developing countries. I thus intend to enrich the understanding of influence of
work-oriented context in the settings of primary healthcare information systems and provide
the implications for the design, implementation and use of information systems.
3.2 Collaborative Work
Collaborative work can be defined as being where different actors work towards a common
goal (Schmidt and Banon, 1992). People engage in cooperative work when they are mutually
dependent in their work and are therefore required to cooperate in order to get the work done
(Schmidt, 1991). There are different types of interdependency among actors in collaborative
37
work – sequential, pooled and reciprocal (Thompson, 1967). In sequential interdependence,
the activities of some individuals must be performed before other individuals’ activities. For
example, in work to develop the software to support the health of children, understanding of
work practices of its health provision must be acquired before the activities of design and
actual programming of the system take place. Another example is that for PMTCT care a
pregnant woman must be first diagnosed as HIV-positive in antenatal care before she can
access PMTCT services. In pooled interdependency, the individual work can happen
concurrently and contribute to the overall goal. Example of pooled interdependency is the
healthcare practices in PMTCT where a pregnant woman is attended by nurses in antenatal
care and at the same time receives care from the HIV clinic. In reciprocal interdependency,
individuals work together and their work relates to each other as both input and outputs.
Example of reciprocal interdependency is caesarean care which depends on the work of the
anesthetist, the doctor who performs the surgery and the nurse who helps the doctor by giving
medical tools or providing medical care for example, during the procedure.
However, in order for the collaborative work to be successful, interdependent tasks of
different professionals need to be aligned with each other; for example, in how individual
goals are related to the bigger goal facing the overall team (Kaptelini and Nardi, 2006). Health
workers have to be aware of the actions of other participating individuals and adjust their
actions to those of other people. However, task interdependencies are tractable only if
appropriate coordination mechanisms are in place (Schmidt and Simonee, 1996). A
coordination mechanism is “a coordinative protocol with an accompanying artifact such as for
instance a standard operating procedure supported by a certain form” (Simone and Schmidt,
1998, p. 295). The use of coordination mechanisms by actors participating in cooperative
work might assist them to reduce the complexity of managing their task interdependency
(Schmidt and Simonee, 1996; Simone and Schmidt, 1998).
There are numerous examples of coordinative artifacts that have been designed to support
collaborative work such as, group calendar systems, scheduling systems, booking systems,
document management systems, systems designed to support work flow within groups,
computer mediated communication (e.g. desktop conferencing, video conferencing, and
email), and automated meeting rooms with a network structure to support voting systems
(Grudin, 1994; Bowers et al., 1995; Dourish et al., 1996; Schmidt, 2011). The coordinative
protocols are usually incorporated within the coordinative artifacts (Schmidt, 2011).
38
Collaboration practices has been researched within the many organizational settings,
including urban planning and design, healthcare, architecture, software engineering and IT
work, e-science and cyber-Infrastructure (Star and Ruhdler, 1994; Schmidt and Simone, 1996;
Berg, 1997; Hersbleb et al., 2000; Bardram, 2000; Aanestad, 2003; Grinter, 2003; Vikkelsø,
2005; Lee et al., 2006; Schmidt et al., 2007; Maquil et al., 2007; Reddy et al., 2010; Dittrich
et al., 2009; Jirotka et al., 2013). However, researchers have been criticized for conducting
studies that are of small-scale, co-located groups rather than distributed collaboration
(Monteiro et al., 2013). In addition, the literature has been criticized for paying less attention
to studying clients’ care trajectories distributed in multiple settings (Fitzpatrick and Ellingsen,
2013). Thus, Fitzpatrick and Ellingsen (2013) suggest that researchers expand their fieldwork
settings by following the patient trajectory around multiple settings in which care is provided.
Furthermore, the literature offers limited insights on collaborative practices around clients’
care trajectories in resource-restricted settings. Fitzpatrick and Ellingsen (2013) suggest that a
context of low resource settings can complicate the collaboration among health professionals
within developing countries since there is significantly lower access to resources than in the
developed world (Fitzpatrick and Ellingsen, 2013, p.33). For example, there are potential
shortages of human resources, and medical material and advanced ICT. Also the health
systems rely heavily on donor support due to the existing low GDP. Moreover, the
populations are illiterate and have less income relatively compared to developed world. In
Europe, the collaboration between distributed providers, for example in diagnosis and
treatment of cancer, may be facilitated by the use of advanced technologies such as video
conferencing, making it possible to facilitate interactions among them (Li et al., 2008; Kane
and Luz, 2009; Robertson et al., 2010). Hence, challenges to collaboration in a context with
fewer resources and infrastructures need more attention, empirical research and understanding.
Thus, this study focuses on collaborative work to facilitate care trajectories. I use the concept
of coordination mechanism for developing the understanding of collaborative practices of
multiple actors who are involved in facilitating clients’ trajectory within primary healthcare
settings.
3.2.1 Collaboration in Facilitating Clients’ Trajectories
In this section I review the collaborative literature with an emphasis on studies that address
collaboration and coordination along the patient trajectory in particular. Client trajectory can
be defined as the client’s journey through different places while accessing care (Strauss,
39
1991). As such, client trajectory accomplishment involves multiple clinical professionals and
often of different specialties, such as physicians, nurses, pharmacists, laboratory technicians,
radiologists (Reddy et al., 2010). Moreover, non-clinical professionals are involved in the
management of the client’s trajectory. For example, administrative staff are involved
facilitating patient transfer in inpatient admission care services in hospitals (Abraham and
Reddy, 2008; 2010). Common to other collaborative work, health providers also need to
manage their interdependency; for example, by inter-relating their work tasks (Berg, 1999b).
The interdependencies of collaborators involved in client trajectories are mediated by health
guidelines and patient-oriented information tools (Berg, 1999a). Health protocols or
guidelines guide providers on which treatment/care standards should be followed in certain
healthcare situations (Engestrom, 2001; Cabitza et al., 2009; Amir et al., 2015). Health
guidelines are usually developed by authorities such as countries, legal bodies or states.
Patient-oriented information is distributed through patient records and other patient-oriented
artifacts which exist in different places where a client’s trajectory takes place.
Patient records contain different types of client information: for example, past medical
history, physical assessment, medication administration, referral, immunization and
population based health records (Häyrinen et al., 2008). Patient records are contained in
different information artifacts, including progress notes, referral letters/forms, prescription
charts, clients’ cards, facility registry books, and radiology/laboratory reports (Wyatt and
Wright, 1998; Mønsted et al., 2011).
Patient-oriented information tools play different roles in facilitating collaboration in both
short-term and long-term interactions among collaborators of care trajectories (Bardram and
Bossen, 2005). Some patient-oriented information tools are used to support short-term
collaboration among health professionals. Examples include use of an appointment sheet,
whiteboard, schedules and plans (Bardram, 1998; Bjørn and Hertzum, 2011). Patient records
provide an overview of care provided by other actors involved in the care trajectory (Ren et
al., 2008; Robertson et al., 2010). Also, patient records facilitate the interaction between
clients and health providers during clinical treatment and care processes (Berg, 1996;
Pedersen and Ellingsen, 2011). Moreover, patient records are used to support the coordination
task by non-clinical staff, such as within patient transfer activities (admission, discharge and
transfers) (Abraham and Reddy, 2008). Additionally, patient records are used to support
communication among collaborators in different departments. For example, drugs prescription
40
information written by the physician in the outpatient section will guide the pharmacist in the
pharmacy section on the types of drugs to dispense to the clients. Moreover, electronic
medical records can be used to create reminders which can prompt providers to perform
certain healthcare tasks due for patients (Ash et al., 2004; Patterson et al., 2005; Mayo-Smith
and Agrawal, 2007).
Patient records are also important for secondary purposes besides supporting patient
trajectories (Garfinkel and Bittner, 1967; Berg and Goorman, 1999; Pedersen and Ellingsen,
2011). For instance, patient records are used to facilitate administrative and reporting
functions which can facilitate decision-making practices of health administrators and policy
makers (Martin et al., 2005). Also, patient records can be used for research purposes
(Garfinkel and Bittner, 1967).
Besides the presence of patient-oriented information artifacts, various information practices
need to be conducted by collaborators in order for the collaborative practices to be successful
(Berg, 1999a). For example, information flowing and sharing between health providers is an
important aspect in enhancing the management of patient trajectories. Health providers also
need to perform information-related activities such as marking, checking, reading and writing
patient-oriented information artifacts. These activities also contribute to the coordination of
collaborative activities of multiple practitioners across different times and places. Furthermore,
medical records staff have to make sure that information in patient records is complete,
correct and can be accessed easily (Bossen et al., 2012).
The facilitation of clients’ care trajectories can be done by distributed health providers in
health facilities (e.g. Robertson et al., 2010; Li and Robertson, 2011), or between health
providers and home care services (e.g. Piras and Zannuto, 2010). It is also common that the
client trajectories, especially in chronic disease management, move to home and community
settings (Kovner et al., 1996; Dalsted, 2012) and thus, facilitation of client trajectories require
the work of non-health professionals. For example, in telemedicine practices, clients may be
required to perform certain care activities (e.g. their own weight- and blood checks) and
communicate the results to health providers (Nicolini, 2013). Moreover, clients and other
people perform information-related activities in order to contribute to the coordination of their
care trajectories and clients and other people can coordinate their care trajectories by filling in
diaries, preparing documents, and managing their personal health records (Piras and Zanutto,
2010). Hence, in chronic disease management there are more actors than in co-located settings
41
who are involved in the coordination of care trajectories in different ways. Moreover,
coordination of collaborative activities is also done by non-health professionals.
From the above discussion, there are multiple coordination artifacts and protocols that are
used to facilitate care trajectories. Moreover, multiple actors are involved in the facilitation of
trajectories– both clinical professionals and non-clinical professionals (e.g. medical recorders
and patient transfer staff). These actors could be co-located or distributed in different places
such as between health facilities and home settings. However, chronic disease management
involves more actors than in co-located settings.
3.2.2 Challenges of Collaborative Work in facilitation of Clients’ Trajectories
The literature acknowledges that the design, use and implementation of coordination artifacts
and protocols is challenging. For example, Cabitza et al. (2006) assert that artifacts may not
be rich enough to facilitate communication between health providers involved in the
management of care trajectories. The deficit in coordinative artifacts may lead to the
development of informal artifacts (Mellini, 2013). For instance, Fitzpatrick (2004) describes
how clinicians tailored and augmented patient records in order to support their own roles in
care trajectories and practices. Likewise, Safad and Faraj (2010) show how workarounds
emerged after the implementation of Electronic Patient Record (EPR) system in ambulatory
clinic because the software did not have facilities for adding new diseases to the list of
common diseases. Accordingly, providers used medical history field to record new diseases to
support the care.
Researchers acknowledge that involvement of multiple stakeholders and diverse professional
groups increases the complexity of cooperative work (Hartswood et al., 2003; Fitzpatrick and
Ellingsen, 2013). This is because complexity increases due to increased coordination work in
managing the interdependences among actors, activities and goals that need to be aligned to
fit task of the collaborative ensemble (Carnsten and Sørensen, 1996). Subsequently, it might
be difficult for EPR systems to satisfy multiple different specialty demands due to the
variation of their individual practices and, consequently, health providers may need to adjust
the use of the systems in order to fit their practices and also to support collaborative practices
(Oborn et al., 2011).
However, a number of researchers acknowledge that distributed settings offer coordination
challenges related to how to maintain awareness of other actors who are not in the co-located
settings (Reddy et al., 2001). The strategies which can be used by collaborators include
42
informal talks, monitoring body activities and using artifacts (Heath et al., 2002). Moreover,
physical co-location of collaborators can make the work of other actors visible. For example,
it offers the opportunity for collaborators to resolve any breakdown or get clarification on
some issues. In contrast, the coordination work between teams involved in distributed care
trajectory can be challenged by a lack of awareness of other actors’ activities. Accordingly,
other coordination challenges can emerge in distributed collaborative practices due to
inadequate coordination protocols, information management practices and contextual
conditions.
Coordination protocols: It is possible that health protocols will not support care trajectories
between health facilities (Engestrom, 2001; Amir et al., 2015). For instance, it might require
the formulation of new health protocols to facilitate collaboration among health units involved
in care (Engestrom, 2001).
Information management practices; There could be poor information sharing between the
groups involved (Abram and Reddy, 2008). Team members can lack a shared view of the
clients’ care because they are unable to access each other’s clients’ records if providers
belong to different organizations (Cabitza et al., 2009; Amir et al., 2015). Also, patient
records can be fragmented in the distributed institutions involved and providers thus do not
have clear picture about clients’ conditions (Cabitza et al., 2009; Bansler et al., 2011).
The challenges related to information management in distributed collaboration can be
addressed through different strategies. Furthermore, Cabitza et al. (2009) indicate that health
providers may improve the overview of care trajectories through peer interactions; i.e.,
communication among themselves. Distributed collaborators may be supported by advanced
collaborative technologies; for example, video conferencing tools or videophone system can
be used to support multidisciplinary team meetings for diagnosis and treatment of chronic
diseases (e.g. Robertson et al., 2010). The challenge related to slow information flow can be
addressed by requiring clients, their family members and other people to become involved in
the information flow needed to facilitate the care trajectory among health providers situated in
different health facilities (Piras and Zannuto, 2010; Amir et al., 2015). Accordingly,
collaboration is mediated through both artifacts and human mediators.
Contextual conditions; Context can be defined as “any information that can be used to
characterize the situation of entities (i.e., whether a person, place, or object) that are
43
considered relevant to the interaction between a user and an application, including the user
and the application themselves” (Dey et al., 2001, p.100).
A number of researchers recognize that contextual conditions can support collaborative
practices. For instance, researchers recognize that a coordination task among health providers
can be improved through the design of context-aware systems (Doryab and Bardram, 2011).
“Context-aware” artifacts are digital artifacts capable of automatically adapting their
operations to fit the existing contextual information (Baldauf et al., 2007; Wischweh and
Bade, 2012). Doryab and Bardram (2011) describe the design of an activity-aware
recommender system used to support the activities of a laparoscopic surgical team. For
instance, the designed system was capable of recognizing the current activities of the
laparoscopic surgical team, and then used such knowledge to suggest the next the action to be
undertaken by team members.
In contrast, other researchers acknowledge that the collaborative work can be limited by
contextual conditions. For example, physical settings have been identified as causing a
variation of interactions between collaborators and the use of coordinative artifacts in
distributed collaborative practices (Robertson et al., 2010; Li and Robertson, 2011). Thus,
subsequently, interactions between distributed collaborators need also to be mediated by
humans.
Context can also limit design and implementation of EPR systems. For example, context can
limit successful customization or implementation of the software to manage client information.
It has been identified that configuring standard packages of EPR software to systems to meet
health facilities’ local practices can be difficult (Martin et al., 2005; Koppel and Lehmann,
2014). Moreover, the configuration of the same software in different settings can yield
different experiences and lessons. This was identified by Koppel and Lehmann (2014), who
report that the standard package with the same language translations (English and Spanish)
was implemented within two nearby hospitals, yet yielded significant differences in data and
interface customization outcomes (Koppel and Lehmann, 2014). Accordingly, the contextualcondition influence may result in a higher demand for the customization of the standard EPR
package or result in resistance from users (Safad and Faraj, 2010) or in delays or unsuccessful
or undesirable outcomes (Harrison et al., 2007). On the other hand, the local contextual
conditions may not be visible during the analysis and design of EPR standard packages, yet
44
the effect will be felt when the implemented software does not work according to the original
understanding of the work flow (Harrison et al, 2007; Smith and Koppel, 2013).
Researchers propose that a better understanding of context can be used to facilitate design and
implementation of collaborative systems (Ciolf et al., 2008). In relation to the healthcare
information systems, researchers recommend that designers accommodate local variability of
healthcare work practices caused by contextual conditions in the configuration of standard
software for managing healthcare services (Balka et al., 2008). Previous CSCW research has
not paid much attention to studying the impact of other contextual features on collaborative
practices (Harrison and Dourish, 1996) and researchers are concerned that contextual
understanding is mostly equated to physical space or as the organizational context that
influences the work practices (e.g. Harrison and Dourish, 1996; Ciolfi et al., 2008). The
understanding of collaborative practice and how it is influenced by contextual factors might
improve the knowledge how ICT tools can be used to improve healthcare services such as
retention of clients.
The aim of the CSCW field is to understand the nature of collaborative work in its natural
settings with the objective of designing computer-based technologies to support work
practices. The aim of work practice research is to understand practitioners’ actual practices as
a basis for technology design (Luff et al., 2000; Karasti, 2001a). One strategy for
understanding practitioners’ practices is to employ ethnographic studies which can then be
used to inform the actual design of the technologies, or for spotting the implications for
design (Blomberg and Karasti, 2013). The ethnographic study can be combined with the
participation of practitioners who could contrast actual design of the technology (Karasti,
2001). Moreover, it is recommended adopting the participatory design approaches in the
design of collaborative systems in order to make the system fit the local contexts of
practitioners (Kensing and Blomberg, 1998). In the following section, I review the literature
on PD.
3.3 Distributed Participatory Design in Patient-Care Software
Design is a socio-technical process which involves changes of artifact, people, organization
and practices (Gregory, 2003; Barcellini et al., 2015). Moreover, the actual work of designing
an information systems artifact is the outcome of collaborative activities of many people, such
as IT professionals, managers, training institutions, donor organizations, NGOs, and other
45
stakeholders (Bødker et al., 1988; Kyng, 1991; Dittrich et al., 2014). Also, practitioners can
contribute to the design process by evaluating design suggestions; for example, by contrasting
old and newer work practices (Karasti, 2001b). The implementation of ICT systems also
requires close collaboration between users and IT facilitators (Markussen, 1994; Hartswood et
al., 2002; Bandon et al., 2005). Thus, the design, implementation and development of patientcare software require close collaboration between practitioners and IT professionals for its
successful implementation.
The Participatory Design (PD) approach emphasizes sharing of decision making in the design
process between all IT professionals and non-IT professionals, (Kensing and Blomberg, 1998;
Kensing and Greenbaum 2012; Brandt et al, 2013). The rationale for emphasizing
participation of non-IT professionals in the design process is related to both democratic and
functional empowerment goals (Bjerknes and Bratteteig, 1995; Iivari, 2011; Robertson and
Simonsen, 2012). The aim of democratic empowerment is so that practitioners who will be
affected by the new IT system have the right to have a say in the design process. Functional
empowerment aims at practitioner participation, which may lead to the development of a
better ICT product that will improve their working situation (Suchman, 2002; Gregory, 2003).
Other core elements included in PD are situatedness and mutual learning (Kensing and
Greenbaum 2012, p. 33). Situatedness relates to the recognition of practitioners’ skills in the
design activities, and working with them in their actual settings in order to improve their
participation in the design process (Suchman, 2002; Kensing and Greenbaum 2012; Van der
Velden and Mortberg, 2014). Mutual learning aims at creating shared understanding between
practitioners and designers in organized settings of collaboration, whereby practitioners try to
learning technological features and IT professionals strive to learn about users’ work practices
(e.g. Kyng, 1991; Mursu et al., 2007; Bratteteig et al., 2012; Robertson and Simonsen, 2012).
Various tools and techniques have been developed to facilitate collaboration between IT
professionals and practitioners so that mutual learning might occur between the two groups.
IT professionals can learn about users’ practices by using ethnographic approaches (Karasti,
2001; Hartswood et al., 2002; Spinuzzi, 2005). Users and IT professionals can learn from
each other, for example, through workshops, design games, and cooperative prototyping
(Gregory, 2003; Johansson and Linde, 2005; Bratteig et al., 2012).
Previous research within the PD community recognizes the importance of design facilitators
who enable dialogue and collaboration in situated participation practices (Kensing and
46
Madsen, 1992; Light and Akama, 2012). Light and Akama (2012) argue that it is necessary
not only to pay attention to how methods matter but also “how designers practice
participation”. Light and Akama (2012, p. 62) characterize design facilitation as a continuous
process which cannot be foreseen.
Previous research also acknowledges that design continues in use and there is a need for
participatory designers to focus on design in use (e.g. Henderson and Kyng, 1991; Ehn, 2008;
Bratteig et al., 2012). However, Ehn (2008, p. 96) is more explicit in his focus on design-inuse and writes:
Rather than focusing on involving users in the design process, focus shifts towards
seeing every use situation as a potential design situation. So there is design during a
project (‘at project time’), but there is also design in use (‘at use time’). There is
design (in use) after design (in the design project).
Bratteteig et al. (2012, p. 138–39) also reflect upon “design-after-design.” The rationale of
their discussion is a shift in systems design towards software and e-services built on platforms
that allow further development. Tailoring possibilities or postponing some design decisions to
design-in-use is Bratteteig et al.’s (2012) suggestion to deal with the transformation. While
previous research acknowledges the importance of designers paying attention to design in use,
most of the participatory design practices have focused on design-before-use. The focus in my
research is design-in-use, with an emphasis on design facilitators. I will now demonstrate why
more understanding on this theme is needed.
New technologies and application domains other than workplace settings have affected the
use of PD approaches in the collaborative design (Karasti, 2010; Bratteteig et al., 2012). In the
text, I will focus on two application domains: developing countries and the distributed design
environment. The use of PD approaches in the settings of developing countries has met
challenges in the participation of practitioners related to both healthcare and specific IT skills
(Nhamposa et al., 2004; Kimaro and Titlestad, 2008; Chawani et al., 2014b). There are also
contextual challenges in the use of PD approaches in developing countries; for example, busy
schedules of health practitioners have made for their limited involvement and, as a
consequence, the design process can be dominated by IT professionals (Elovaara et al., 2006).
In order to enhance the participation of practitioners, several strategies have been
recommended, such as IT facilitators training practitioners in computer skills, and altering
participation through extended (& recurring) interactions with other collaborating actors
47
during the design process (Puri et al., 2004; Winschiers-Theofilus et al., 2012; Chawani et al.,
2014b). However, the use of PD approaches within such a design environment has been
recognized as a challenging endeavor due to the geographical distribution of collaborating
actors (Titlestad et al., 2009; Iivari, 2011).
Overall, the literature review shows that the PD literature seldom focuses on design-in-use
practices. Meanwhile, it is now common for healthcare organizations to adopt and use
standard Open Source Software (OSS) packages such as ClearHealth, FreeMed, OpenEMR ,
WorldVistA, Epic, Care2X and DHIS Tracker (Faus and Sujansky, 2008; Gizaw et al., 2012;
Koppel and Lehmann, 2014). The design, development and use of OSS applications involve
multiple actors who may be in geographically-dispersed work settings (Iivari, 2011). Globally
distributed software developers will create generic software whereby some of its features are
left open (half-finished product) to be finished in workplace settings (Pollock et al., 2003;
Pries-Heje and Dittrich, 2009; Gizaw, 2014). The design of the generic software will be
finished in the workplace by configuring it to match organizational characteristics (Titlestad
et al., 2009). The design activities done in work place settings include tasks of setting up of
data elements, redesign of standard forms and reports, and specifying workplace workflow
processes (Saugene, 2013; Ngoma et al., 2012; Chawani, 2014b). Thus design of OSS
involves design-after-design practice. In some cases, the re-configuration of generic software
might not be sufficient to fit organization tasks and thus customization activities will be
conducted as a next step (Pries-Heje and Dittrich, 2009). Customization of the software
involves the change of the source code and add-ons that will modify the base functionality of
the generic software. The steps between customization-configuration-customization can be a
continuous process throughout the life cycle of the software. Hence, design-in-use practices in
the context of standardized software are common activities. Thus, we need more attention,
empirical research and understanding of design facilitators in order to gain insights of how the
systems can be designed to improve healthcare activities.
The main challenges to the use of PD methods in distributed OSS environment stem from the
nature of PD methods that largely depend on the collaborative techniques based on proximity
(Bratteteig et al., 2012). Research shows various techniques and tools which can be used to
improve participation of non-developers in the design activities. For example, websites,
workshops, training, mailing lists and academies provides mechanisms of acquiring feedback
from non-developers (Iivari, 2011; Saugene, 2013). Moreover, the role of intermediaries or
48
implementation mediators has been recognized as one way to facilitate participation of nondevelopers within distributed design settings (Iivari, 2011; Titlestad et al., 2009; Braa and
Sahay, 2012). For example, implementation mediators can provide design, implementation or
customization suggestions and training support for the other actors (Iivari, 2011; Saugene,
2013). There are many actors who could plays the role of intermediary, for example, donors,
software customizers, university academies institutions, implementers or NGO (Puri et al.,
2004, 2009; Pries-Heje and Dittrich, 2009; Saugene, 2013). While previous research
acknowledges the importance of mediators in facilitating participation of practitioners within
distributed design settings (Titlestad et al., 2009; Iivari, 2011; Braa and Sahay, 2012), most of
these studies rarely discuss the design-in-use practices of mediators. Implementation
mediators are also design facilitators (Light and Akama, 2012).
Therefore, as I have demonstrated from the literature review, we lack knowledge about
design-in-use within the PD community. We also lack knowledge in the distributed design
environment about mediators’ practices in the local settings when they are interacting with
practitioners. I argue that in order to understand the introduction of standardized information
tools (paper- or computer-based) in primary healthcare settings we require insights from PD
about the design-in-use practices of design facilitators. In the next chapter I will describe the
theoretical framework which has been used to guide the study.
49
CHAPTER 4: THEORETICAL FRAMEWORK
This chapter presents the theoretical framework which influenced the understanding of my
empirical research. My thesis is informed by practice theories. All practice theories share the
assumption that working practices are the unit of analysis, and objects play a crucial role in
organizing practices (Gherardi, 2012; Nicolini, 2013). Furthermore, practice theories share the
assumption that working practices are situated in time and place and thus offer a way of
understanding the social aspects of their accomplishment. Finally, practice theories recognize
that change, organizing, meaning, coordination and learning happen in work practices
(Huizing and Canavagh, 2011). Objects can include information artifacts, meaning various
tools, systems, interfaces, and devices which are used for storing, tracking, displaying, and
retrieving information, whether paper, electronic, or other material (Star et al., 2003).
Practice theories include such theories as actor-network theory, structuration theory,
community of practice, and activity theory (Nicolini et al., 2003; Nicolini, 2013; Kuutti,
2013). My thesis is guided by activity theory, which is a theoretical framework used to
understand work practice and also for informing the design process of information systems
(Kaptelinin et al., 1995, Korpela, et al., 2001; Igira, 2008). The use of activity theory in this
study was influenced by the insights I gained from the literature review (Gasser, 1986; Gerson
and Star, 1986; Engestrom, 1987, 2001; Kuutti, 1996; Law, 1992; Walsham, 1997; Berg,
1997; Aanestad, 2003) and the practical motivation of the study. I elaborate these insights in
the following. First, there are claims that activity theory is more open than other practice
theories on the role of artifacts in work practices (Nicolini, 2013). Second, scholars
acknowledge that the motives of individual do not disappear in the analysis, which is the case
in the analysis of human practices when using the actor-network theory (Kaptelinin and
Nardi, 2012, p. 74). Third, scholars suggest that activity theory is useful for analyzing
information and communication technologies for development practices (Kaptelinin and
Nardi, 2012; Karanasios, 2014) where the major theme is avoiding the failure of technologies,
is also a practical motivation for this study. Finally, it is suggested that activity theory is
fundamentally oriented to contextual considerations (which is also one of the research
questions addressed in this study), because the work practices are seen as culturally-, sociallyand historically situated (Nardi, 1996). Thus, it is suggested that the use of activity theory to
understand information systems can support the development of information systems that
back the situated nature of work rather than resist it (Kaptelinin and Nardi, 2006, p.88).
50
The use of activity theory as an analytical framework for understanding human practices is
common in many research fields such as education, healthcare, and management (e.g.
Engeström, 1987; 2001; Cole, 1995; Jarzabkowski, 2005; Milne et al., 2011). In Information
Systems (IS) research, activity theory has been used not only for understanding human
practices but also for informing the design of information systems (e.g. Kuutti, 1996; Bardram,
1998; Korpella et al., 2000, 2001; Sangiorgi and Clark, 2004; Mursu et al., 2007; Igira, 2008b;
Kaptelinin and Nardi, 2012).
The usefulness of employing activity theory as a theoretical framework to understand human
practices stems from its emphasis that the motives for undertaking activities should be taken
as the point of departure for the analysis (Nardi, 1996). Moreover, activity theory stresses
those motives are not static since they are partly given and partly emergent (Blackler et al.,
2003; Nicolini et al., 2003). For instance, the motives of some practices may hold orientations
or interests that are contradictory for others. By stressing the understanding of human motives,
activity theory provides an analytical framework to understand how to intervene in order to
improve the human practices through the design of information systems (Igira, 2008b).
The remaining part of the chapter is organized into two sections. The first section presents the
history of activity theory while the second section presents specific concepts which I used to
inform the understanding of distributed collaborative practices in my case study.
4.1 Activity Theory
Activity theory has its roots in Soviet psychological studies. The first generation was
established in 1978 by Vygotsky, who based activity theory on the basic idea that human
activities are motivated by some object/purpose. Another important idea was that human
beings do not interact directly with their environments; rather it is through the mediation of
artifacts that the interactions take place. Vygotsky’s ideas were expressed pictorially by
Engestrom as a triad of subject, object and mediating artifact (see figure 6).
Mediating Artefacts
Object
Subject
Figure 6: The first generation of activity theory as presented by Engestrom (1987)
51
However, the limitation of the first generation is that the unit of analysis was individually
centered and lacked regard for the social and communicative facet (Engestrom, 2001). The
limitations were overcome in the second generation developed by Leontev (1978; 1981).
Building on Vygotsky’s theory of human activity, Leontev provided distinctions between
activity, actions and operations (see Figure 7). Activities are achieved through a series of
actions. While activities are oriented towards objects/motives, actions are directed towards
goals. Moreover, actions are implemented through operations, which are oriented towards the
conditions under which the subject is trying to attain a goal.
Activity
Motive
Actions
Goals
Operations
Conditions
Figure 7: The hierarchical structure of the activity system (Kaptelinin and Nardi, 2006)
Moreover, Leontev introduced the term “division of labour” to differentiate between an
individual action and a collective activity. Extending, Leontev’s ideas, Engestrom (1987)
expanded the activity system triangle to include six interrelated elements which have direct
relationship with each other. These elements include an object, a subject, artifacts, division of
labor, rules and community (Engeström 1987), see Figure 8. The subject represents the
involved actor(s) who are motivated by the object to produce the outcome of the activity. The
object can be a material thing (e.g. patient, software) or intangible element (e.g. idea that
clients should be monitored to finish their care trajectories), which the subject aims to
accomplish (Nicolini, 2013). Moreover, the subject uses tools and artifacts to transform the
object into the outcome. The community element represents all people who share in some way
the object of a certain activity system. The division of labor refers to explicit and implicit
allocation of tasks between the community and the subject elements, so that the
transformation process of the object into the outcome can take place. The activity’s rules and
procedures cover explicit and implicit norms, expectations, conventions, and moreover are
used to regulate actions and interactions (Kuutti, 1996; Nicolini, 2013).
52
Artefacts
Subject
Rules and Procedures
Object
Outcome
Division of Labour
Community
Figure 8: The structure of human activity system (Engeström, 1987)
Activity theory acknowledges that an activity system interacts and overlaps with other activity
systems. For example, the activity system may be part of wider network of activity systems,
either because it is the outcome of another activity or it produces resources to be consumed by
another activity or it cooperates with other activity systems to produce the outcome (Allen et
al., 2011). Thus, the third generation of activity theory was developed by Yrjö Engeström in
order to account for wider perspective on human activities and their collective efforts
(Engeström, 2001). The third generation of activity theory proposes that the unit of analysis
should be a minimum of two interacting activity systems (see Figure 9).
Artefacts
Subject
Rules and
Procedures
Artefacts
Object
Community
Division of
Labour
Potentially
Shared Object
Object
Division of
Labour
Subject
Community
Rules and
Procedures
Figure 9: Two interacting activity systems (Engestrom, 2001)
Activity theory is also built upon a dialectical perspective that changes and developments are
triggered by contradictions (Engeström, 1987). Contradiction can be defined as historically
accumulating tensions which emerge in daily work practices causing people to change their
activities (Engeström, 2001). Furthermore, contradictions manifest in work practices as
problems, ruptures, breakdowns, or clashes or deviations from the normal flow of work (Helle,
2000; Blin and Munro 2008; Milne et al., 2011), generally called disturbances. Change of
practices can occur if end-users will not just ignore disturbances and will instead decide to
address them, for example by developing new solutions (Cluts, 2003), which can be a
deviation of formal practices such as adaptations or improvisation (Gasser, 1986; Folcher,
2003; Alter, 2015).
53
Engeström (2001) categorizes contradictions into primary, secondary, tertiary and quaternary
which are elaborated in Figure 10. Primary contradictions (1) occur within elements of a
single activity system; such as within the subject or rules. Secondary contradictions (2) occur
among the elements of the activity system such as between subject and rules. Tertiary
contradictions (3) occur when actors introduce the object of a culturally more advanced form
of a central activity or when they introduce a new element of the activity system. Quaternary
contradictions (4) occur between an activity and nearby interacting activity systems as
depicted. This study has mainly focused on quaternary contradictions within a network of
activity systems.
3
4
1
2
2
4
1
4
1
2
2
4
2
4
2
1
4
Figure 10: Network of interacting activity systems (Engestrom,
4.2 Relevant Activity Theory’s Concepts for Analysis
This study will use insights from the second and third generations of activity theory to
understand work practices in PHC. Leontev’s notion of hierarchy of activity, as activity,
actions and operations (Leontev, 1978) provides a framework to describe and understand
granular level practices with the purpose of understanding the operational PHC information
systems. Activity theory offers a set of concepts for describing human work practices such as
object-orientedness,
multi-voicedness,
activity
system,
mediation,
internalization-
externalization, contradiction and development (Engestrom, 2001). Three of activity theory’s
concepts are especially useful for this research: activity system, mediation and contradiction.
The combination of activity-actions-operations structure from the second generation and the
54
selected concepts from the activity theory provide an analytical framework to understand how
problems or conflicts occur in PHC work practices. Subsequent sections introduce the
concepts of the structure of the activity system, mediation and contradictions and also how
they have been used in the study.
4.2.1 Activity System
The main argument in activity theory is that human activities are directed towards motives
which are constructed through negotiations, alignments or ignoring of different motives,
interests and aspirations represented in the community (Nicolini, 2013). IS researchers have
used the concept of the activity system to identify important actors to be involved in the
design process (Hyysalo and Lehenkari; 2000; Pilemalm et al., 2000).
Moreover, the activity system is seen as culturally-, socially- and historically situated.
Therefore, context is also integral to the activity system. According to Nardi:
Activity theory, then, proposes a very specific notion of context: the activity itself is the context. What
takes place in an activity system composed of object, actions, and operation, is the context. Context is
constituted through the enactment of an activity involving people and artifacts. Context is not an outer
container or shell inside of which people behave in certain ways. People consciously and deliberately
generate contexts (activities) in part through their own objects; hence context is not just “out there.''
Context is both internal to people—involving specific objects and goals—and, at the same time,
external to people, involving artifacts, other people, specific settings. The crucial point is that in activity
theory, external and internal are fused, unified. (Nardi, 1996, p.38)
The theory offers a richer understanding of context as both internal and external to the activity.
The emphasis on “internal context” necessitates a solid analysis of the activity itself. The
recognition of “external context” facilitates the incorporation of context beyond the activity
system itself. For instance, the understanding gained from multiple interacting activity
systems can be the connecting points to the relevance of studying the relationship between
one activity system and other interacting activity systems, such as economic system, industry,
supply chain, and organization (Boers et al., 2002) that encompass the larger landscape
(Korpela et al., 2008).
The hierarchical structure of the activity system (activity-action-operation) can also assist in
the analysis of how context impacts activities. For example, IS researchers have attempted to
study the influence of context at the actions and operations levels and to use the gained
understanding to guide the design of computer-based systems (Kaptelinin et al., 1999;
55
Kaenampornpan and O’Neil, 2004; Cassen and Koffod-Petersen, 2006; Koffod-Petersen and
Cassen, 2006; Huang and Gartiner, 2009; Bardram and Doryab, 2011; Doryab and Bardram,
2011).
Building on the work of these scholars, this study has used the concept of activity system to
identify which activities and actors are involved in healthcare provision, clients’ retention
practices and design of the software. Furthermore, the hierarchical structure of activity system
has been used to identify how external contextual elements (e.g. facility location) impact the
use of the patient registry at the activity-action-operation levels. Also, I have analyzed how
internal contextual conditions (e.g. the object of the specific disease that the health program is
trying to address) impact interacting activity systems.
4.2.2 The Concept of Mediation
Humans do not interact with the world directly; rather, their activities are mediated by
different artifacts developed over time in a specific social-cultural and physical context
(Bardram and Doryab, 2011). The concept of mediation is thus important when we want to
understand the influence of context on work practices since it is through mediation that social
issues such as culture, history and power manifest in action (Kuutti, 1996; Nicolini, 2013). In
particular, activity systems are mediated by artifacts, rules and division of labor (Engeström,
2001; Blin and Munro, 2008; Nicolini, 2013). Artifacts can be categorized into tangible, less
tangible and abstract artifacts (Engeström 1987; Kuutti, 1996; Nicolini, 2013). I present these
categories with the use of PHC workers’ practices in the following sentences. Tangible
artifacts are the ones which health workers consciously utilize during health services
provision, for instance, information artifacts (e.g. patient records in patient registries,
appointment sheets, and mobile phones), and medical supplies (e.g. drugs, stethoscope and
surgical equipment). Less tangible artifacts include healthcare and information systems
procedures and guidelines. Abstract artifacts are the ones which health workers unconsciously
use to organize their work, such as language, human skills/knowledge, means to divide work
or interaction, norms, and social relations within a community (Engestrom, 1987; Bannon and
Bødker, 1989; Kaptelinin and Nardi, 2012; Lee et al., 2013).
This study uses the concept of mediation to understand which information artifacts and other
social infrastructures (such as division of labor and rules) mediate achievement of the patient
trajectory across health programs, facility and community levels. Moreover, the concept has
56
been used to understand how mediation of the activity-action-operation is happening in health
facilities or how actors interact with each other through mediation of artifacts.
Furthermore, the analysis centering on mediation concepts offers a framework to understand
how context impact IS work practices (e.g. Bannon and Bødker, 1989; Nardi, 1996; Huang
and Gartiner, 2009; Bardram and Doryab, 2011). The most significant context for an artifact
is therefore the situation of use, where a subject works to achieve an object. This
conceptualization of the artifacts centers the analysis on their relation to actual use, rather than
as standalone entities. This study has also analyzed how social artifacts are mediating the
activity systems at activity-action-operation levels. For example, defaulter tracing practices
can be supported by rural settings (Papers II, III and V).
4.2.3 The Concept of Contradiction, Disturbance and Mediation Breakdown
The existence of activity networks which have objects/motives with varying orientations or
interests can be a source of conflict and at the same time a source of transformation of the
activity systems (Blacker et al., 2003; Engestrom et al., 2003; Nicolini, 2013). For instance,
the presence of contradictions can result in the expansion of an object (new practice, or spatial
or temporal expansion) which may need new mediating artifacts (Engestrom et al., 2003;
Nicolini, 2013, pp.130–131). Accordingly, studying contradictions that exist within or
between activity systems provides a framework for understanding the motives behind changes
or deviation of activity systems.
Researchers have used activity theory to analyze how artifacts get changed during
information-systems development or use situation. For example, Issroff and Scanlon (2002)
used the concept of contradiction to analyze the use of technology in higher education and to
highlight how problems arise in learning and teaching activities after the implementation.
Also, Murphy and Rodriguez (2008) have used the concept of contradiction to understand
how introducing technology-enhanced learning had an influence on teachers’ practices and
the need for restructuring. Igira and Aanestad (2009) have used the concept of contradiction to
understand challenges facing the design and implementation process of computerized health
management information systems. Ngoma and Igira (2012) used the concept to identify
problems that exist in operational information systems in order to highlight how to improve
quality of information in maternal and child healthcare. In the same way, Igira (2012) argues
that identification of contradiction provides an analytical toolkit for identifying gaps in the
design or implementation of information systems.
57
In this study the concept of contradiction is central to the analysis. I have identified
contradictions which exist in healthcare, defaulter tracing practices and the design of software.
Based on Engestrom (2001), the study has focused on contradictions within activity systems
and quaternary contradictions between the interacting activity systems in all dimensions of
collaborative practices (Papers II and V). These contradictions manifest themselves as
problems, disturbances or inconsistencies between the components of an activity system
(Engeström and Escalante, 1996; Issroff and Scanlon, 2002; Blackler et al., 2003; Lehenkari,
2000). Practitioners see disturbances as interruptions in their work practices that cause
deviations from the normal flow of work (Engeström, 2000; Helle, 2000; Barab et al., 2002;
Issroff and Scanlon, 2002; Blacker et al., 2003; Allen et al., 2011; Milne et al., 2011;
Karanasios, 2014).
Bannon and Bødker (1989) have argued about the importance of looking at conditions which
result in mediation breakdown in the use situation. It is recognized that changes of practice
will occur if users choose not to ignore disturbances. For example, users can respond to the
occurrence of disturbances by changing the object of the activity system or by creating new
artifacts (Cluts, 2003; Murphy and Rodriguez-Manzanares, 2008). This study has specifically
focused on disturbances with the purpose of gaining an understanding of the operational HIS
operating in PHC. Using the insights provided by Helle (2000, p.8) and Engestrom and
Mazzocco (1995), disturbances are defined as “different kinds of disruptions in work and
communication which can exist in action, work task, artifacts or interactions among actors.”
In addition, the term “mediation breakdown” is used in the study to refer to the kind of
disruption in work or communication which happens solely because of information artifacts.
The concept of disruptions has helped to identify mediation problem from information
artifacts and existing changes (Miettinen and Hasu, 2002). For example, the concept of
disturbance was used in combination with the hierarchical structure of an activity at the
action-operation levels in order to identify interruptions occurring during healthcare service
provision and also during piloting of the software.
Much of activity theory’s discussion on the use of activity theory to understand the impact of
context is limited to the understanding of how context can be modeled into artifacts (Chen et
al., 2013). This is not a focus of my research since the approach I take here aims to understand
work practices in their actual work situation. Thus, a combination of the concept of activity
system, contradictions and disturbances from activity theory by Engestrom (1987; 2001), and
58
Leontev’s notion of hierarchy of activity (activity, actions and operations) in the analysis has
provided a holistic view on sources of problems occurring in PHC work practices. For
example, the study has used the concept of contradictions to identify disruptions in healthcare
trajectories which originate from public health concerns (e.g. clients not attending their
appointments), and how its impact is felt on the operational practices around HIS (recording,
IS artifacts, etc). This study has also analyzed how social artifacts are causing mediation
breakdowns at activity-action-operation levels. For example, defaulter tracing practices can be
limited by urban settings or the nature of the disease it aims to support (Papers II, III and V).
Hasu (2001) suggests that improvisation by users in information artifacts provides a means of
identifying breakdowns in work practices. The study has also investigated the practitioners’
improvisations to understand the reason for them. The focus on practitioners’ improvisation at
the action and operation levels helped the study to understand how context impacts the work
situation and also to provide design implications with a focus on the context. Overall, the
analysis of disturbances at activity-, action- and operation levels provides a framework to
understand how context impacts the work situation.
59
CHAPTER 5: RESEARCH APPROACH
This chapter outlines the research methodology utilized in this thesis. I begin by highlighting
the research strategy, and then describe the research design and fieldwork as well as
techniques used for data collection and analysis. The chapter concludes by reflecting on the
research methodology.
5.1 Research Strategy
All research is based on philosophical assumptions, which could be positivism, interpretive,
critical or pragmatist (Orlikowski and Baroudi, 1991; Myers, 1997; Walsham, 2006).
Positivist researchers assume that the world can be understood objectively by researchers
defining certain variables to be focused on during the research periods (Silverman, 2005). On
the other hand, interpretive researchers assume that understanding the world can be gained by
focusing on a social-cultural context according to the meanings assigned by participants and
the researcher’s orientation (Orlikowski and Baroudi, 1991; Myers, 1997). Conversely,
critical researchers assume that social reality is produced and reproduced by people and, thus,
focus on oppositions or conflicts facing the society, and seek to be emancipatory (Myers and
Avison, 2002). In my study, I have adopted an interpretive approach as I considered it suitable
for gaining an understanding of primary healthcare information systems offered in different
social-cultural and geographical settings.
Interpretive researchers in information systems studies can use qualitative, quantitative or
mixed research methods (Myers, 1997). Qualitative research methods help researchers to
understand people and their socio-cultural phenomena in their settings in a detailed form
(Myers, 2002; Silverman, 2005). Quantitative research methods are useful for predictive
studies and involve researchers committing themselves to a prior hypothesis to be tested
through controlled studies or statistical analysis (Kaplan and Duchon, 1988). Since the
practical motivation for the study was to gain an understanding of health providers’ work
practices in primary healthcare information system, the qualitative research strategy provided
me with a methodological approach for achieving the objective of the study without
predefining what aspects of the practice to focus on.
Qualitative research approaches can involve ethnography, action research, grounded theory or
case study (Myers and Avison, 2002). Ethnography is the study of people in naturally
occurring settings which capture their social meaning and ordinary activities (Macaulay et al.,
60
2000; Silverman, 2005; Walsham, 2006; Blomberg and Karasti, 2013). Consequently, I
adopted an ethnographic approach as a strategy to understand primary healthcare because of
these methodological characteristics. In ethnographic studies, researchers participate directly
in the settings of the lives of people who are to be studied, and fieldwork is conducted over a
prolonged time in order to gain deeper understanding of the settings (Markus, 1995; Aanestad
and Olaussen, 2010). The second characteristic of ethnographic studies is that they produce a
thick description of the ongoing interaction between people and technologies they use in their
everyday work. Walsham (2006) argues that the thick description is important for creating an
understanding of what is happening in connection with information systems and its users.
Another reason for employing an ethnographic approach in my study is because the approach
has been deemed useful for both understanding the work practices and for informing the
design and development of information systems (Dourish and Button, 1998; Mortberg et al.,
2010). I have participated in the implementation of the DHIS Tracker system in Tanzania.
Further details about the DHIS Tracker are presented in Section 5.2. Accordingly, I have used
the ethnographic study to gain detailed insights on primary care health workers practices and
use this knowledge to inform the development of the DHIS Tracker.
There are three ways of using ethnography in information system development: ethnography
for IS development, ethnography of IS development, and ethnography within IS development
(Beynon-Daviess, 1997). Ethnography for IS development concerns the use of the
ethnographic approach to highlight the insights on work accomplishment and the tacit
knowledge involved in it. It can also involve studying the use of information systems in an
organization. Ethnography for IS development is a situation where the ethnographic approach
is used for studying IS developers and IS development workplaces. Ethnography within IS
development represents the use of ethnographic approaches as information system
development techniques, such as for requirement elicitation, evaluation and training. I have
used the ethnography approach for IS development and within IS development
In the conduct of the fieldwork in the primary healthcare settings, I have used an multi-site
ethnographic technique to collect empirical data (see for example Marcus, 1995; Hine, 2007;
Nicolini, 2009; Madden, 2010). A multi-sited ethnography technique involves a spatiallydispersed field through which the researcher moves physically to two or more places, or
conceptually, by means of techniques of juxtaposition of data (Falzon, 2012). The benefit of a
multi-sited ethnography technique is that it can provide insights into both local contexts and
61
the wider social systems in which the two are situated (Marcus, 1995). As such, multi-sited
ethnography can be used for comparison and contrastive purposes (Madden, 2010; Blomberg
and Karasti, 2013). This study, therefore, selected multiple-sited ethnographic techniques in
order to get a broader understanding how information systems work in different places. This
allowed me to understand comparisons and similarities between the different settings.
Ethnographic fields in the multi-site studies can be related to geographical, social or mental
constructs (Madden, 2010), such as understanding different sub-cultural groups in different
geographical locations (Hage, 2005). My fieldwork involved different geographical settings
(e.g. urban and rural), different health programs (e.g. vaccination, PMTCT, HIV, tuberculosis),
different facility types (health centers and dispensaries) and health levels (facility and
community).
5.2 Research Design
The research reported and analyzed in this thesis was conducted in Dodoma region, which is
centrally positioned in Tanzania. Dodoma region was selected to be my research site since it
is one of poorest regions in Tanzania. Being a poor region, it provides better settings for
studying and understanding how social-economic situations impact the primary healthcare
information systems, which could be of value to other similar settings.
My research is situated within the HISP project, which is an action research framework aimed
at strengthening health information systems in the southern hemisphere. One way of
strengthening information systems by the HISP network is through development and
implementation of DHIS2 (District Health Information Software version 2) as well as
capacity development. The design of DHIS2 is aimed at enhancing information-use behavior
among health managers, planners and workers in the health sector of developing countries by
strengthening the local professional capacity and enhancing the sustainability of the system
(Braa et al., 2004). As such, since the beginning of HISP activities in the year 1994,
participatory design with its roots in the Scandinavian PD tradition and action research
approach has been core in the development of the DHIS2.
Currently, the DHIS2 platform is the health management information system in use in more
than 47 countries across four continents (www.dhis2.org). In the DHIS2 platform, there are a
number of modules, such as the DHIS aggregate, the DHIS mobile and the DHIS Tracker
module. The DHIS aggregate is the main module within the DHIS platform and it allows the
62
management of aggregate data. It also includes advanced features for data analysis, such as
data GIS, charts, pivot tables, and dashboards. The DHIS mobile is used for data reporting
from a facility to the district level, while the DHIS Tracker is designed to follow up and report
on individual program service deliveries to beneficiaries, as part of the name-based
component of DHIS2. The DHIS system is a national standard in Tanzania for processing
aggregate data at district, regional, and national levels. In Tanzania, HISP has been located
and coordinated within the Department of Computer Science and Engineering at the
University of Dar-Es-Salaam since 2002. The Tanzania HISP team works closely in
collaboration with the Tanzania Ministry of Health and Social Welfare and other various local
and international partners (Nguyen and Nyella, 2010).
The DHIS Tracker can be used for monitoring clients in health programs, such as for patient
with tuberculosis, expectant mothers in antenatal care, children who attend the children’s
clinic within the maternal and child healthcare program, or patients with HIV/AIDS. The
initial version of the DHIS Tracker module was designed by the Oslo team, using the
experience of an Indian team. The purpose of building the system was in strengthening
institutional measures in order to increase immunization coverage, pregnancy tracking, and
home-based care (Gizaw et al., 2012). The DHIS Tracker system has been deployed to
countries such as Guinea-Bissau, Malawi, Tanzania, and Uganda.
As described earlier, this study involved different geographical settings (e.g. urban and rural),
different health programs (e.g. vaccination, PMTCT, HIV, tuberculosis), different facility
types (health centers and dispensaries) and health levels (facility and community). In the next
paragraphs I elaborate on the multi-sitedness of my research.
The study was conducted in two districts (which will be referred to here as Bwawa and Bahari
districts). The districts were selected based on the fact that they had different levels of socialeconomic development (rural and urban contexts). The fieldwork was divided into three
phases: situation analysis, defaulter tracing practices, and implementation of the DHIS
Tracker. The division of the timeline had nothing to do with the objective of study, rather it
had to do with other commitments I had as PhD student at the University of Oslo and, for
example, had to detach myself from fieldwork to accomplish course work and evaluation
requirements of the faculty in which I was enrolled between the first and second phase of my
study. In the following, I present each phase by elaborating on where the study was conducted
and the reasons behind its selection.
63
The first phase of my ethnographic study – situation analysis – was conducted over 6 months
(August 2011–January 2012). My initial focus in the fieldwork was on understanding how
information technologies could be used to improve access to and quality of maternal and child
healthcare in sub-Saharan Africa. In particular, I aimed at gaining an understanding of how
information artifacts were used to help pregnant women (under the PMTCT service) who
moved in the multiple settings where care is provided. During this period, I selected health
facilities from the two districts used as my empirical research sites. Table 3 presents the
overview of study facilities showing the ownership and type features. The names of the
facilities are represented alphabetically (A–K). The selection of the facilities in both districts
was based on the fact that they provided RCH services, including PMTCT services. For the
Bahari district, I had an additional criterion: that the facilities should be connected to the
national electricity grid, so that I could pilot a computer-based system. I selected eight
facilities in Bwawa and three facilities in Bahari. The selected facilities were of different
types: health centers, dispensaries, and hospitals.
Table 3: Overview of study facilities showing ownership and type features
Public
Private
Total
Bwawa
Health Center Dispensary
2 (C, D)
2 (A, B)
0
4 (G, H, I, J)
8
Health Center
1(E)
0
Bahari
Dispensary
1(F)
0
3
Hospital
0
1 (K)
Table 3 shows also that private facilities did not include the dispensary category. This is
because at the time of the study there was no private facility at the dispensary level which
offered PMTCT services. The actual data collection was performed in the maternal and child
health services, HIV and Tuberculosis health programs. As indicated in Table 4, my initial
plan was to have 11 study sites. However, during the situation analysis period, I only
managed to visit nine facilities (A, B, C, D, G, H, E, F, and K) due to challenges that arose
from the use of the ethnographic approach (I provide the reasons in section 5.6). During this
phase I became aware that among the challenges which facilities faced was the drop-out rate
of clients in maternal and child health services. I then came to learn about defaulter tracing
practices which were conducted within health facilities premises to counter the problem.
Some data collection was performed at community level and involved community health
workers and leaders, since I had learned that they were involved in client retention practices.
The second phase of my study was from July to August 2012, and November 2012 to April
2013. In this phase the emphasis was on gaining a deeper understanding of defaulter tracing
64
practices. The fieldwork was performed at the health facility and also at the community level,
and was conducted in four facilities. Van Maanen (1988) argues that ethnographies are
experientially driven since the writers draw experience from their fieldwork (p.4). Similarly, I
used the acquired fieldwork experience (from the first phase) to select which facilities to
proceed with in the second phase of my research. In particular, I ended up with four facilities
in the two districts (A, C, E, and F), all being in public ownership and where two are health
centers and two are dispensaries. I now demonstrate the reasons for selection of the four
facilities in the next paragraphs.
The experiences I had acquired showed that client referral from dispensaries to health centers
created a challenge in health service provision and subsequently led to defaulter tracing
practices. The experience was obtained through the accounts of the informants in the facilities.
For example, the following comment was made by one informant in a peripheral facility: “The
problem which we have in our facility is that we do not know where our referred PMTCT
clients (to other facilities) are.” From this and similar comments, I then decided that one of
the criteria for selecting health facilities in both districts would be that the two selected
facilities should be of referral- and peripheral facility types in the formal system of PHC
services, according to the national guidelines. This decision led to the exclusion of private
facilities in Bwawa district, since I found all private facilities (G, H, I, J) referred clients to
public health centers (A) in the case where clients could not afford some costs, such as for
laboratory use, malaria drugs or outpatient services (in case any pregnant woman or child was
sick). In Bahari district, I had access to three facilities (E, F, K), which made me drop the
hospital facility (K) because it did not fit within the referral and peripheral criteria.
Another criterion was to exclude facilities which had limited or non-existent information
practices on PMTCT services. For instance, one private facility’s staff (H) said that no client
had preferred to deliver a baby at their facility since the service was not free, as it was at the
publicly-owned facility. Likewise, staff in another privately-owned facility (G) never filledout data in the PMTCT registers because the HIV clinic was just adjacent to the antenatal
room. Health workers explained that they did not fill the PMTCT register because there was a
shortage of staff. They also informed me that their task was to make sure that the client went
to the HIV/AIDS clinic room and, then, all HIV related information would be taken care of by
the HIV clinic, and the antenatal clinic would concentrate on specific maternal and child
healthcare services. Similarly, the other public facilities (B and D) had limited information
65
practices since their data documentation practices in the PMTCT registries were minimal.
Thus, in the first district, I selected two facilities (A, C) by eliminating the other facilities
described above, since they could not offer me an opportunity to observe information
practices in totality. The decision is in line with the argument made by Ramduny-Ellis et al.
(2005), who state that when activities happen infrequently, observation will not capture some
instances of work practices. In this phase, I came to learn how defaulter tracing practices were
affected by contextual issues. Consequently, in this phase, I also placed strong emphasis on
understanding defaulter tracing practices and how they were influenced by the working
situations of the four health facilities. Child health services for under-fives are extensive and
include immunization, weight growth monitoring, and PMTCT services. However, when I
wanted to compare PMTCT and another health program, I decided to focus on defaulter
tracing practices related to immunization since the practices were performed more frequently
compared to growth monitoring defaulter tracing practices. Also changes of information
practices were evident (refer to the finding section of Paper III).
The third phase of my study was from January 2011 to August 2012, and November 2012 to
December 2013. Being involved in the HISP project, I had an opportunity to participate in the
implementation of the DHIS system globally in general and specifically for HISP Tanzania.
Consequently, some research activities had been conducted at the Department of Informatics,
University of Oslo. Because my research orientation was on patient-care information practices,
the DHIS Tracker Module gave me the opportunity to study and understand how it can be
used to support health service provision. The Tanzanian HISP team has plans to extend the
development of the DHIS system to the facility level. In the year 2009, the DHIS Tracker
system was piloted in two public health facilities in Pwani region (Ngoma et al., 2012) and in
this study it was implemented in Dodoma region (see paper IV). The DHIS Tracker system
was implemented as a pilot in two facilities in Bahari using the DHIS Tracker version’s data
model which was previously piloted in Kibaha district. The selection of two facilities (E and F)
was based on the criteria that they were less hectic compared to their urban counterparts (A
and C). The implementation period ranged from November 2012 to April 2013, when I was
part of the implementation team together with two MSc students from the University of Dar
es Salaam, who joined the implementation in November 2012. The aim was to gain more
understanding of how the DHIS Tracker could be designed to improve information
management in maternal and children healthcare.
66
5.3 Data Collection
The empirical work was carried out in health facilities, community premises and at the
Department of Informatics, University of Oslo. Data was collected using ethnographic and
participatory design-inspired approaches (Hughes et al., 1994; Blomberg and Karasti, 2012,
2013). The empirical findings were obtained through the following ethnographic techniques:
observation, document review and artifact examinations, and interviews. The study also
obtained data by piloting DHIS Tracker software in two health facilities, using participatory
design techniques of focus group discussion, training and system experimenting. The
ethnographic study, training and focus group discussions were conducted in Kiswahili, since
it is the most-used language in Tanzania. However, the ethnographic study related to my
interactions with the global team and other non-Tanzanian members were conducted in
English.
Ethnography usually involves the participation of a researcher, either overtly or covertly, in
people’s daily lives through watching and listening, informal and formal interviews and by
collecting document and artifacts (Hammersley and Atkinson, 2007). As such, observation
was the main technique used to collect empirical material in order to gain insights of PHC
work practices around the use of information systems and use the gained insights to inform
the development of the DHIS Tracker. Moreover, the data collection had a focus on
breakdowns (e.g. non-use of official information artifacts or non-complying clients), as they
provided understanding of the instances where work is interrupted during the use situation
(Bødker, 1996).
To present the data collection methods used in this study, I will first demonstrate data
collection in primary healthcare-related fieldwork. For the second part, I will present the data
collection in the implementation of the DHIS Tracker.
5.3.1 Primary HealthCare Related Fieldwork
In my fieldwork, I collected data in a total of 11 Primary healthcare facilities, described in
Table 3. The main data collection techniques used were observation, document review and
artifact examination, and interviews. These techniques for data collection were used in all
phases.
67
5.3.1.1 Observation
Observations were conducted at the health facility and community levels. Bryman (2008)
categorizes a participant observer’s roles into complete participant, participant-as-observer,
observer-as-participant and complete observer. A complete participant is a fully-functioning
member of a social setting and other members are not aware of the researcher’s identity. A
participant-as-observer is similar to complete participant but other members are aware of the
researcher’s status. An observer-as-participant researcher performs some observation with
very little participation while in complete observation they only observe and do not interact
with people. At both facility and community levels, my roles altered between observer-asparticipant, and participant-as-observer depending on whether informants were aware of my
status and why I participated.
The observations were performed in the 11 antenatal care clinics, 11 children’s clinics, 4
delivery units, 4 HIV clinics, and a TB clinic. In most facilities, maternal and child health
services are offered within one building, albeit in separate rooms. PMTCT services, antenatal
and for children, could be in a single room, though health providers would attend one client at
a time. Accordingly, it was easier for me to observe multiple services. The aim for conducting
observation was to understand how information artifacts were actually used to facilitate
healthcare services. The following questions were used as guiding principles during my
observations:
I.
II.
III.
Which information artifacts are used to support health service provision?
What breakdowns exist in the use of information artifacts?
How do health workers fit things together in the case of breakdowns?
I used different methods to conduct observation in health service provision. I observed
activities in different clinics while taking field notes on how health providers were
documenting or accessing information in health programs. I also followed health providers 16
and clients in different places where care trajectories were taking places in order to follow up
on which information artifacts were used in each place. For example, on several occasions in
facilities with an HIV/AIDS clinic, I followed nurses when they were escorting PMTCT
clients between antenatal and HIV clinics, and vice versa. I also observed meetings between
nurses and clients and I listened to their conversation and took field notes. Based on that, I
was able to understand how healthcare services were mediated by both formal and informal
16
I use the terms health worker, health provider, practitioners interchangeably
68
artifacts. Figure 11 provides insights of observed practices in children services in the outreach
and institutional premises of health facilities.
A
B
Figure 11: Observation of child health services provision
A: In outreach premises
A: In health facility premises
In some cases, I took the role of a participant-as-observer. For example, when I helped nurses
in registering data for pregnant women/children or helping nurses to fetch drugs/files in
PMTCT or HIV/AIDS clinics, I took the role of an observer-as-participant. Initially I was
asked but later I volunteered to help. The role of observer-as-participant had both advantages
and disadvantages. By helping to register data on antenatal cards or registers, I quickly
learned how working situations affect information system artifacts and practices. For instance,
by interrogating clients on their residential information, I noted that the problem of the
clients’ movements was huge, and also how nurses distributed documenting practices in
formal and informal information artifacts. However, observer-as-participant had a negative
impact on the documentation of research in terms of field notes since I had to wait for when
there were no clients about or do it after I had left the fieldwork settings.
I also made observations concerning defaulter tracing practices where I took a role of an
observer-as-participant. For example, I could watch how health workers searched paper
registries to identify defaulters, and the challenges they encountered in the use of paper
registries. Similarly, I observed and listened when nurses were using mobile phones to contact
clients or health workers in other health facilities about missed appointments. Furthermore, I
listened and took notes when I observed conversations between nurses and community health
workers (CHW) about informing certain clients who had defaulted. For phone conversations,
it was possible for me to hear their conversation since in most cases they worked in pairs and
69
used the loudspeaker mode of the phone so that the co-worker could record the outcome of
client tracing. The observation of and listening to the conversations between clients and health
providers or among health providers also made me understand the reasons why clients did not
show up in care trajectories. On several occasions, I followed nurses and CHW when they
were visiting clients at home. I paid attention to the conversation they had with clients or
other community inhabitants they asked about the clients. Figure 12 illustrates my observation
concerning defaulter tracing practices during the fieldwork period.
In such instances, I observed their information practices for how the outcome of clients’
information was recorded or retrieved from different community-based information artifacts
such as CHW’s notebooks and village registration systems. This aspect deepened my
understanding of how the community level participates in health service provision, and how
their information artifacts interact with facility-based information artifacts. Also, the
observation helped me to understand other challenges facing health workers in provision of
primary healthcare services. For instance, I noticed how challenges arise in defaulter tracing
practices due to CHW’s lack of notebooks.
A
B
Figure 12: Some illustration from the observation sessions concerning defaulter tracing practices
during the fieldwork period
A: Mobile phone calling by nurses to the PMTCT clients in health facility premises
B: CHW visiting a mother of a child on her farm
While observing health services in health facilities, I could also see and hear the interaction
between facility and other community actors. For example, I observed how health providers
informed community leaders about mobilizing clients to bring their children when they did
not attend services or during immunization campaigns. Likewise, I could see and hear how
nurses asked the audience of clients to convey the messages to their neighbors who had not
showed up for vaccination services. I also attended several meetings conducted at health
70
facility level about facilitating defaulter tracing practices. My main focus was to gain more
understanding of how different artifacts (e.g. the invented health programs reports) were used
in the meetings or other places to facilitate defaulter tracing practices.
At the community level, I made observations by attending several events regarding healthcare
services where I took a role of observer-as-participant. These meetings were conducted at the
community level with the aim of facilitating client-retention practices. For example, in
February 2013 in Bwawa district, I observed a campaign coordinated by government and
various donors calling all people who had dropped out of HIV care to return. I also attended
community gatherings, such as the inauguration of a new vaccine, at regional, district and
ward levels in order to understand how community leaders participate in health service
provision. These observations provided me with further understanding of how defaulter
tracing systems were recognized and realized by various actors from the regional, district,
facility and community levels.
I observed work practices in the 11 health facilities, usually from 8.00 am to 3.00 pm.
However, Marcus (1995) argues that multi-site ethnography is mobile and emergent. I made a
plan that each facility had to be visited at least once a week. Sometimes, I would observe for
four hours in a peripheral facility and then move to a referral facility on the same day. In total,
101 sessions of observation in facilities, each lasting between 3 and 7 hours, were included in
the data collection. In communities, the duration of observation in the meeting ranged from 3
to 5 hours. I had encounters with 87 health providers; 42 nurses, 13 doctors, 5 data clerks, 2
patient-tracing coordinators (community health workers who coordinate defaulter tracing
activities in certain facilities), 13 community leaders, 6 village health workers and 6
community health workers (CHW) who were involved in PMTCT services.
5.3.1.2 Document Review and Artifact Examination
Several policy documents were reviewed, such as Tanzanian health policy, Reproductive and
Child Health (RCH) strategic plans, (regional & districts) health annual plans and reports,
PMTCT guidelines, home based care service guidelines, national guidelines for the
management of HIV/AIDS, and HIV act and policy. These documents provided background
information about PMTCT and other collaborating actors and healthcare provision protocols.
Some guidelines like PMTCT also provided background information on the information
documentation and data flow process in client trajectories. Moreover, I reviewed letters
71
written from district level to facilities about defaulter tracing practices and letters written to
facilities by CHW/community leaders regarding maternal and child healthcare. The review
helped me to understand the role of community leaders or district administrators in the
defaulter tracing practices.
Further, I conducted artifact examination – an approach of describing the artifact by focusing
on its users and uses in supporting actual work practices (Carstensen and Sørensen, 1996). For
example, I examined the patient-oriented artifacts to understand how they facilitated
collaborative work by making visible the work of others and coordinating one’s own work.
Artifacts reviewed included clients’ cards, clients’ folders, referral forms, laboratory forms,
village registration system, community health workers’ notebooks and facilities’ registers. I
took photos of patient registries in PMTCT and vaccination programs in my observations
from all facilities. The comparison of photos of the registries in the different phases of my
research provided me with insights on what has changed and made me understand the reason
for change.
5.3.1.3 Interviews
Madden (2010) asserts that interviewing is one of the potential ethnographic methods that can
be used for uncovering informants’ perceptions of the world under study. I conducted some
interviews to get an overview of the observed health workers’ practices, at both the facility
and community levels. I focused on issues like overall work practices, such as work flow and
institutionalization of informal practices/artifacts due to breakdowns. As such, I did not have
interview guidelines; rather, I had some question that emerged from the observation sessions I
had conducted. The style of interviewing I used is according to Madden (2010), who argues
that an ethnographer can use a range of interviewing techniques, beginning with an informal
approach and then moving on to formal interviews. Interviews can be categorized as highly
structured, semi-structured or unstructured (Madden, 2010). I chose unstructured and semistructured interviews because I did not want to limit the responses from the informants. I also
judged that it was appropriate in the context of the healthcare service because of the higher
work load of health providers. The informants in the study comprised nurses, doctors, data
clerks, patient tracker coordinators, community health workers, health officers and
community leaders who were all participating in immunization and PMTCT services.
72
I cannot state the exact duration of informal interviews but the duration of semi-structured
interviews ranged from 30 to 90 minutes. Semi-structured interviews were conducted in
offices before and after healthcare sessions, while the informal interviews were conducted at
the place of service provision. The semi-structured interviews were planned ahead, for
example by appointment, while the informal interviews were conducted on ad-hoc basis.
Bryman (2008) categorizes interviews into individual and group interviews. Using group
interviews, people who were known to have had a certain experience could be interviewed in
a relatively unstructured way about that experience. Accordingly, I have conducted both
individual and group interviews.
I performed informal interviews with health providers based on observed work practices for
the purpose of getting clarification of such issues as under-use of formal artifacts, and
documentation in informal artifacts. The interviews took place during healthcare sessions, but
I did not interrupt the process of care; rather, I waited until there were no clients or until the
end of the day’s session. The informal interviews were also conducted during observations I
performed; for example, during or after observation of defaulter tracing activities.
Apart from interviewing health providers, I also conducted four informal interviews with five
clients (four women and one man) who were attending the PMTCT services. One interview
was conducted with a couple in one PMTCT clinic, while the other three interviews were
conducted with individuals in the outpatient and HIV/AIDS clinics. The theme of the
interviews was their experience of going to multiple settings carrying information artifacts,
and their experience following healthcare in different health programs in different facilities. I
waited until they had finished their healthcare consultations and then asked if I could have a
little chat with them.
Phone interviews can be a viable way of communicating with informants in geographically
diverse locations (Berg and Lune, 2012). I conducted four phone interviews with the first
peripheral facility with one health provider, and four phone interviews with two health
providers at the second peripheral facility in order to get clarification on an issue I had
encountered in the examination of the information artifact. The phone interviews were also
used to follow up on some clients who had missed appointments or been referred and had not
come back. I later performed some informal interviews with health workers regarding the
reason for a change of information artifacts after I had first examined them.
73
I also conducted interviews with some community leaders in either their home settings or any
other place suitable for them. The theme of the interviews was related to their participation in
defaulter tracing practices, specifically for both immunization and PMTCT services. Figure
13 provides some illustrations of group interviews conducted during the fieldwork.
A
B
Figure 13: Group interviews illustrations
A: With community leaders
B: With community health workers and a fellow researcher
Table 4 presents the summary of the interviews conducted with health providers working in
primary health facilities during the fieldwork period.
Table 4: Summary of interviews at facility and community levels
Type of informant
Number of
CHW
Community leaders
Nurses
Medical doctors
Patient Tracker coordinator
Total
informants
12
13
14
8
2
49
Total number of interviews (with repetitions)
Individual
Group
7
4
10
2
13
2
9
1
5
44
9
Interviews were also conducted with health administration officers for health services
supporting maternal and child health at district, regional and zonal levels. The theme of
interviews was the elaboration of observed work practices, such as the essence of defaulter
tracing practices and collaborative practices and their relation to information work practices at
the PHC levels. Table 5 presents the summary of the interviews done with health
administration officers.
74
Table 5: Summary of interviews at district, regional and zone levels
Location
Health Administration
Number of
Total number of interviews (with
Department
Informants
repetition)
Bwawa district
RCH
HIV/AIDS
HMIS
Health officer
3
2
1
1
9
4
2
2
Bahari district
RCH
HIV/AIDS
HMIS
Health officer
5
3
1
1
11
8
1
2
RCH
HIV/AIDS
Health officer
Total
3
1
1
22
4
1
2
46
Regional and
Zone Levels
5.3.2 DHIS Tracker Implementation
The DHIS Tracker implementation was conducted in the third phase of the research.
Observation, artifact examination, and interviews were methods employed in data collection.
5.3.2.1 Observation
Observation of the DHIS Tracker system occurred in two places: at the department of
informatics and in health facilities. I performed the role of participant-as-observer in the
DHIS Tracker implementation system in the health facilities as I was an implementation
facilitator who participated in preparation, training and implementation of the DHIS system.
While participating in the training for using the DHIS Tracker system, I also listened to the
conversations between health providers regarding the use of the DHIS Tracker or the
comments made by community health workers or leaders. The following questions were used
as guiding principles during my observations of the DHIS Tracker:
IV.
V.
VI.
Which features of the DHIS Tracker can be used to support health service provision?
What breakdowns exist in the use of DHIS Tracker?
What needed to be changed in the DHIS Tracker to fit healthcare practices?
The first part of the observation was done by attending 4 DHIS implementers’ workshops
organized by HISP with the aim of disseminating DHIS skills. The workshops were usually
conducted for a period ranging between one and two weeks. My participation in DHIS
training workshops provided me with opportunities to learn about the DHIS system – and
DHIS Tracker in particular. Further, I also learned about different roles of the stakeholders
75
involved in its design and also how design practices are distributed between the local and
global level. The learning I acquired from the workshop was supplemented with informal
learning through informal interactions and discussions I had with colleagues from other
countries. For example, I learned about the DHIS Tracker module in use in Tanzania through
colleagues from Tanzania. One of them was a PhD student at the University of Oslo and the
other was a programmer who developed the modules. I also had discussions with DHIS
Tracker implementers from Malawi and Mozambique, also PhD students at the University of
Oslo.
The shortcomings in the DHIS system were communicated to the software development team
at the University of Oslo using several approaches, ranging from email conversation, short
meetings and formal powerpoint presentations where I invited some developers and another
coordinator to attend. In one instance, I sat with the coordinator of the DHIS Tracker
Developer who interacted and worked with developers and discussed the data format for
functionality of a defaulter list and health program summary reports. I also prepared a mockup to demonstrate how the system should look in order to facilitate SMS to non-client actors.
During the whole period of the study, I had interaction with 9 DHIS core developers at the
University of Oslo from August 2012 to December 2013. In health facilities, I took notes on
the conversation/discussion I had with fellow implementers; for example, on several
occasions we had conversations on the shortcomings we observed when deciding where to
locate the computers and about the training of health workers.
The time spent in the observation within the facilities related to the DHIS Tracker
implementation ranged from 2 to 6 hours. The implementation involved the task of
installation and training health workers and experiencing the system in practice. Figure 14
provides insights from the fieldwork of the observation conducted. Moreover, this period of
implementation and interaction with the global team was used to understand design issues,
evolution of the DHIS design, and the facilitator’s role. In April, we 17 conducted 10 focus
group discussions with health workers with the purpose of obtaining their views on the digital
information system. I prepared discussion topics based on the comparison of their work
practices around defaulter tracing and DHIS Tracker’s assumption on defaulter work practices.
We also demonstrated the system to some community leaders in order to get their views on
the design of the system. Moreover, on several occasions, I had discussions with fellow
17
The three people involved in the implementation of the DHIS Tracker
76
implementers about the shortcoming of the DHIS Tracker which we had observed, based on
our experiences with health workers. We discussed and agreed which data item or form
should be changed and in what way and, consequently, we re-designed some forms.
A
B
C
D
Figure 14: Illustration of observation session conducted during the DHIS Tracker fieldwork
A, B: DHIS Tracker training sessions in the two facilities
C: Demonstrating the system to a community leader
D: CHW using DHIS Tracker to get defaulters names from downloaded excel format sheet
5.3.2.2 Artifact Examination
The review of artifacts was also performed on the DHIS Tracker software. I examined DHIS
software functionalities in order to find out how they matched with work practices observed
in the facilities. I have examined how DHIS features related to layout, contents, and
functionality have changed over time in the different versions released, and linked this with
suggestions gained from the ethnographic fieldwork. The examination was done by
scrutinizing emails related to DHIS release sent on the DHIS implementers’ mailing lists and
also by experimenting with the demo version of DHIS Tracker from the HISP website.
77
5.3.2.3 Interviews
I further conducted 9 informal interviews with three coordinators of the DHIS Tracker
development at the department of informatics about the change of the DHIS Tracker’s design.
These interviews were conducted before and after the implementation of the DHIS Tracker
with the aim of understanding the history of DHIS Tracker development and gaining an
update about a request I had made for the changes on the DHIS Tracker. These dialogues
ranged from 20 to 40 minutes. The interactions were also supplemented by short informal
talks, whenever the opportunity arose in day-to-day activities related to the implementation
and use of the Tracker.
5.4 Field notes and Data Analysis
I have used ethnographic and practice-based approaches to present and understand primary
healthcare facilities and work practices. Ethnographic practices involve three phases: writing
field notes, organizing data and creating stories (Madden, 2010). Field notes can be
categorized as primary or secondary (Madden, 2010); primary field notes are taken during the
fieldwork while the secondary field notes are expanded primary field notes and have an
analytical and reflective tone (p. 123). Field notes were taken during observations and
interviews, and document reviews. Both English and Kiswahili were used in the field notes.
The ethnographic study, training and focus-group discussions were conducted in Kiswahili.
However, the ethnographic study related to my interactions with the global team and other
non-Tanzanian members which were conducted in English. Field notes can also be
constructed from a diary (Madden, 2010). I did keep a fieldwork diary, which documented my
reflection about primary healthcare work practices and the way they evolved over time. The
diary also helped me to follow up certain issues which were not clear during the next
fieldworks visit. The diary was kept on the implementation of the DHIS Tracker in order to
document it. The primary notes were later expanded into detailed field notes which were
translated into ethnographic descriptions (Emerson et al., 1995) in English.
To construct the ethnographic descriptions, several approaches had been used. First,
ethnographic descriptions were prepared for all 11 facilities, with the main focus on
understanding information needs, tools and the facilities’ practices and their socio-economic
profile (e.g. population statistics, type of health facility and service provided, organization of
their work practices, challenges facing facilities, etc.). The facilities’ ethnographic
78
descriptions were read several times in order to obtain an understanding of the data. The
theme of change of information artifacts emerged from reading and reflecting on descriptions
over a period of time. In particular, the analysis indicated that there was a tension (Engestrom,
2001) between the official information system and the work-related demands that resulted in
the adaptation of information artifacts by health workers. Kaasboll et al. (2010) argue that for
users to develop information systems competence, they need to be fluent in three domains:
information, technology and work tasks. The descriptions from health facilities showed that
practitioners changed their information artifacts’ content, layout or introduced new
information artifacts. Hence, Kaasboll et al.’s (2010) insights helped me to identify existing
changes made by health providers in artifacts (e.g. layout, add/skip or remove information
fields in paper records) or development of informal guidelines (defaulter tracing guidelines).
Researchers have noted that the construction of ethnographic descriptions in multi-site
ethnographic approaches can be done by using such techniques as following the people, the
thing (artifact) or work practices, or the story (Markus, 1995; Blomberg and Karasti, 2013).
These insights also guided me in further organization of the ethnographic description into
narrative descriptions (Emerson et al., 1995). In this study, I used the technique of following
work practices around healthcare provision in health facilities or, defaulter tracing practices.
This was aided by a characterization strategy (Van Maanen, 1988) through a combination of
stories from multiple clients put into a character of a certain client in order to improve
readability of work practices around care trajectory (see Paper I). The description of defaulter
tracing practices was also based on how the work of tracing clients between facilities and
communities was carried out. Further, the concept of the patient trajectory helped me to
identify coordination artifacts and protocols used in each place where a health service is
provided. I also used the technique of following the DHIS artifact (in both local and global
instances). The ethnographic description of the DHIS Tracker was prepared based on how (in
both local and global instances) the DHIS Tracker system changed (functionality/layout)
depending on the ethnographic study I had conducted, or participatory practices carried out
(by engaging health workers in the design practices).
However, in ethnographic studies, the process of data collection, analysis and writing can be
intertwined (Madden, 2010). Accordingly, throughout the study I have been going back and
forth in activities of reading primary field notes or data collection, and expanding
ethnographic descriptions from the facilities. For example, further reflection of ethnographic
79
descriptions of the four facilities in the second phase of the study revealed the differences
related to health programs, specifically vaccination and PMTCT programs. This description
helped me to distinguish the health programs and to understand their differences in the pace of
change of coordination artifacts. Further reading of the ethnographic description showed that
there was a (social-economic) contextual influence that had a role in the change of artifacts/
practices. This made me to go back to primary field notes and investigate the influence of
social-economic conditions in 11 facilities. The extracted information was incorporated into
ethnographic descriptions related to defaulter tracing practices or clients’ stories, etc. Finally,
the ethnographic description was re-arranged for each of the different papers related to
collaboration across health programs and facility (Paper I); defaulter tracing practices (Paper
II); Vaccination (Paper III); DHIS Tracker implementation (Paper IV); and PMTCT (Paper V).
The individual papers were analyzed using different theoretical frameworks under the practice
paradigm. I have been challenged about how to understand different phenomena of health
workers’ practices I encountered throughout the data collection period. It is suggested that a
better understanding of context can be obtained by combining empirically- and theoreticallyinspired analytical approaches (Kaptelinin et al., 1999). Therefore, I have combined bottomup (and empirically driven) and theoretically-inspired approaches to understand the domain,
where I have combined insights from practice-based studies and from information system
design to understand primary healthcare information systems’ practices. From practice-based
studies, I have used activity theory and other concepts (see Paper III) in order to understand
work practices. From the design-oriented discourses within the broad information systems
field (including computer-supported collaborative work and participatory design), I have used
insights about coordination work, and participatory design practices. The plethora of concepts
from the two fields have deepened my understanding of the domain regarding the way health
workers utilize information artifacts in their work practices and the challenges they face.
The macro level of analysis was performed in the dissertation in order to sum up the issues
encountered in primary healthcare information systems. The theme of distributed
collaborative practices around care trajectories across health programs and facilities, health
administration, community and implementers and global software development were evident
from reading and reflecting on the individual papers. The analysis in the dissertation is done
on two levels. First, I have prepared new narrative descriptions (Emerson et al., 1995) of
collaborative work practices (from attached papers) with a focus on aspects of distributed
80
collaborative practices. Second, I have performed the analysis through the use of theoretical
concepts as a base for framing broader contribution to the literature of collaborative work
practices. The concept of coordination mechanisms was used to make sense of the practices
observed when I scrutinized the narrative description for coordination artifacts or protocols
used to mediate both primary and articulation work activities. Also, I decided to use activity
theory to perform a comprehensive, repeated analysis across all papers (Chapter 6) in order to
provide a more extensive analysis of the theme of distributed collaborative practice. The
reason for using activity theory in summarizing key findings is because it has been used
implicitly or explicitly in all papers. In Papers I, II and V, the findings have been analyzed by
explicit use of the concepts of activity system, mediation and contradiction. On the other hand,
Papers III and IV present findings of challenges in work practices by analyzing where
disturbance, conflict or misfit occurs.
Using the concept of activity system, I was able to identify actors who are involved in
collaborative practices. Moreover, the concept of mediation helped me to identify information
artifacts used in each activity system and across collaborating activity systems (e.g. the
referral form). The concept of contradiction was used to identify problems in the mediation,
and changes within collaborative work practices. However, in order to elaborate the problems
occurring in the collaboration within multiple providers, I specifically focus on disturbances
(Engestrom, 2000; Helle, 2000) or mediation breakdowns (Bannon and Bodker, 1989) in
collaborative practices. I identified disturbances or breakdown of the artifacts’ mediating role
across activity, action and operational levels and linked these with contextual conditions.
5.5 Ethical Considerations
The research proposal for conducting this study was cleared by the PhD board of the
Department of Informatics at the University of Oslo. I am also one of the academic staff
employed at the Department of Information Systems at the University of Dodoma in Tanzania.
The empirical study was approved by the University of Dodoma’s Research and Publication
ethical committee. The permit granted by the university allowed me to visit any premises
where maternal and child healthcare was provided unless the health management restricted
me. Consequently, I gained written permission prior to the fieldwork from both district
managers, and obtained oral consent from facility management and health workers prior to
conducting interviews or observations. In addition, clients who were interviewed gave oral
consent first to health providers and then to the researcher.
81
Madden (2010) states that ethnographers have to make ethical decisions in the writing
process, for example, about what material to include or exclude, about evolving issues of
privacy and confidentiality that arise in the writing process (p.33). De-identification strategy
(Rothstein, 2010) during transcription, in paper publication and in the thesis has been used by
omitting the names of clients. For example, though I photographed the patient registry book
pages containing personal information which I used for my research, they were concealed
when presenting or publishing these photos (Papers I, II and V). Moreover, no pictures of
PMTCT clients are included in paper publications and in the thesis.
Also, researchers are supposed to be aware and responsible for their informants, and think
through the possible consequences of their actions (Silverman, 2005; Madden, 2010).
Accordingly, the names of the districts, facilities, and health workers have also been omitted
as I reasoned that some of the issues, like health workers not filling in registers or not
providing PMTCT services, might have consequences.
5.6 Methodological Reflection
In this section, I reflect on the challenges which were encountered in the conduct of this study.
Madden (2010) argues that validity is one of the things which challenges the credibility of
ethnographic studies. He suggests that one way of improving validity is “for ethnographers to
outline clearly why they did what they did when they did it; a case of data transparency”
(p.26). To improve the validity of my study, this chapter has elaborated clearly on the
decisions I made regarding the selection of empirical settings as well as my roles in the
fieldwork. I have also sought to strengthen the reliability of my findings by doing the
fieldwork at different healthcare levels: community, facility, district, regional and zone levels.
By doing this multi-level study, the findings and my interpretation have been discussed with
health workers at different levels of the health sector. I also co-authored four papers with
senior researchers, the process which helped to cross check my interpretations. The findings
have also been analyzed by using studies which have discussed primary healthcare or
contextual influence on information systems, both within the developed and developing
countries environments.
I also understand that my overlapping roles – ethnographer, facilitator, researcher, and trainer
– have validity implications. The multiple and overlapping roles provide an intellectual
resource for producing more realistically attuned data. On the other hand, Madden (2010)
82
claims that if the ethnographer is both a method (tool) and methodological assessor, the
influence of the ethnographer on the research process needs to be assessed. For instance, there
are ethical/validity concerns in my approach as a facilitator where I use myself as a unit of
analysis. I understand that this is not an optimal method in conducting research. I did so in
order to provide insights into the role of mediators in distributed global software development
and the challenges they may face. At the time of writing, there was limited understanding of
what mediators do within the HISP community, and the challenges they face in the settings of
developing countries. However, the study was undertaken with other implementers so that my
findings are based not only on the reflection of my role but also how we as a group
participated as implementation mediators. Also, for ethnographic studies, it is normal for an
ethnographer to relate confessional tales where the researcher gives an account of the
fieldwork (Van Maanen, 1988; Madden, 2010) and a description of hardship endured or
overcome, and accounts of the effects of the fieldwork on the fieldworker (Van Maanen,
1988, p.73).
The multisite-site ethnographic approach was useful for gaining a deeper understanding of the
work practices in different places. However, it was difficult for me to observe large numbers
of facilities dispersed around the two districts. The challenges of conducting multi-site study
have been reported in other studies (see for example, Hage, 2005). As a consequence of
challenges arising from the ethnographic studies, I decided to reduce the number of field sites
which has further implication on the generalization of my findings. Macaulay et al. (2000,
p.40) argue that designers will be better informed by generalization that they can base on their
design decisions. Furthermore, they argue that it can be difficult to attain useful generalizable
conclusions if the fieldwork has been conducted in limited field settings. This argument is
valid for my study, since the findings are drawn from only 11 out of 107 health facilities in
the two districts because of the difficulties in conducting multiple ethnographic studies at the
same time.
Also, Hine (2007) argues that there is a potential for loss of depth in ethnographic studies
when the researcher employs a multi-site approach (p. 665). Thus, my study does not provide
a complete view of client retention practices or implementers’ practices. I view this as a
limitation of my study, as I was not present at all times in the facilities; the findings are based
on the participant observations conducted only during the visits to the facilities. To minimize
this, I used phone interviews (cf section 5.3.3.1) with some facilities to balance the difficulty
83
of being present at many sites. However, the use of phone interviews to follow up on certain
issues was limited as some health workers could not elaborate as effectively on the use or
change of information artifacts by mobile phone as they could in face-to-face, in-situ
interviews. The limitation of the phone interview has been noted by other researchers (Berg
and Lune, 2012). Accordingly, my study does not provide a complete view of implementers’
practices since I have researched only two facilities.
Another limitation of conducting ethnographic studies is that researchers can have difficulties
in accessing some sites (Van Maanen, 1988; Hammersley and Atkinson, 2007). I was denied
access to one religious facility because the administrators, reading my study title, felt that they
had only minimal use of ICT tools and, therefore, concluded that it was not a suitable place
for me. I was unable to visit one private facility due to time limitation and with some other
sites I felt that I was not very welcome; for example, some health workers asked me how long
I would be staying. It is well known that researchers may face difficulties in gaining access
related to interviewing (Hammersley and Atkinson, 2007). In my research, I faced challenges
related to interviewing many clients. Though it is advised that relations between researcher
and interviewee need to be established and identities co-constructed (Hammersley and
Atkinson, 2007), the efforts to interview more clients in the PMTCT services did not bear
fruit as in some cases the nurses disapproved or we 18 sensed that clients were not in good state
or were unwilling to cooperate. Furthermore, we arranged some appointments where clients
who had agreed to be involved did not show up. The reasons I give for this problem of clients
not showing up are based upon my interpretation of conversations I/we heard between
nurses/CHW and clients or what health providers/CHW said during the interviews.
There is also the limitation of time, since this study was undertaken for my PhD, which was
initially scheduled for three years. Thus I was not able to observe changes of the digital
artifact use in the two facilities over a prolonged time. Such extended observation would have
provided more insight into local changes of the DHIS Tracker system arising from the actual
use of the system.
18
People who co-authored Paper I
84
CHAPTER 6: ANALYSIS OF THE KEY FINDINGS
This chapter builds on the research findings presented in the previous chapter as preparation
for the next chapter that elaborates on the contribution to knowledge and practice that have
emerged in the dissertation. As stated in Chapter 1, the main objective of the thesis is to
strengthen the understanding of distributed collaborative healthcare practices in resourcerestricted settings and to use the understanding gained to inform the design of HIS to support
PHC. The research findings presented in the appended papers were analyzed by using
different concepts under practice-oriented research. Nonetheless, in this chapter, I present a
summary of the analysis of the research findings across the papers by using activity theory as
a theoretical framework. The reason for employing activity theory in this analysis is to offer a
coherent and comprehensive analytic lens to the key findings presented in the individual
papers.
In particular, I will analyze the research findings of the papers by identifying the systemic
relations within the element of the activity system and between activity systems, as described
by Engestrom (1987), which are involved in the collaborative practices. The concept of
coordination mechanisms is used to make sense of the collaborative practices observed since I
use the concept of mediation to identify the coordination artifacts or protocols that are used to
mediate both primary and articulation work activities. Moreover, the mediation concept is
used to identify information artifacts used in each activity system and across collaborating
activity systems. The concept of contradiction is used to identify problems in the mediation,
and changes within collaborative work practices. I will also analyze activity by looking at its
actions and operation of corresponding objects as described by Leontev (1978) in order to
identify disruptions (see Engestrom, 2000; Helle, 2000) in collaborative practices by
searching for artifacts mediating breakdowns or interacting with activity systems.
The analysis aims to respond to the two research questions posed in the introductory chapter:
1. What are the characteristics of distributed collaborative work of the PHC settings?
2. How does context impact the distributed collaborative work of the PHC settings?
The analysis is divided into three parts. The first part focuses on collaborative practices
among health providers from different health programs to facilitate care trajectories. The
second part focuses on the collaborative practices between facility and community actors in
facilitating care trajectories both within facility- and community settings. The third part
focuses on the collaborative practices between health providers, implementers and global
85
software developers in the design and implementation of software for monitoring patients in
the health programs. In each of the three parts, I present the activity systems and mediations
involved, and resulting disruptions and their resolutions. At the end of the chapter, I present
the summary of key findings that the summarized analysis has provided.
6.1 Collaboration in Facilitating Care Trajectories across Health
Programs and Facilities
This section presents collaborative practices among health providers across health programs
and facilities in facilitation of patient trajectory.
6.1.1 Activity Systems and Mediation
To demonstrate collaboration in facilitating care trajectories across health programs and
facilities, three main activity systems were identified in this study (see Figure 15): the Health
Care Activity System (HCAS), which is viewed from the perspectives of health providers; the
Health Management Activity System (HMAS), which is viewed from the health
administration’s perspective; and the Clients’ Activity System (CAS), which is viewed from
the clients’ perspective. In each activity system, I identify the object of the activity systems
together with its subject and community elements. Moreover, actions of corresponding objects
as described by Leontev (1978) are identified. Furthermore, the activity systems’ mediations
are identified.
The object of HCAS in relation to activity theory analysis is the clients’ physical wellbeing.
More specifically, HCAS has different actions, as described by Leontev: HIV counseling and
testing for pregnant women and their partners 19, enrolling them into the PMTCT program,
provide further HIV counseling and education for them, and providing medication (ARV
prophylaxis and cotrimoxazole) for pregnant women and their new-born children. However,
PMTCT services are provided in more than one clinic. For example, pregnant women
diagnosed as HIV-positive would receive PMTCT care from the antenatal, delivery, and
postnatal clinics. Moreover, infants born of the HIV-positive women were required to be
enrolled into the PMTCT program as well as being enrolled into the under-fives clinic. The
subjects of HCAS are its health providers who work in PMTCT-related activities. The
community element of HCAS comprises nurses, laboratory technicians, public health officers,
health managers, doctors and the catchment population.
19
Clients are allowed to decline HIV testing after a counseling session (PMTCT guidelines, 2013)
86
HCAS
CAS
HMAS
Wellbeing of
the society
Figure 15: Three interacting activity system in the PMTCT services
There are multiple tools which mediate the collaborative provision of healthcare services. For
instance, patient registries (PMTCT) located within health programs mediated the interaction
between providers and clients during healthcare provision. The collaborative practices among
multiple providers were also mediated by healthcare guidelines and clients’ cards (Paper I).
PMTCT healthcare guidelines prescribed what PMTCT task should be provided in each of the
clinics. Also, clients were given personal cards by health providers from each clinic where
they received services. For example, antenatal care clients were provided with an antenatal
card which contained information about their PMTCT status. The PMTCT information
recorded in the antenatal card conveyed information to other health providers (within health
programs) on what kind of health activities are required in their clinics. For example, a nurse
in delivery care, seeing PMTCT codes on clients’ cards, will be alerted that PMTCT services
need to be offered to the pregnant mother during labor and after the child is born. Similarly,
the PMTCT code on the antenatal card of a woman who has delivered a baby will alert the
care provider in the under-fives clinic to also enroll the child into the PMTCT program and to
have the PMTCT code written on the top of its under-fives clinic card.
Moreover, HCAS is intertwined with HIS-related actions in accessing clients’ data to
determine the care trajectory’s status and also to guide service provision, and documenting
87
data after care is provided (Papers I, III, and V). Thus, the subject of HCAS also coordinates
clients’ trajectories by accessing and documenting information in patient registries and
clients’ cards.
HCAS is also impacted by the health authority of Tanzania (from the national to district level),
which plays a crucial role in management of healthcare service provision in the country
through creating information artifacts and health guidelines, and supplying medical and
human resources. Moreover, the health authority has the responsibility of supervising
healthcare services in order to make sure that the goals and policies set by the country are
achieved. Here, I refer to the national management of healthcare of Tanzania as Health
Management Activity System (HMAS). The subjects are managers located at the district,
regional, zonal and national levels. The community element of HMAS is composed of health
providers and other collaborating partners (e.g. economic donors). The artifacts which
mediate the practices of managers include national policies, strategies and goals, and
information collected via the HMIS reports sent from the HCAS (Papers III and V).
Clients attend health services because they aim to become physically well through prevention,
treatment and care. The Client Activity System (CAS) is mediated by personal cards, which
are given by health providers from each clinic where they received services (e.g. antenatal
card, under-fives card). The CAS is also mediated by other artifacts, finance, and the time to
go to health facility to continue with healthcare. The community element of CAS comprises
family members, relatives, friends and the catchment population where clients are residing.
However, PMTCT care guidelines indicate that clients have to receive health services from
other, multiple clinics as well. For instance, a one-month old child who is diagnosed HIVpositive in the PMTCT clinic will attend three clinics in parallel: under-fives clinic to receive
vaccinations and growth monitoring; PMTCT services to continue with ARV medication,
counseling, and HIV testing; and the HIV clinic to receive palliative care. Equally, pregnant
women who are diagnosed as HIV-positive would receive care from an HIV clinic. Besides
this, in case such clients contracted TB, they were supposed to attend the TB clinic. Moreover,
clients would be required to seek services from other facilities if there was unavailability of
all the required services within a single PHC facility. From an activity-theoretical perspective,
the PMTCT service provision is connected to other multiple individual activity systems of
each health programs and facilities involved. Figure 16 gives an overview of the network of
activity systems for facilitating care trajectories across health programs.
88
Object:
prevention of
MTCT
PMTCT HCAS
Patient Registries
Referral forms
Other HCAS
Object: Well
being of clients
in MCH
Patient Registries
Referral forms
HMAS
Object:
Management
of
Healthcare
National Policies, Strategies,
Goals and HMIS Reports
Client AS
Client’s Cards
Mobile Phones
HIV HCAS
Tuberculosis
HCAS
Patient Registries
Referral forms
Object: Prevention,
Care and Treatment
of TB infected
Figure 16: The network of activity systems involved in the PMTCT Care
Clients carrying Referral
or Laboratory Forms
Object: Physical,
Social-Economic,
and Psychological
Wellbeing
Object: Wellbeing of Clients
through prevention of
infections and treatment of
diseases
Object: Prevention,
Care and Treatment of
HIV-infected People
Patient Registries
Referral forms
89
The diagram shows that collaborative work in PMTCT is achieved through multiple HCAS
and also through HMAS and CAS. The figure also shows that clients provide crucial
coordination work in the healthcare services by mediating the information flow between
health providers located within different facilities and programs.
The objects of HCAS, HMAS and CAS overlap in their services offered to create the physical
wellbeing of clients. The subjects of each HCAS are its health providers while the community
element of each individual HCAS is formed by other health providers in other collaborating
services as well as health managers, economic donors, clients, and other catchment population.
The collaborative practices of multiple HCAS were mediated by multiple artifacts and clients’
practices. For instance, healthcare guidelines (e.g. PMTCT, Vaccination or HIV) prescribe
what care should be provided in each clinic and what care should be provided by
collaborating healthcare services. The collaborative practices among multiple providers were
also mediated by clients’ cards from each clinic where they received services (e.g. antenatal,
under-fives, TB, and CTC) (Paper I). Also, patient registries/folders located within health
programs offered insights about other collaborating programs involved in the care of the client.
Similarly, clients also contributed to coordinating their care trajectories by carrying referral
forms and X-ray and laboratory test results between health facilities/programs which were
used to inform other health programs/facility about the types of services they required. In the
next section, I present how interaction between the three activity systems resulted in
disruptions.
6.1.2 Disruptions within Care Trajectories and their Resolutions
The interaction between the three activity systems resulted into disruption due to mediating
artifacts, conflicting interests in the object of interacting activity systems, and the influence of
contextual conditions. My research indicated that there could be disruptions within healthcare
provision caused by medical resource shortages (e.g. ARV medication, HIV testing kits), and
the staff’s limited PMTCT skills (Paper I and V). Moreover, there could be disruptions within
the facilitation of clients’ care trajectories due to information artifacts; either because of
inadequate design of referral form (without a feedback part) or due to a shortage of forms
(Paper I). The emergence of previous disruptions can be interpreted as a contradiction
between the HMAS and the HCAS because the HMAS has the responsibility of designing
information artifacts and supplying medical resources. Hence, the first source of disruption
90
occurring in achieving the object of the HCAS is caused by the HMAS’ inadequate resource
supply or poor design of the information artifacts.
My research indicates that conflicting interests between the object of HCAS and CAS can be
a second source of disruptions in the care trajectory. The trajectories of individual PMTCT
clients are accomplished when the clients receive services in all required/connected HCAS as
directed by the PMTCT guidelines. On top of that, clients have to attend the health services so
that health providers can provide the services. Disruptions in healthcare-provision trajectories
occur when clients do not show up for their scheduled appointments. The reasons why clients
do not show up for a scheduled appointment can be related to economic issues; for instance
they prioritize work and income generation activities over healthcare appointments. This is
identified in my research when clients temporarily moved to work on distant farms when their
appointments are due (Papers I, II, III and V). Also, issues related to the need for clients to
maintain psychological and social wellbeing can influence a client’s willingness to show up in
the HCAS as, for example, clients in the PMTCT services might stop accessing health
services in order to avoid the social stigma attached to HIV (Papers I and II). Hence, the
examples identified in this research show that clients have multiple objects (according to
activity theory analysis) related to their wellbeing (e.g. physical, social-economic and
psychological wellbeing) and, depending on these objects, clients may not use the health
services or show up as required in the HCAS (cf Figure 1, Paper I, p.7 which shows clients’
folders for those who have missed their appointments). This can be interpreted as a
contradiction between the CAS and the HCAS because clients can prioritize some objects
over their physical wellbeing. Furthermore, the act of clients missing their appointments also
has an impact on HCAS and HMAS efforts to reduce maternal and child mortality rates in
Tanzania.
The third source of disruption in healthcare provision can be related to the ineffectiveness of
mediating artifacts in the HCAS due to external contextual features of clients’ choices. The
findings of my research have shown that client compliance with the requirements and
directives given by health providers is fundamental for the information artifacts to work as
designed in supporting the collaboration of health providers located in multiple HCAS. For
example, a client called Prisca tore up the referral form she was supposed to bring to a facility
for PMTCT care (Paper I). Meanwhile, the same artifact and procedures worked for Joyce,
another client who facilitated the information flow between health programs and facilities.
91
Thus, the success of the referral form in informing other health facilities depends on the
client’s willingness to follow the trajectory. This implies that the occurrence of disruption in
the care trajectory due to mediating artifacts has contextual dependence (the client’s
compliance status). Hence, in distributed collaborative practices, there may be a variation of
occurrence of mediation breakdowns for the same information artifacts used to support
collaborative practices.
Resolution of disruptions: My research shows that the disruptions were resolved either
through health guidelines or by the adaptation/improvisation done by the subjects of HCAS.
For example, disruptions caused by a shortage of resources (medical supplies or skills) were
resolved by the subjects of HCAS following the referral procedure indicated in healthcare
guidelines where PMTCT clients were asked to attend other health facilities if the services did
not exist at their facility (Papers I and V). Also, the disruptions caused by the inadequate
design, or absence, of referral forms, were resolved by the subjects of HCAS through
improvisation. This is identified in my research where some health workers invented informal
patient registries or information practices to improve the overview of client attendance at the
HIV clinic (cf Figure 4 in Paper I, p.12) or improvised to facilitate client referrals.
Moreover, my research indicates that the disruptions caused by conflicting interests between
the object of HCAS and CAS or contextual conditions were also resolved through different
mechanisms. For example, the disruptions in care trajectories caused by clients’ decisions to
prioritize their non-physical-wellbeing objects were resolved by health providers within their
health programs by performing additional tasks. Examples of such coordination practices to
ensure that clients attended all required services were a series of actions as described by
Leontev (1978), for instance, when health workers escorted clients to other health programs,
or interrogated clients in order to get an overview of care conducted in other services, or
reminding clients to attend other services (Paper I). Another form of coordination practices
performed by the health providers is communication by mobile phone with their peers located
in other facilities. This is identified in my research by examples of the interaction between the
(PMTCT and HIV) health providers of Korongo and Chemchem facilities regarding certain
clients’ attendance (Paper V) or interactions between facilities which facilitated Prisca’s
continuation of her care trajectory (Paper I).
Furthermore, the disruption in care trajectories caused by clients’ decision not to attend their
care trajectories as required were resolved by many actors through a new or changed object,
92
which then caused a new activity system to emerge that ensured that clients’ completed their
health programs. I call the new activity system a Defaulter Tracing Activity System (DTAS)
(presented in section 6.2). Figure 17 illustrates the resolution of the contradiction between
HCAS and CAS resulting in the DTAS. The diagram shows how a contradiction emerges
between HCAS and CAS when clients prioritize some of their other objectives over their
physical wellbeing.
CAS
HCAS
Clients
complete their
health programs
DTAS
Figure 17: Resolution of contradiction between CAS and HCAS resulting in DTAS
The previous paragraph indicates that disruptions in care trajectories were resolved through
interrogating and reminding clients or peer interactions among health providers. These
interactions between health providers were mediated by artifacts such as the PMTCT code
written on children’s cards to signal to the health providers to interrogate clients about their
attendance in other, collaborating health programs.
However, further analysis shows that the context has an impact on the resolutions of
disruptions; the resolutions of disruptions in distributed collaboration in the coordination of
care trajectories within health programs and facilities have a contextual dependence of a
higher workload for health providers and in rural/urban disparity. My research shows that, due
to higher workload, health providers might forget to use PMTCT codes (as indicated in health
protocols from HMAS) to remind clients to attend other programs or health facilities (Paper
I). Also, rural/urban disparity might impact on how health providers (in multiple HCAS)
communicate between themselves as a mechanism to resolve absence of clients’ attendance.
For example, in PMTCT services in an urban context, clients could have access to many HIV
clinic options (Papers I and V) and thus it might be difficult for providers in one health
program/facility to get an overview of clients’ care trajectories due to the larger geographical
93
area. It can be relatively easier for health providers in rural facilities to monitor care trajectory
where there are fewer options for HIV clinics. This is identified in my research in Paper V,
where the health provider of the Korongo facility in a rural setting could monitor clients’
attendance at HIV clinics in Chemchem facility because it was the only facility offering the
PMTCT services in the geographical division. At the same time, the providers of the Bwawa
facility in an urban setting could not have such easy monitoring of the care trajectory due to
the availability of many HIV and PMTCT options in this district. Thus the resolution of
disruptions occurring in care trajectory may be subject to the size of the geographical area.
6.2 Collaboration in Ensuring that Clients Complete their Care
Trajectories
This section presents DTAS and other activity systems which are crucial for its facilitation.
6.2.1 Activity Systems and Mediation
To demonstrate collaboration in facilitating DTAS across health programs and facilities, the
following activity systems were identified in this study as important in the accomplishment of
the DTAS: the HCAS, the CAS, the HMAS (all of these represent similar activity systems as
in section 6.1.1); the Community Health Workers Activity System (CHWAS), which is
viewed from the perspectives of the community health workers; the Catchment Population
Leaders Activity System (CPLAS), which is viewed from the perspectives of the catchment
population leaders; and the activity systems of community inhabitants, viewed from their
perspectives. In the next paragraphs, I identify the objects of each of the activity systems
together with their subject- and community elements. Also identified are actions of
corresponding objects as described by Leontev (1978). Moreover, activity system mediations
and also the actions and operations of the DTAS are identified.
The DTAS has the objective of making sure that all clients complete their health programs. It
is recognized that the motive of an activity system can come from global or national concerns
or political considerations (Mursu et al., 2007). The motive of the DTAS is enforced by the
Tanzanian government’s concern about higher maternal and child mortality rates and the
effort to reduce it. Collaboration between different stakeholders from the national to the
community levels is seen as an important aspect in achieving the reduction of maternal and
child mortality rates. Consequently, some community actors (e.g. catchment population
94
leaders and health workers) are also required to participate in non-clinical such health services
as health education, advocacy, social - and community mobilization.
The subjects of CHWAS are community health workers who are designated to work for
specific streets/hamlets. The community element of CHWAS is formed by health providers,
health managers, hamlet leaders, and the catchment population. The artifacts which mediated
CHWAS include their individual mobile phones and also the notebooks used to keep records
of the non-clinical health services activities they conducted within their catchment population
jurisdiction. Similarly, the subjects of CPLAS are community council members who have
been elected to serve in the hamlet, streets or ward local government structures. The artifacts
which mediated the CPLAS include their mobile phones and village registration systems
which keep information about catchment inhabitants, e.g. data on births and death. The
community element of CPLAS is formed by health providers of multiple health programs,
CHW, health managers and its catchment population.
The DTAS is also impacted by the health authority (HMAS) of district and regional levels
which play a crucial role in management of healthcare service provision in the country
through supervision of health facilities and in providing the economic support for the DTAS
(Paper II and III). Moreover, health managers at districts and regions could intervene by
directing health facilities to perform the DTAS (Paper II). Two types of artifacts were used to
mediate HMAS’s action of monitoring healthcare services in health facilities. First, HMAS
set health target to be achieved by each HCAS of a certain health facility (Paper III). Also,
HMIS reports generated from health facilities were the artifacts which informed the managers
if the healthcare targets set for a certain catchment population had been achieved or not.
Activity systems of other community inhabitants could also be involved in facilitating care
trajectories. The objects of community inhabitants include one related to taking care of friends,
relatives, family members and other community inhabitants. The subjects forming community
inhabitants are many. For example, community-based groups of People Living with HIV
(PLWHIV) were required by the ministry to work as role models in preventing ongoing HIV
transmission of HIV. Also, family members, relatives and friends had to play a crucial role in
HIV services, for example, by being adherent assistants who ensure that clients take their
daily medicines and attend their clinic appointments. The artifacts which mediated
community inhabitants include their individual mobile phones, existing social relations with
each other, and also knowledge of family friends, neighbors and other inhabitants. From an
95
activity-theoretical perspective, the achievement of DTAS’ objectives requires the
collaboration of multiple individual activity systems of health professionals, health
administrators, catchment population leaders, community health workers, and other
community inhabitants. Figure 18 gives an overview of the network of activity systems for
facilitating the DTAS.
96
HCAS
Object: Management,
Supervision and
Monitoring of Healthcare
Services
Object: Clients complete their
care trajectories in the
enrolled health programs
Object:
Physical
wellbeing of
hamlet/street
inhabitants
CHWAS
Notebooks, Mobile phones
CPLAS
Mobile phones, Village
Registration Systems
Object: Wellbeing
of the Catchment
Population
Figure 18: The network of activity systems involved in defaulter tracing practices
Other Community Inhabitants Activity Systems
Object: (community spirit of)
Taking care of friends, relatives,
family members and other
community inhabitants
Mobile phones, social relations with each
other, knowledge of family, friends,
neighbors and other inhabitants
Object: Clients
Physical
Wellbeing
Patient registries, Health
program Summaries
HMAS
Healthcare targets
HMIS reports
97
The collaboration between facility and community actors in DTAS is evident in its actions.
The DTAS has three actions: identifying defaulters, tracing defaulters, and providing
feedback to the facility. The action of identifying defaulters is performed within health
facilities in two operations according to Leontev (1978). First, health providers would search
the patient registries/folders to determine which clients are overdue in their health program
schedules (Papers I, II and V). The second operation is to record the name of a client
identified as having missed their appointment(s). The two operations are repeated for all
clients’ records in the patient registry. The outcome of the action of identifying defaulters is
the defaulter list (cf Figures 2, and 4, in Paper II, p. 10 and 12) which contains information
about clients who had missed their appointment(s). Thereafter, health providers would
distribute the defaulter names to other participants who are to be involved in the tracing
defaulter action. Also, health program summaries (on a daily or weekly basis) generated by
each health program mediated identification of defaulters as the summaries provide an
overview of clients’ attendances and defaulting status within healthcare programs.
The action of tracing defaulters involved informing clients about their non-attendances in
facilities and the requirement of resuming the services. The actual action of tracing defaulters
was performed through a number of operations. First, clients would be informed about their
missed appointments. Two main approaches were used to inform clients: by mobile phone or
through home visits. Health providers could employ mobile phone-based follow up to contact
clients who were overdue in their schedules (Paper V). First, health providers would find the
mobile phone number from the defaulter list and call the client. Second, health providers had
to record the responses from the call. Responses obtained from the mobile phone approach
include clients not able to be reached or clients not picking up the phone; client had received
services in other facilities; or is traveling. The two operations had to be repeated for each
client recorded on the defaulter list.
Visits to a client’s home were mostly done by community health workers (CHWAS) (cf
Figure 3, Paper II, p.12) and in a few cases by health providers (Paper V). The home visits
could also be done by catchment population leaders (CPLAS). This is demonstrated in Paper
III where the description of facility four indicates how the village executive officer requested
defaulter names from the health facility for him to take the action of tracing them through
hamlet leaders. The main tool which mediated a home visit was the defaulter list.
Community-based IS tools (e.g. CHW notebooks and village registration systems) could also
98
be used to inform actors about clients’ house location or an overview of recent healthcare
status of clients (Paper II).
Besides information artifacts, the action of tracing defaulters was mediated by humans. For
example, a primary school teacher could be involved in facilitating communication between
clients and providers by telling pupils to inform their mothers about immunization schedules
in the health facility (Paper IV). Also, PLWHIV (People Living with HIV) could convey
messages from the health facility to clients about the need for them to attend their
appointments in healthcare trajectories (Papers II). Thus the primary school teacher, pupils
and PLWHIV become mediators between clients and health providers. Also, health providers
could contact other collaborating HCAS regarding certain clients’ attendance or ask them to
convey messages to clients who reside in their catchment area (Papers IV and V).
The action of providing feedback to health facilities involved informing health providers of
the outcome of traced clients. The following operations were identified as part of providing
feedback to health facilities. First, the person must update the defaulter list based on the
information they had received from the client. Examples of the documented outcome could be
that clients were informed, or they could have received the service in another facility, or they
may have migrated, or were deceased (Papers I and III). The defaulter list with the updated
status of clients has to be returned to the health facilities. Lastly, the health provider had to
update patient registries by documenting the outcome of the action of the tracing of defaulters.
6.2.2 Disruptions and their Resolutions
My study also identified disruptions in the DTAS which were caused by interaction with other
activity systems or the influence of contextual conditions (PMTCT program, clients’ choices,
and rural/urban settings).
My research has indicated that a disruption may occur in the DTAS where tracing the wrong
person is performed by participants because health providers had either wrongly calculated
(due to the manual tools) that clients had missed appointments or they might not have
recorded the data properly during the health service provision (for instance immunizing a
child without recording any information in the child registry) (Papers II and III). My research
indicates that there were mediation breakdowns in the DTAS due to the lack of information
artifacts (e.g. defaulter tracing guidelines and official tools to support generation of health
programs summaries) or economic support (Papers II, III and IV).
99
The previous paragraph implies that the disruptions occurring in the DTAS are mainly caused
by secondary contradictions (erroneous identification of defaulters) and quaternary
contradictions from HCAS (incomplete recording of information) and HMAS (lack of
information artifacts or economic support). However, it is acknowledged that the activity
system itself is the context which involves specific objects and goals (Nardi, 1996). It also
became obvious from my findings that the PMTCT services acted as a context that impacted
on the emergence of disruptions in the interactions of the activity systems that were
supposedly collaborating in DTAS. HIV/AIDS is infectious/communicable to other people –
for example, to infants in the womb, health professionals attending the clients, or
family/friend members. As a consequence of its infectious/communicable nature, PMTCT
services demand close follow-up in order to prevent further infection of other people. In
contrast, in the vaccination services there are no serious short-term side effects and missed
immunization can be conducted once a client is found.
As a result of the nature of PMTCT services, their health provider had stronger incentives and
means to trace referred clients in other health programs and facilities belonging to different
districts or catchment population settlements (Papers I, IV and V). However, Paper III shows
that tracing defaulters in vaccination services was not conducted for clients who reside outside
the catchment population area designated by a certain health facility. Thus, the coordination
tasks performed by health providers across health programs and facilities can differ due to the
influence of the object of the specific diseases it targets.
Moreover, my research shows that there were variations in participants, tools and mediation
breakdowns between the two programs. For example, in contrast to the PMTCT program,
there is a lack of economic support for the vaccination program so that CHW may not have
notebooks to facilitate the action of tracing defaulters. Also, fewer community workers
participated in tracing defaulters in vaccination services because they did not get any
incentives from HMAS for tracing defaulters (Paper II). On the other hand, PMTCT DTAS
drew the support of donors who provided economic support to facilitate the practice. For
example, CHWs working with HIV-related health services (including PMTCT), were
provided with notebooks. Moreover, donor support allowed health providers within PMTCT
services to use mobile phones to contact clients about their testing schedules or missed
appointments. There are no such practices within the vaccination program.
100
Furthermore, due to the social stigma attached to the HIV-AIDS pandemic, there were
differences between the two programs in the manner of interaction between facility and
community actors during the tracing of defaulters. My research shows that in the PMTCT
services, the contacting of clients was performed discreetly to protect the privacy of
individuals (Paper II). Moreover, the need for protecting the patients’ privacy in PMTCT
services resulted in restricted involvement of CPLAS, or other community inhabitants. For
example, in the PMTCT program there were designated CHW used to support home-based
care services for HIV program. The designated CHW could involve other inhabitants/nonhealth staff but they were not informed about the reason for the search for the client. In
contrast, in the vaccination services client contacting could be conducted more freely and
involved any available community member, drew on the support of the local government
leadership (CPLAS) and the reason for the search for the client was made explicit to the
participants. Moreover, my research indicates that in the rural area, village authorities
(CPLAS) can use community guards to bring children to a facility to be immunized (Paper II,
p.12). In comparison, the home visit may not be very effective since clients might refuse to be
visited or the CHW might not inform the client in case there were family members who were
not aware of client’s HIV status (Paper II, p.13). In such cases, the CHW had to wait for other
opportunities to inform the client, such as in religious meetings, burial or wedding
ceremonies, or on the road. Consequently, in the HIV-related DTAS, documenting of the
outcome action might not be conducted immediately.
Other external contextual features such as clients’ decisions, and rural/urban disparity also
had an impact in the DTAS. This was identified in my research as some clients would not
immediately come to facilities when informed about their defaulter status (Paper II). This
disruption arises because clients still prioritized their other objectives over their physical
wellbeing. This kind of disruption in the work done could result in the tracing action having to
be repeated several times.
My research has also demonstrated that rural/urban disparity could influence the emergence
of mediation breakdowns/disruptions between different activity systems which are supposedly
working together in tracing defaulters’ practices. For example, since mobile phones are in
wider use in urban than in rural areas, PMTCT services located in urban areas use them to
contact clients (Paper V). On the other hand, PMTCT service located in rural areas used
community health workers to contact clients.
101
The physical location of a facility can also determine whether a patient registry could function
effectively as a mediating artifact to facilitate interactions between health providers and
clients during care service provision. This is identified in my research by mediation
breakdowns faced by the health facility located in the town center when using patient
registries (Facilities; one in Paper III, and Mto in Paper V). For example, due to a higher flow
of clients, such health facilities faced difficulties when searching for clients’ names to identify
who had missed appointments or to get an overview of clients’ testing activities. The same
patient registries were used successfully by health providers in health facilities in rural areas
(Facility four, and Korongo in Papers III and V respectively) or in the outskirts of urban
settings with fewer clients (Facility two).
Resolution of disruptions: My research shows that some of the disruptions within DTAS were
resolved by the subject of HCAS. For example, health providers had created informal
defaulter tracing guidelines (Paper II) and patient registries to support identification of
defaulters (cf Figure 5, p. 13 in Paper V). Moreover, health providers adapted patient
registries so that they can support multiple practices of documenting, reporting and tracing
defaulters (Papers III and V). Also, disruptions in the interactions between health providers
and clients caused by contextual conditions of the facility (higher flow of clients due to its
physical location) resulted in health providers changing the layout of the patient registries
(Papers III and V).
They had also created informal health programs summaries which were used in daily and
weekly facility meetings to gain an overview of the health programs’ performance (Papers III,
IV and V). Furthermore, the problem of economic limitation was resolved by using the
activity systems of other people (e.g. relative, a client’s neighbor) who mediated between
health providers and clients. For example, the problem of limited economic support from
HMAS within the vaccination program was resolved by using neighbors to convey messages
from health providers to their clients about the need to attend their care trajectory in the
facilities (Papers II). Also, a client’s confidante (e.g. a relative) could be used to facilitate
communication between client and health providers regarding testing schedule reminders in
PMTCT services, where a client did not own a mobile phone (cf Figure 6 in Paper V, p.16).
However, some of the resolutions made by health providers were impacted by contextual
conditions of the rural or urban setting. For example, it is difficult for urban health workers to
convey messages to clients through neighbors and other community habitants because of the
102
attitude of urban dwellers, where they have less interaction and hence less knowledge of each
other. This is exemplified in Papers II, III and V, where it is indicated how community
inhabitants’ involvement in defaulter tracing practices was stronger in rural areas as compared
to urban areas.
6.3 Collaborative Work in the Design and Development of the DHIS
Tracker
This section presents an analysis of collaborative work in the design and implementation of
the DHIS Tracker software.
6.3.1 Activity Systems and Mediation
To demonstrate collaboration between global and local actors, three main activity systems
were identified in this study: the Global Team Activity System (GTAS), viewed from the
global teams’ perspective; the Implementation Mediator Activity System (IMAS), viewed
from the perspectives of people who are involved in the implementation of DHIS Trackers
(including myself); and the Defaulter Tracing Activity System (DTAS), viewed from the
perspectives of both the facility and community actors involved in facilitating clients’ care
trajectories. 20 Moreover, actions of corresponding objects as described by Leontev (1978) are
identified. Furthermore, activity systems’ mediations are identified. In Figure 19, the various
activities representing GTAS, IMAS and DTAS are illustrated.
In the research as shown in Paper IV, the GTAS had the objective of developing the standard
DHIS Tracker software which can support monitoring of clients within their care trajectories.
More specifically, GTAS has two actions: designing and developing the DHIS software, and
capacity building with the purpose of disseminating DHIS skills. The subjects of GTAS are
people based at the University of Oslo involved in the development and capacity building
tasks for the DHIS Tracker software. The community element of GTAS includes DHIS
implementers and other HISP researchers/members who are located in different countries.
20
DTAS is represented in 6.2
103
Understanding of
DHIS Tracker work
practices
GTAS
Understanding of
PHC practices
IMAS
DTAS
Figure 19: The network of activity systems involved in DHIS Tracker Design
The DTAS has the objective of making sure that all clients complete their health programs.
The achievement of these objectives requires the collaboration of multiple actors, including
health professionals, health administrators, catchment population leaders, community health
workers, and other community inhabitants. The community element of DTAS is formed by
health providers of other collaborating HCAS, other catchment population inhabitant (e.g.
PLWHIV, primary school teacher), and DHIS implementers, and developers. Multiple
artifacts mediated DTAS, including patient registries, defaulter lists, health program
summaries, mobile phones and actors’ acquired skills related to how work is accomplished.
Paper IV shows my role and those of other people as implementation mediators of the DHIS
Tracker software. The objectives of IMAS are to customize/re-design and implement the
DHIS Tracker in health facilities. In particular, IMAS performs three actions:
customizing/configuring the standard DHIS Tracker system to fit the work practices of health
facilities; gathering software requirements; and providing feedback to the GTAS. The subjects
of IMAS are people who participate in implementation of the DHIS software. The community
element of IMAS includes global software developers, other DHIS implementers who are
located in different countries, and health professionals together with non-health professionals
who are involved in ensuring that clients complete their care trajectories.
The artifacts which mediate IMAS include their gained skills, which are related to both the
local work practices around DTAS and DHIS software. This is exemplified in the findings
where I (as an implementation mediator) acquired an understanding of DTAS through
104
ethnographic studies conducted in health facilities. This is also exemplified in the findings
where I acquired the understanding of the DHIS Tracker through the activities conducted by
GTAS by participating in mailing lists, group discussions, and training workshops. For
example, I learned that defaulter tracing practices in the DHIS Tracker could be performed
through personal dashboards, multiple data entry forms or program-tracking functionalities.
Moreover, different colors in the data entry form or personal dashboard signals if the client
has completed the program (green), or has an incomplete status (purple), or has an overdue
status (red), or is scheduled for a future appointment (yellow). Figure 20 21 shows sample
DHIS Tracker screen elaborating the status of under-fives’ attendance in the immunization
and PMTCT services health programs in one health facility.
Figure 20: Sample DHIS Tracker screen with different colors showing children status in the
children under-five program
I also learned that information about defaulters could be viewed through the program-tracking
functionality. The SMS functionality included in the DHIS Tracker system could be
configured by implementers to send automatic reminders to clients before their scheduled
appointment or after a missed appointment. Also, the personal dashboards included
functionalities for rescheduling clinic appointments and creating comments (such as
“migrated,” “deceased” or “travelled”) about clients’ attendance status.
Thus, my research indicates that IMAS mediates the interactions and learning between DTAS
and GTAS in the design of DHIS Tracker.
21
The figure was removed due to lengthy limitations
105
6.3.2 Disruptions and their Resolutions
The interaction between the three activity systems resulted in disruption, since GTAS and
DTAS take place on difference physical sites: the global at the University of Oslo and locally
in the health facilities, e.g. at PHC in Tanzania. Thus, IMAS’ practices (when they try to
connect the interaction between the local with the global) are prone to the emergence of
disruptions which may require resolutions. This section presents disruptions and their
resolution which may exist in the interaction between the GTAS and the DTAS as mediated
by the IMAS.
It has been argued that needs of health practitioners can also be identified through analysis of
the mediation breakdowns they face in their work practices (Igira, 2008b); this was identified
in this research. Having a role of implementation mediator, I had to identify which artifacts
mediate practices and, in addition, the actual mediation breakdowns. For example, the
ethnographic approach helped me to identify some mediation breakdowns in the defaulter
tracing practices: an error in processing; being faced with an economic limitation; and a lack
of defaulter tracing guidelines. Besides that, I had learned that users had resolved some
mediation breakdown by creating informal artifacts (e.g. health program summaries) (cf
Figure 2, Paper I, RCH reports in Facility three’s description, Paper III) which were not
included in the software developed by the GTAS.
The understanding I had acquired from the DHIS Tracker informed me that it could be used to
address some of the identified challenges in the ethnographic study; for example, the DHIS
Tracker could be utilized to facilitate the searching process, thereby eliminating errors in the
defaulter identification. Also the SMS functionality included in the DHIS Tracker system
could be used to facilitate automatic reminders to clients about their scheduled appointment or
missed appointment, thereby reducing the impact of problem of economic support limitation.
The understanding of health workers’ practices gained through the ethnographic study helped
me as the implementation mediator to identify mediation breakdowns within the DHIS
Tracker’s functionality. For example, within the DHIS system there was no functionality for a
health program summary (on a daily or weekly basis) or to print defaulter lists which could be
carried by community workers during their home visits. Thus, the first source of disruption in
the use of the DHIS Tracker as mediating artifact for DTAS could be interpreted as the
contradiction between GTAS and DTAS. This is because the GTAS has developed a software
functionality which is different from the DTAS’s actual practices.
106
The resolution of the first disruption was done by me and GTAS. As the mediator, I had to
convey the gained understanding (regarding DTAS and DHIS Tracker) to the GTAS. The
mediation done by IMAS resulted in the GTAS re-considering the design of the DHIS
Tracker. The DHIS Tracker was changed by the GTAS because of the improved
understanding around the use of artifacts by health providers such as the defaulter list and
their invented health program summaries. Figure 21 presents a standard DHIS Tracker screen
from the demo version that was created by the GTAS after the ethnographic study by IMAS.
Improved
Feature
New
Feature
Figure 21: Standard DHIS tracker system with changed features
The implementation of the DHIS Tracker in the two health facilities was conducted by three
people including myself. The implementation of the DHIS Tracker faced disruptions arising
from contextual features of the two health facilities. For example, there was a lack of physical
space with enough security and electrical power for the computers. The resolution of this
disruption needed to be configured by the implementation mediators in order to provide the
connectedness between the design and use of DHIS Tracker system. We 22 discussed with
health providers a possible way to counter the disruptions, for example, by finding alternative
places for the computers. Moreover, we had to improve physical security around doors and
windows and install electrical outlets and light in the rooms where computer was to be kept.
The resolution of the disruption emerging from the physical context’s constraints illustrates a
form of articulation work done by IMAS which may be invisible (see Star and Strauss, 1999)
to other people including the GTAS.
22
The three people who had participated in the implementation of the DHIS Tracker
107
The third source of disruption in the participation of practitioners in the design activity was
related to their lack of skills in the general computer application and the DHIS Tracker. We
resolved this disruption by training health workers and some community workers in both
general computer applications and the DHIS Tracker. In addition to the training of the DHIS
Tracker, we engaged health workers in post-implementation design practices, for example
through the use of the system in the real practices and in focus-group discussions.
The understanding gained by health workers through a combination of training and use of the
system enabled them to indicate the differences in the understanding of the implementation
mediator (me) gained during the earlier ethnographic study. For instance, health providers
called our attention to the local context where few people owned mobile phones. Moreover,
the participants made us aware of need for further design and what has to be conveyed to the
global design team so that reminders could also be sent to people not registered in the
programs, such as the head of the primary school, a client’s neighbor, hamlet leaders or
anyone who could facilitate the defaulter tracing action. This was in contrast to the DHIS
Tracker’s design with its focus only on people registered within health programs and an SMS
could be sent from the facilities to clients only.
Thus, the disruption in the use of the DHIS Tracker as a mediating artifact for DTAS
discovered in post-implementation participatory practices can be interpreted as the
contradiction between GTAS and DTAS. For example, the disruption emerged due to cultural
differences between the GTAS and the DTAS: Western (individual perspective) versus
Ubuntu tradition (relational or community perspective). Besides, the occurring disruption
illustrates how in the design of DHIS by the GTAS, it is taken for granted that everyone owns
a mobile phone. Hence, the global team had a limited understanding of work practices in
terms of who had access to a mobile phone. On the other hand, the disruption in the DHIS
functionality illustrates the contradiction between the IMAS and GTAS. This is because the
request to change the software I had made as implementation mediator to the global team did
not include the need for the software to include community actors’ practices within the DTAS.
The previous paragraphs indicate that the interaction between GTAS and DTAS through
IMAS is impacted by external contextual conditions: physical context of facilities (placement
of computers, security & electrical power); cultural differences (individual versus relational
perspectives); and social-economic situation (limited ownership of mobile phones).
108
We as implementation mediators resolved this disruption when we expanded local
participative activities, by doing more interviews, and demonstrating a system to community
leaders. The local participative design activities resulted in the implementation mediators’
decision to re-design the local DHIS Tracker system by adding mobile phone numbers of
catchment population leaders and anyone the client had access to (e.g. a neighbor, friend, and
relative). This was added to the client registration form for all health programs. Furthermore,
we decided that we to use an ad-hoc SMS solution (as an improvised solution) at the local
level to cater for a missing functionality in the DHIS Tracker. Additionally, the demand for a
more permanent functionality was communicated to the global level (GTAS).
6.4 Summary
Table 6 presents the key points arising from analysis of the research findings in relation to
answering the research questions addressed in the study.
Table 6: Summary of the Key Findings
Research Questions
What are the
Research Answers
Multiple actors are involved in facilitating care trajectories (health providers,
characteristics of
community health workers, community leaders, health managers at district and
distributed
regional levels, clients’ neighbors, etc.)
collaborative work of
The interaction between distributed actors is mediated both by artifacts and
the PHC settings?
humans
Contextual features can support or limit interactions between health providers
How does context
who are located across health program, across health facilities; between health
impact the distributed
facility and community actors; or between the global team, implementers and
collaborative work of
healthcare providers
the PHC settings?
Contextual features can limit effectiveness of coordinative artifacts used to
mediate interactions between distributed actors
109
CHAPTER 7: DISCUSSION AND CONTRIBUTION
The aim of this chapter is to provide an account of how the dissertation relates and contributes
to the academic discourse in which it is positioned.
7.1 HIS in Developing Countries
This thesis contributes to HIS in developing countries by providing rich insights of
operational primary healthcare information systems operating in maternal and child healthcare
services. I have argued in the introduction chapter that the HIS research field has put more
emphasis on the managerial aspect of HIS and less on understanding operational primary
healthcare information systems as researchers have tended to focus on design of
functionalities or design practices (see for example, Korpela et al., 2002; DeRenzi et al., 2011;
Ngabo et al., 2012; Karara, 2013), and less on understanding the work practices around the
use of information systems tools to accomplish the work. Thus, this research extends the
understanding of PHC-IS by providing a rich analysis of health workers’ practices related to
the use of IS tools to support coordination of clients’ trajectories. Further, the research has
shown that official tools are inadequate to support the coordination of care trajectories and,
thus, before designing computerized systems, there is a need to get a better understanding of
the work domain (such as, the work practices of health workers in coordinating care
trajectories, the ways health workers handle challenges, and their current solutions) and, in
addition, how to approach design and implementation in an appropriate manner (e.g. to how
to conduct ethnographic studies and participatory design by IT facilitators). I argue also that
we should not only focus on understanding how ICT can be designed to support PHC but also
try to have better understanding of the work practices within which the technology will be
used.
In particular, this research has demonstrated the challenges that health workers face in the
utilization of patient-oriented information systems. It is acknowledged within the HIS
discourse that design and implementation of information systems at the primary healthcare
level are faced with challenges related to poor design of information system tools, a lack of
tools and the non-use of tools (see for example, Igira, 2008a; Ngoma and Igira, 2012; Kanjo,
2011; Damtew and Moges, 2013; Heunis et al, 2011). This dissertation extends the
understanding of the challenges facing primary healthcare information systems by showing
that there are also coordination challenges arising from inadequate design or shortage of
110
coordination artifacts and protocols, such as referral forms without a feedback section, and
lack of defaulter tracing guidelines, and health program summaries. Moreover, the
dissertation indicates that clients’ decisions and other contextual conditions can challenge the
effectiveness of the designed patient-oriented artifacts (e.g. patient registries, referral forms).
Furthermore, my research contributes to the understanding of context in the HIS field. It has
been argued that the understanding of context is important in the design, development, and
implementation of appropriate HIS (Lippeveld and Sauerborn, 2000; Igira, 2008; Avegrou,
2010; Tiihonen et al., 2010). I have argued in the introduction chapter that the HIS research
field does not explicitly emphasize the impact of context in the use situation but, rather, puts
more emphasis on context for information system development activities (computer literacy,
skills, limited resources, etc.). Thus, my dissertation contributes to the understanding of
context within HIS literature through highlighting the impact of contextual features on health
workers’ work practices, for example, by showing that context affects the interaction between
distributed health workers and their use of coordination artifacts. Moreover, I have elaborated
the other types of contextual conditions which shape health workers’ practices (e.g. the nature
of the health services) which could be useful for design of PHC information systems.
Lastly, my research provides specific contributions to the HIS field by focusing on the
analysis of defaulter tracing practices in maternal and child healthcare settings. Defaulter
tracing practices have been recognized as an important aspect in facilitating care trajectories
within tuberculosis, HIV/AIDS and PMTCT (see for example, Fraser et al., 2007; Thomson et
al., 2011; Nglazi, et al., 2010; Chadambuka et al., 2012), but these studies have paid greater
attention to the design of EMR and less to the practices of health workers in using the EMR
system to support care. The dissertation argues that to retain clients effectively, there is also a
need to pay attention to policy, practice and design of health information systems. Previous
research also provides limited insights about the involvement of community inhabitants in
defaulter tracing practices. Therefore, my research has extended the understanding of
defaulter tracing practices, where the involvement of the community in the practice has been
contested, and indicates the need to improve the design of IS (e.g. patient registries, defaulter
guidelines and the DHIS Tracker) to support facility and community collaboration.
111
7.2 Research on Collaborative Practice
This research contributes to the understanding of distributed collaboration around the
facilitation of the patient trajectory. Research on collaborative work has been criticized for
lack of empirical insights about distributed collaboration around patient trajectories (Grisot,
2008; Fitzpatrick and Ellingsen, 2013). Previous research has shown that the facet of
distributed collaboration can be between health facilities (e.g. Robertson et al., 2010; Li and
Robertson, 2011), or between health providers and home care services (e.g. Piras and Zannuto,
2010). Moreover, design, implementation and development of patient-care software require
collaboration between health practitioners and IT professionals (Bandon et al., 2005). While
most of these studies discuss one of the facets at a time, this study broadens the perspectives
on the facets of distributed collaboration by demonstrating that collaborative practices around
the facilitation of patient trajectories can span health facilities and programs; health facility
and community levels; and between health practitioners, implementers, and global software
development teams. The extended facets of distributed collaboration may have implications
for the understanding of the actors involved in facilitating care trajectories within the
distributed healthcare services; mediation of interactions between distributed actors; and the
impact of work-context on collaborative practices and design practices. Taking these insights
into consideration has implication for how we design and implements distributed
collaborative systems.
7.2.1 Actors involved in Facilitating Care Trajectories
Previous research acknowledges that multiple actors are involved in the accomplishment of
distributed healthcare services (e.g. Schmidt et al., 2007; Robertson et al., 2010). Likewise,
my research indicates that the accomplishment of patient trajectories depends on the
collaborative practices of multiple professionals who are located across health programs and
facilities, community and administrative levels. The findings thus emphasize roles played by
multiple actors in the accomplishment of patient trajectories. In addition, my research extends
the understanding of the involvement of non-health professionals in the care trajectories.
While the previous research offers insights on the involvement of different types of nonhealth professionals in the facilitation of care trajectories, such as medical clerk recorders,
patient transfer staff, IT personnel, social workers and client/family members within
healthcare trajectories (see for example, Bandon et al., 2005; Abraham and Reddy, 2008;
Bossen et al., 2012), the dissertation also highlights that other types of community inhabitants
112
are involved in the facilitation of care trajectories: community actors (e.g. the client’s
neighbors, the primary school teacher, PLWHIV, catchment population leaders). More
specifically, my research points to the involvement of any non-health staff in the facilitation
of distributed cooperative work. To build on previous research, the study emphasizes the need
to consider other types of non-health professionals groups in the design and implementation
of healthcare information systems to support coordination of patients’ trajectories (see for
example, Abraham and Reddy, 2008; Bossen et al., 2012). The understanding is important so
that development of collaborative systems can also pay attention to practices of other types of
non-health professionals who are involved in coordinating patient trajectories. This is
exemplified in the research where health providers requested the modification of the DHIS
Tracker so that SMS messages can also be sent to other people instead of just being confined
to the relationship between clients and their health providers.
7.2.2 Mediation of Interactions between Distributed Actors
My research indicates that multiple protocols and artifacts (e.g. healthcare guidelines,
defaulter tracing guidelines, patient registries/folders, mobile phones, referral forms, defaulter
lists, community health workers’ notebooks, village registration systems, and health program
summaries) are used to support interactions between distributed actors involved in facilitating
patient trajectories. Thus, the dissertation strengthens the understanding that interactions
between multiple actors of distributed collaboration is mediated by multiple coordinating
artifacts and protocols, which are both given and invented (e.g. Fitzpatrick, 2004; Schmidt
and Wagner, 2004; Reddy et al., 2010; Bansler et al., 2013).
Previous research acknowledges that interactions between distributed collaborating actors is
not only artifact-mediated, but also human -mediated (Engestrom et al., 2003; Piras and
Zannuto, 2010; Li and Robertson, 2011). It became obvious in the study that mediation
between distributed collaborators can also be done by human beings. This is exemplified in
the findings by roles played by catchment inhabitants (e.g. catchment population leaders,
primary school teachers and pupils, PLWHIV, CHW) who mediate between clients and health
providers. Likewise, it is illustrated by the role played by clients who move information
artifacts between distributed health providers across health programs and facilities. It is also
exemplified by the role played by implementation mediators who mediate the interactions
between local and global actors in the design of the software. Hence, the study strengthens the
113
understanding that interactions among distributed collaborators are mediated both by artifacts
and human beings.
The role performed by human mediators is crucial in the accomplishment of the care
trajectory because they facilitate coordination between the facets of distributed collaboration
such as between clients and health providers, or across health programs and facilities or local
PHC and global team. However, while previous studies recognize the role of human
mediators, they have not explicitly discussed the challenges human mediators face when they
try to manage the interdependencies between distributed actors. On the other hand, the
research extends the understanding within distributed collaborative research that disruptions
can emerge not only because of artifact design, but also due to practices of the people who are
mediating activities. This is exemplified by disruptions caused by clients when they choose to
default on their care: the referral form will not function as a means of coordination between
health providers, or implementation mediators who may forget to convey some software
requirement. This implies the need within the collaborative research to pay attention not only
to artifact mediation but also the practices of people who perform mediation between multiple
actors.
7.2.3 Understanding of the Impact of Context on Distributed Collaboration
My research contributes to the theme of distributed collaboration by emphasizing the role of
played by contextual conditions in collaborative practices around the facilitation of the patient
trajectory across multiple settings. Previous research suggests that spatial arrangement in the
workplace can shape collaborative practices by limiting or supporting interactions between
collaborators due to differences of coordinative artifacts between the physical settings
(Robertson et al., 2010; Li and Robertson, 2011). Similarly, my research has shown that
contextual features can support or limit interactions between collaborators. For example, the
dissertation has demonstrated how rural/urban disparity has an influence on the interactions
between health providers who are in different health programs or facilities; or how the nature
of the disease influences interactions between community and facility actors; or how the
physical context of the facilities and cultural differences influence the collaboration between
local and global actors in the design of the software. Also, my research has shown that
contextual features can support or limit the use of coordination artifacts. Moreover, my
research has demonstrated that mobile phones are used more in health facilities within urban
settings compared to the rural area, or more in the PMTCT program compared to the
114
vaccination program. The findings thus emphasize the role played by contextual conditions in
shaping distributed collaborative practices.
Beyond that, the study highlights that contextual conditions have a role to play not only for
the interaction between collaborators or variation of coordination artifacts but also for the
effectiveness of coordination artifacts and, hence, the resolution of disruptions. For example,
clients’ compliance status can make the referral forms mediate practices between health
providers (across health programs/facilities) effectively in some instances and not in others. It
is also demonstrated in the research that urban health practitioners invent patient registries to
monitor clients’ attendance in HIV program area. At the same time, there are fewer
disruptions and adaptation of patient registries in rural settings because the contextual features
support easier monitoring of clients. This implies that, in rural areas, the contextual features
are more adapted to collaborative practices in comparison with the contextual features of
urban settings. It also implies that the emergence of mediation breakdowns due to
coordination artifacts within distributed collaboration is dependent on contextual features.
Thus, we need to understand how some contextual conditions make possible or impossible the
coordination practices of distributed collaborators.
While the study emphasizes the role of context, my study argues that it is also relevant to
understand the role of context beyond a specific spatial setting. While the literature
acknowledges that context has influence on collaborative practices, it is mostly equated to
physical space (e.g. Harrison and Dourish, 1996; Ciolfi et al., 2008; Robertson et al., 2010). In
contrast, my research has shown that contextual features that influence work practices can be
both of macro- (e.g. socio-political, geographical, organizational culture, human resources –
higher workload – and economy) and the detailed level/micro- work practices (e.g.
relationship between clients and health services). For example, the use of human mediators
(e.g. community inhabitants) can be limited or supported by contextual issues such as the
nature of the disease or urban geographical settings. It also became obvious in my research
how health providers faced mediation breakdowns in the use of coordination artifacts due to
social-cultural-economic characteristics of the working situations, specifically at the
operational and action levels of the work practices.
Thus, my research goes beyond a limited spatial understanding of context by showing the
other forms of contextual conditions which influence distributed collaboration. This implies a
need to pay attention to the influence of work-oriented contextual conditions, such as disease115
specific issues, clients’ decisions, or the socio-geographical characteristics of the facility
(whether located in a town center or in a rural setting), or cultural issues (e.g. non-Western
culture).
As such, my findings indicate that the design of collaborative systems needs to take into
consideration the aspects of local contextual conditions of health facilities. This is also
identified by Balka et al. (2008) and I specifically discuss more about work-oriented
contextual conditions. This extended view of influence of the work context in collaborative
practices is useful because it can have design implications. This is exemplified in the
dissertation where the re-design of the standard DHIS Tracker system by the global team
included a form with a defaulters’ list which could be printed and hence carried about by
community health workers during home visits. The change is due to the fact that in the
settings of rural Tanzania there is limited use of sophisticated mobile phones and computer
systems. It is also exemplified in the dissertation where the implementers re-designed the
local instance of DHIS Tracker system to include any mobile phone numbers to which clients
had access in the client registration form for any of the health program (e.g. a neighbor, friend,
relative, or catchment population leaders). The redesign of the software would not be needed
if all clients owned mobile phones.
7.3 Participatory Design
This research contributes to the PD community on the role of design facilitators (here called
implementation mediators) in distributed design by focusing on their activities in design-inuse. Previous research acknowledges that design continues in use and the need for
participatory designers to focus on design-in-use (e.g. Henderson and Kyng, 1991; Ehn, 2008;
Bratteig et al., 2012). Also, previous research recognizes the importance of design facilitators
who enable cooperation and dialogues in situated participative practices (Kensing and
Halskov, 1991; Blomberg et al., 1993; Light and Akama, 2012). However, most of the
participatory design practices have focused on design-before-use, though the call that design
continues in use has been there for a long time. In addition, the design facilitators’ discussion
in the literature is based on their practices in the design-before-use sessions. This study has
focused on the role of implementation mediators who facilitate the interactions between
distributed practitioners and software developers during the design of software for managing
patient data. Moreover, my study has indicated that facilitators may face challenges when they
try to facilitate participation of practitioners within distributed design activities. Hence, my
116
study broadens the understanding of implementation mediators by indicating their practices in
design-in-use situation; for example, by indicating how their role is crucial in facilitating
participatory activities in the design-in-use.
PD as a design approach recognizes the importance of the skills of practitioners and design
facilitators in the design and implementation of technologies (Hartswood et al., 2002; Light
and Akama, 2012; Kensing and Greenbaum, 2012; Van der Velden and Mörtberg, 2014). This
approach together with ethnography has helped to address the existing contradictions between
the understanding of global developers and local practitioners about practices around
facilitating clients’ care trajectories. Implementation mediators gained knowledge of health
workers’ practices and their artifacts through ethnography. Such understanding informed
mediators about how tools mediated the use situation and contrasted the existing DHIS
Tracker design, which is the software developed globally. The global development team
gained the understanding through the implementation mediators about use practices that
allowed them to re-design the software.
Thus, my study implies that it is necessary to pay attention to design facilitators and
practitioners’ practices when the system is in use; to pay attention not only to design-beforeuse but also to design-in-use. This is because further learning and understanding of the
implemented system can be acquired by design facilitators and practitioners. My research
indicates that, during this period, implementation mediators were able to organize
participatory design activities at the local level that allowed the practitioners to contribute to
the design process. This is demonstrated in the findings where practitioners gained
understanding of the software through training, by using the system, and through focus-group
discussions. These activities helped the implementation mediators to gain an understanding
that contrasted with the previous understanding from the ethnographic studies. The
understanding contributed to further development of the software. This indicates that
interventions of design facilitators in the design were based on the experience gained in the
use situation. Thus the design facilitators’ role is very important in facilitating participatory
activities during design-in-use.
Furthermore, previous research recognizes that design process and participation activities are
situated and there is a need for design facilitators to take into account practitioners’ practices
and views on cultural issues (Suchman, 2002; Kensing and Greenbaum, 2012; WinschiersTheophilus et al., 2010; Light and Akama, 2012). This situatedness was also experienced in
my study where implementation mediators’ activities in the practitioners’ settings were
117
constrained by contextual elements related to practitioners’ working conditions. As identified
in the findings, this forced the implementation mediators to perform both design- and nondesign-related activities (e.g. related to placement and security issues in the facilities) in order
to facilitate collaboration between global and local actors. In addition, my research indicates
that it may be difficult to envision the situatedness of collaborative practices in the designbefore-use phase. This is indicated in the findings, since it was difficult for implementation
mediators to gain knowledge on how cultural issues both supported and limited collaborative
practices through the ethnography study phase alone.
Though it is identified in the PD research that it might be relevant for design facilitators to
have an understanding of the cultural issues (Winschiers-Theophilus et al., 2010; Light and
Akama, 2012) this may be difficult to achieve in distributed global software development
when software developers are not aware of the cultural issues that are embedded in local
practices. This research has for example shown that few people owned mobile phones and that
social connections between people affect the relationship between health providers and
clients. That is, in western culture, the provider and client relationship tends to be individual
and more personal. In Africa, the provider and client relationship is more communal as it
involves other people, like the family, neighbors and other community inhabitants. The
representations of such relationships in the DHIS Tracker were far from the reality. In order to
understand/deal with the situatedness of design process and participation activities in
designing distributed global software, my study suggests that there is a need for global
developers to pay attention to implementation mediators’ situated activities in distributed
participatory design in order to facilitate cooperation between designers and practitioners.
This knowledge was impossible for global developers to gain without the implementation
mediators. By this, my finding emphasizes the need to adopt PD approaches in the design of
collaborative systems in order to make the system fit the local contexts of practitioners
(Kensing and Blomberg, 1998). Therefore, the role of implementation mediators in distributed
design remains crucial in facilitating design-in use.
This thesis contributes also to the PD literature of developing countries in two ways. First, the
study supports findings from previous studies done in Malawi, Tanzania and Mozambique
regarding challenges to participation of practitioners in the design and implementation of ICT
at health facility level. For instance, previous research reports that participation of health
practitioners is challenged by their lack of computer skills and busy schedule, and also a
limited domain knowledge among primary healthcare workers (Nhamposa et al., 2004;
118
Elovaara et al., 2006; Kimaro and Titlestad, 2008; Chawani et al., 2014b). Confirming
previous PD research in developing countries, the study emphasizes the problem of computer
skills which can challenge PD activities and identifies the need for training. This is identified
in my research as health providers had limited computer skills and, as a consequence,
computer training was mandatory. Second, beyond confirming findings from previous studies,
the thesis addresses a gap in the PD literature concerning the mediators’ role across different
levels of software development and implementation (local health facility and global level).
While previous research acknowledges the importance of mediators in facilitating the
participation of practitioners within distributed design settings (e.g. Titlestad et al., 2009; Braa
and Sahay, 2012), most of these studies seldom describe design-in-use practices of mediators.
Thus, the thesis addresses a gap in the PD literature concerning mediators’ role across
different levels of software development and implementation (local health facility and global
levels).
Previous studies highlight the importance of the role of implementation mediators as a liaison
between a global software development team and the practitioners in the collection of
requirements, customization and participatory practices (see for example, Puri et al., 2009;
Titlestad et al., 2009; Starring, 2011; Braa and Sahay, 2012; Saugene, 2013). Also, previous
research recognizes that implementation mediators can be challenged by lack of
customization skills during the adaptation of software to local requirements and need training
(Saugene, 2013), or by power asymmetry in the interaction between implementation
mediators and the global team in requirements’ collection process (Chawani, 2014b). This
study adds to previous research by pointing to challenges facing the participation of
implementation mediators due to work-contextual conditions of the practitioners’ settings.
This research, therefore, highlights the need not only to focus on understanding the
interactions between implementation mediators and the software development team but also
to pay attention to the interaction of implementation mediators and practitioners during
activities that take place after the implementation in order to facilitate design and use of
global software.
Finally, it is recognized that users of the systems can participate in the design by providing
their views on the new system; for example by contrasting old and newer work practices
(Karasti, 2001b). It is recognized also that practitioners’ improvisations are important for
facilitating their participation in the design process (Caroll, 2004; William et al., 2005; Safad
and Faraj, 2010). When it comes to designing distributed global software, implementation
119
mediators play a crucial role in including the users and their views (Iivari, 2011; Titlestad et al,
2009; Braa and Sahay, 2012). This study has shown that these views and improvisations were
learned during system use. First, the dissertation provides an ethnographic description of
practitioners’ improvisations during work practices and how gained understanding of
adaptations (e.g. health programs summaries) by the implementation mediator was used to
inform the global development team. Second, the participatory design activities conducted
after post-implementation helped health workers to contrast the design of the software. Thus,
this research highlights the need within the PD community of developing countries to pay
attention to the practitioner’s acquired competency during the implementation and use of a
system by investigating their improvisations and innovations. For example, implementers can
explore what information artifacts are used in practice by the practitioners in particular
working conditions (e.g. related to the specific location of town, rural/urban) and use the
understanding to improve the design of IS (e.g. national HIS or open source software) in
developing countries. This implies that the practitioners’ adaptations can be used by design
facilitators in developing countries to enhance their participation in the design of the software.
7.4 Activity Theory Literature
The analysis of collaborative practices in this dissertation was done interpretively (Walsham,
2006) by using activity theory’s concepts of activity systems, mediation and contradictions.
Previous research recognizes that analysis centered on the concepts of structure of the activity
system and mediation offers a framework to understand how context impacts everyday work
practices (e.g. Leontev, 1981; Kaptelinin et al., 1995; Kaptelinin et al., 1999; Kaenampornpan
and O’Neil, 2005; Koffod-Petersen and Cassen, 2006; Huang and Gartiner, 2009; Bardram
and Doryab, 2011). Similarly, the use of these concepts from activity theory helped to identify
the impact of context in collaborative practices. In addition, my research contributes to
activity theory by extending the discussion around the impact of context in general and
specifically by showing that the influence of context can be investigated at the granular level
(of actions and operations) by looking at mediation breakdown arising from work-contextual
conditions.
My research adds to the activity theory literature by demonstrating the usefulness of the
disturbance concept in understanding the impact of contexts in work practices. Bannon and
Bødker (1989) have argued about the importance of looking at conditions which result in
mediation breakdown in the use situation. This study has focused on mediation breakdowns at
120
activity, action and operation level which provided an understanding of why and how
information artifacts and other work practices were adapted to fit the contextual condition.
For instance, defaulter tracing practice emerged to counterbalance influence from clients with
their need to fulfill their cultural or social-economic issues; e.g. the perceived greater need to
acquire income and psychological satisfaction than their physical wellbeing. Moreover,
patient registries became adapted in facilities located in town centers in order to provide the
overview of care during the search operation in identifying defaulters’ action. Therefore, my
research has shown that the impact of context happens not only in the elements of activity
system (as elaborated by previous researchers) but also at other levels within the hierarchy of
activity system.
Activity theorists have been criticized for limiting their discussion on the use of activity
theory to how context can be modeled into artifacts (Chen et al., 2013). In contrast, my
research provides insights on the use of activity theory to understand work practices around
the use of information tools and how gained insights can be used to inform the design of
primary healthcare information systems. Hence, my research also contributes to the discussion
on the use of activity theory to understand the impact of context on work practices rather than
only on the design of artifacts.
7.5 Design and Implementation Implications
This section present the design and implementation implications arising from the gained
understanding related to collaborative practices and also of the influence of the work-oriented
contextual condition in the settings of primary healthcare services.
7.5.1 Primary Healthcare Information Systems
The rich insights gained through the research of defaulter tracing have direct implications for
design of primary health information systems. The official patient registry in the PMTCT
services can be designed with the space to add community leaders’ information, including
their mobile phone numbers. Also, patient registries/forms for facility units can be created to
facilitate the creation of daily/weekly health program summaries. Additionally, the design of
the patient registry and other artifacts should allow an easy overview of the care trajectory of
clients and drop-out rates in order to facilitate defaulter tracing practices.
On a policy level, the institutionalizing of defaulter tracing practices in maternal and child
healthcare is important. For example, health authorities may develop defaulter tracing
121
guidelines and enforcing protocols (e.g. reminder practices for health workers). There is also a
need for providing monetary incentives and information artifacts (e.g. notebooks) to CHW
who are involved in tracing defaulters in the maternal and child healthcare services.
The rich empirical insights contribute to a better understanding of the contextual challenges
that primary healthcare information systems have to deal with. For example, each primary
health facility can be enabled to collect additional context-specific data in order to address the
work-context challenges, rather than having a uniform design of patient registries in all PHC
units. Moreover, the patient registries to support primary healthcare can be designed with a
general flexibility, for example, with extra field(s) to enable health workers to collect
additional context-specific data. Also, during the HIS training, trainers may also alert health
workers to the anticipated work-contextual challenges and how they should be prepared to
address them through improvisation of patient registries and their practices. Finally, the
digital system for primary healthcare services might be designed flexibly to allow health
providers to adapt them with little assistance from IT professionals.
The insights about the role of implementation mediators and the challenges they face can be
used to inform the design, implementation and development of distributed OSS software in
developing countries. The literature recognizes that the use of FOSS (Free and Open Source
Software) EPR within healthcare settings is viable both economically (e.g. it will eliminate
costs related to acquisition of software and license issues) and skills development (local
software development and implementation skills) (Seebregts et al., 2009; Omary et al., 2010).
Based on the findings of my research, I argue that it is also important to pay attention to the
role of the implementation mediators and their (participatory) practices within PHC facilities
so that the FOSS systems can be appropriately designed and implemented in order to support
the work practices of each health facilities.
This implies that practitioners’ adaptations can be used by implementation mediators in
developing countries to enhance practitioners’ participation in the design of the software.
Thus, this research highlights the need within the PD community of the developing countries
to pay attention to the practitioner-acquired competency during system use by investigating
their improvisations and innovations. For example, implementers can explore what
information artifacts are used in practice by the practitioners in particular working conditions
(e.g. related to the specific location of facility – rural/urban) and use the understanding to
improve the design of IS (e.g. national HIS or FOSS). This might require more capacity
122
building for implementers/IT facilitators (e.g. how to conduct ethnographic studies and
participatory design activities) for sustained use of computerized patient-care IS within the
PHC units.
7.5.2 DHIS Tracker
This study was conducted within the HISP project, which is involved in the design,
development, implementation, and capacity development activities of the standard DHIS
system. As part of my research, I was involved in the implementation of the DHIS Tracker
system in Tanzania and had the opportunity to interact with the global development team. The
importance of using the DHIS Tracker for monitoring care trajectories of pregnant women
and of children has been recognized within the HISP community (Gizaw et al., 2012; Ngoma
et al., 2012; Chawani, 2014a; Gizaw, 2014), but the understanding of the use of the DHIS
Tracker to support retention of clients within health programs is limited. My research
contributes to the HISP project by extending the understanding of the design and use of the
DHIS Tracker to support the retention of clients by demonstrating how the DHIS Tracker can
be used to facilitate defaulters’ tracing practices within health facilities. This is exemplified in
Paper IV, which describes how different functionalities within the DHIS Tracker (e.g. a
personal dashboard with different colors indicating clients’ attendance status, or the use of
SMS reminders to defaulting clients) can be used to facilitate defaulter tracing practices.
Furthermore, the research has provided design and implementation implications to support
defaulter tracing practices. First, I have shown that the design of automated SMS reminders to
clients who miss their scheduled appointments in Tanzania needs to be built on a
community/relational perspective (rather than individual one) to support existing defaulter
tracing practices in rural areas which depend on community health workers, community
leaders, neighbors and relatives.
Second, I have also provided the implementation implication related to how the DHIS Tracker
can be used to facilitate communication among health workers in different health programs
and facilities. The DHIS’ messaging component is recognized as an important tool for
collaboration among practitioners in monitoring the performance of the health system and
also to influence the design of generic DHIS system to accommodate the local needs of
countries (Shaw and Braa, 2014). However, within the HISP team, the understanding of the
use of the messaging component as a tool to facilitate collaboration is limited to facilitating
interactions between practitioners (health administrators), HISP facilitators, implementation
123
partners (e.g. international NGOs), and global software development teams. My study has
demonstrated that the DHIS Tracker’s communication facilities can also be used to facilitate
horizontal coordination: interactions between health providers in monitoring care trajectories.
For example, health workers can use mailing lists amongst themselves to send queries about
clients’ attendance in other facilities or health programs across districts even regions (Paper I).
Beyond the contribution related to the understanding of the use of the DHIS Tracker to
support the retention of clients, my research has contributed to the improvement of the generic
DHIS Tracker module by providing requirements to the global development team that
resulted in the inclusion of new features in the software (Paper IV).
7.6 Concluding Remarks
In this dissertation, I have presented my research with the aim of contributing to the HIS
literature of developing countries through improving the understanding of work practices in
primary healthcare settings. The aim of the dissertation was to increase the understanding of
HIS related to use, design and its implementation in order to facilitate health service provision
in maternal and child services around client retention. The study research had practical
motivation and both research contributions and practical implications have been proposed.
The main research questions introduced in Chapter 1 were addressed directly in Chapter 6.
The collaborative practices around retention of clients in the programs of immunization and
PMTCT services in Tanzania were the empirical focus of the study. The analysis in the thesis
has been done interpretively by using activity theory’s concepts of activity systems, mediation
and contradictions. The empirical findings were obtained mainly through multi-site
ethnographic studies conducted in multiple health programs, facilities and catchment
population located in different socio-geographical settings. Also, participatory design
approaches were used to inform the design and implementation of the software for monitoring
clients’ care trajectories. The thesis adopted an interpretive strategy to analyze empirical
findings. A key concept used to inform the analysis was a coordination mechanism, which
offered insights how PHC information systems can be designed to facilitate coordination of
patient trajectories within primary healthcare settings.
This thesis contributes to HIS and work practice literature. The thesis contributes to the
understanding of collaborative practices around facilitating the healthcare trajectory by
highlighting the involvement of different kinds of actors, both health and non-health
124
professionals. The study has also illustrated the role of work-oriented contextual conditions in
shaping their practices around the use of patient-oriented information artifacts. This
contextuality of collaborative work has implications for how work practices are coordinated
and, thereafter, for the design of primary healthcare information systems to support clients’
care trajectories. Activity theory was useful for analyzing patient-oriented information
artifacts, specifically by focusing on mediation breakdown in the situation of use, at activity,
action, and operation levels. By doing this study, the dissertation contributes to work practice
studies through illustrating how multiple actors are collaborating around care trajectories in
the resource-restricted settings. The thesis concludes by calling for more attention for the
context-sensitive design, implementation, and use of primary healthcare information systems
in Tanzania and other developing countries.
Beyond the discussions in this dissertation, further research may want to address how clientcentered information systems can be designed. The findings have shown that there could be
conflict between clients and healthcare activities and, consequently, clients could miss their
appointments. However, due to constraints arising from the PMTCT clinic which was the
main empirical data collection area, I was not able to obtain clients’ views on their care
trajectory. Hence, the findings of the dissertation are more oriented to the views of health
providers and managers obtained mainly through participant observation and interviews.
Accordingly, future research might be required to increase the understanding of HIS related to
how its design and implementation can incorporate clients’ views. For example, future
research can investigate how IT facilitators/designers can use the participatory design
approach to align different values (clients versus other actors) in the design and use of
technologies supporting defaulter tracing practices like the DHIS Tracker.
Further research is also needed to gain more understanding of mediators’ participatory
practices in larger-scale information systems (e.g. at national and regional levels) or within
open source software design and implementation activities or in different geographical and
cultural contexts of developing countries (e.g. Asian or Latin America countries).
125
References
Aanestad, M. (2003). The camera as an actor design-in-use of telemedicine infrastructure in surgery. Computer
Supported Cooperative Work (CSCW),12(1), 1-20.
Aanestad, M. and Olaussen, I. (2010). IKT og samhandling i helsesektoren. Digitale lappetepper eller sømløs
integrasjon?. Trondheim, Norge: Tapir Akademisk Forlag.
Abdusamadovich, L.M. (2013). Global Standards and Local Health Information System Applications:
Understanding their interplay in the context of Tajikistan. PhD, Faculty of Mathematics and Natural
Sciences, University of Oslo, Norway.
Abraham, J. and Reddy, M.C. (2010). Challenges to inter-departmental coordination of patient transfers: a
workflow perspective. International journal of medical informatics, 79(2), 112-122.
Abraham, J. and Reddy, M.C. (2008). Moving patients around: a field study of coordination between clinical and
non-clinical staff in hospitals. In Proceedings of the 2008 ACM conference on Computer supported
cooperative work (pp. 225-228). ACM.
Adaletey, D.L., Jolliffe, B., Braa, J., and Ofosu, A. (2015). Peer-Performance Review as a Strategy for
Strengthening Health In-formation Systems: A Case Study from Ghana. Journal of Health Informatics in
Africa, 2(2).
Allen, D., Karanasios, S., and Slavova, M. (2011). Working with activity theory: Context, technology, and
information behavior. Journal of the American Society for Information Science and Technology, 62(4), 776788.
Alter, S. (2015). A Workaround Design System for Anticipating, Designing, and/or Preventing Workarounds.
In Enterprise, Business-Process and Information Systems Modeling (pp. 489-498). Springer International
Publishing.
Amir, O., Grosz, B.J., Gajos, K.Z., Swenson, S. M., and Sanders, L. M. (2015). From Care Plans to Care
Coordination: Opportunities for Computer Support of Teamwork in Complex Healthcare. Computer Human
Interaction (CHI) 2015, Seoul, Republic of Korea, April 18–23.
Asangansi, I.E. (2013). Exploring the Institutional Logics and Complexity of Health Management Information
System Implementation. PhD, Department of Informatics, University of Oslo, Norway.
Ash, J. S., Berg, M., and Coiera, E. (2004). Some unintended consequences of information technology in health
care: the nature of patient care information system-related errors. Journal of the American Medical
Informatics Association, 11(2), 104-112.
Assefa, Y., Damme, W. V., Mariam, D. H., and Kloos, H. (2010). Toward universal access to HIV counseling
and testing and antiretroviral treatment in Ethiopia: looking beyond HIV testing and ART initiation. AIDS
patient care and STDs, 24(8), 521-525.
Avgerou, C. (2010). Discourses on ICT and Development. Information Technologies and International
Development, 6(3), pp-1.
Avilés, W., Ortega, O., Kuan, G., Coloma, J., and Harris, E. (2007). Integration of information technologies in
clinical studies in Nicaragua. PLoS Med, 4(10), 1578 - 1583.
Bagayoko, C. O., Anne, A., Geissbuhler, A., and Fieschi, M. (2011). Can ICTs Contribute to the Efficiency and
provide Equitable Access to the Health Care System in Sub-Saharan Africa?. IMIA Yearbook, 6, 33-38.
Baldauf, M., Dustdar, S., and Rosenberg, F. (2007). A survey on context-aware systems. International Journal of
Ad Hoc and Ubiquitous Computing, 2(4), 263-277.
Balka, E., Bjorn, P., and Wagner, I. (2008). Steps toward a typology for health informatics. In Proceedings of the
2008 ACM conference on Computer supported cooperative work (pp. 515-524). ACM.
Bandon, D., Lovis, C., Geissbühler, A., and Vallée, J. P. (2005). Enterprise-wide PACS: Beyond Radiology, an
Architecture to Manage All Medical Images 1. Academic Radiology, 12(8), 1000-1009.
Bannon, L.J. and Bødker, S. (1989). Beyond the interface: Encountering artifacts in use. DAIMI Report
Series, 18(288).
Bansler, J. P., Havn, E., Mønsted, T., and Schmidt, K. (2011). A study of the fragmentation of the medical
record. P. Bjørn, et al, 94-97.
Bansler, J., Havn, E., Mønsted, T., Schmidt, K., and Svendsen, J. H. (2013). Physicians’ progress notes.
In ECSCW 2013: Proceedings of the 13th European Conference on Computer Supported Cooperative Work,
21-25 September 2013, Paphos, Cyprus (pp. 123-142). Springer London.
126
Barab, S.A., Barnett, M., Yamagata-Lynch, L., Squire, K., and Keating, T. (2002). Using activity theory to
understand the systemic tensions characterizing a technology-rich introductory astronomy course. Mind,
Culture, and Activity, 9(2), 76-107.
Barcellini, F., Prost, L., and Cerf, M. (2015). Designers' and users' roles in participatory design: What is actually
co-designed by participants?. Applied ergonomics, 50, 31-40.
Bardram J. and Doryab A. (2011). Activity analysis: applying activity theory to analyze complex work in
hospitals. In Proceedings of the ACM 2011 conference on Computer supported cooperative work (pp. 455464). ACM.
Bardram, J.E. (1998). “Designing for the Dynamics of Cooperative Work Activities”, Poltrock and Grudin (red):
Proceedings of the Conference on Computer-Supported Cooperative Work, CSCW'98, ACM, pp. 89-98.
Bardram, J.E. (2000). Temporal Coordination–On Time and Coordination of CollaborativeActivities at a
Surgical Department. Computer Supported Cooperative Work (CSCW), 9(2), 157-187.
Bardram, J.E. and Bossen, C. (2005). A web of coordinative artifacts: collaborative work at a hospital ward.
In Proceedings of the 2005 international ACM SIGGROUP conference on Supporting group work (pp. 168176). ACM.
Bath, G.F. (2006). Telemedicine in Africa. Journal of the South African Veterinary Association, 77(3), 107.
Bennett, D. E., Jordan, M. R., Bertagnolio, S., Hong, S. Y., Ravasi, G., McMahon, J. H., ... and Kelley, K.F.
(2012). HIV drug resistance early warning indicators in cohorts of individuals starting antiretroviral therapy
between 2004 and 2009: World Health Organization global report from 50 countries. Clinical infectious
diseases, 54(suppl 4), S280-S289.
Berg, B. and Lune, H. (2012). Qualitative Research Methods for Social Sciences. 8th edition. Harlew Essex;
Pearson.
Berg, M. (1996). Practices of reading and writing: the constitutive role of the patient record in medical
work. Sociology of health and illness, 18(4), 499-524.
Berg, M. (1997). On distribution, drift and the electronic medical record. In Proceedings of the fifth European
conference on computer supported cooperative work (pp. 141-156). Springer Netherlands.
Berg, M. (1999a). Accumulating and coordinating: occasions for information technologies in medical
work. Computer Supported Cooperative Work (CSCW),8(4), 373-401.
Berg, M. (1999b). Patient care information systems and health care work: a sociotechnical
approach. International journal of medical informatics, 55(2), 87-101.
Berg, M. and Goorman, E. (1999). The contextual nature of medical information. International journal of
medical informatics, 56(1), 51-60.
Beynon-Davies, P. (1997). Ethnography and information systems development: Ethnography of, for and within
is development. Information and software technology, 39(8), 531-540.
Bjerknes, G. and Bratteteig, T. (1995). User participation and democracy: a discussion of Scandinavian research
on system development. Scandinavian Journal of information systems, 7(1), 1.
Bjørn, P. and Hertzum, M. (2011). Artefactual multiplicity: A study of emergency-department whiteboards.
Computer Supported Cooperative Work, 20(1-2), 93-121.
Blackler, F., Crump, N., and McDonald, S. (2003). Organizing processes in complex activity networks. Knowing
in organizations: A practice-based approach. ME Sharpe.
Blin, F. and Munro, M. (2008). Why hasn’t technology disrupted academics’ teaching practices? Understanding
resistance to change through the lens of activity theory. Computers and Education, 50(2), 475-490.
Blomberg, J. and Karasti, H. (2012). Ethnography: Positioning Ethnography within Participatory Design.
Simonsen, J. and Robertson, T. (eds.) Routledge International Handbook of Participatory design Routledge
New York, 86–116.
Blomberg, J L., Giacomi, J., Mosher, A., and Swenton-Hall, P. (1993). Ethnographic Field Methods and Their
Relation to Design. In Participatory Design: Principles and Practices Hillsdale, N.J.: Lawrence Erlbaum
Associates, , 123–56.
Blomberg, J. and Karasti, H. (2013). Reflections on 25 years of ethnography in CSCW. Computer Supported
Cooperative Work, 22(4-6), 373-423.
Blomberg, J., Suchman, L., and Trigg, R. H. (1996). Reflections on a work-oriented design project. HumanComputer Interaction, 11(3), 237-265.
127
Bodart, C., and Shrestha, L. (2000). Identifying information needs and indicators. Design and implementation of
health information systems. Geneva: WHO.
Bødker, K., Kensing, F. and Simonsen, J. (2004). Participatory IT design: designing for business and workplace
realities. MIT press, USA.
Bødker, S. (1996). Applying activity theory to video analysis: how to make sense of video data in humancomputer interaction. In Context and consciousness (pp. 147-174). Massachusetts Institute of Technology.
Bødker, S., Ehn, P., Knudsen, J., Kyng, M., and Madsen, K. (1988). Computer support for cooperative design.
In Proceedings of the 1988 ACM conference on Computer-supported cooperative work (pp. 377-394). ACM.
Boer, N. I., van Baalen, P. J., and Kumar, K. (2002). An activity theory approach for studying the situatedness of
knowledge sharing. In System Sciences, 2002. HICSS. Proceedings of the 35th Annual Hawaii International
Conference on (pp. 1483-1492). IEEE.
Bossen, C., Groth Jensen, L., and Witt, F. (2012). Medical secretaries' care of records: the cooperative work of a
non-clinical group. In Proceedings of the ACM 2012 conference on Computer Supported Cooperative
Work (pp. 921-930). ACM.
Bowers, J., Button, G., and Sharrock, W. (1995). “Workflow from Within and Without: Technology and
Cooperative Work on the Print Industry Shopfloor” Marmolin m.fl. (red): Proceedings of the Fourth
European Conference on Computer-Supported Cooperative Work, ECSCW'95, Kluwer, pp. 51-66.
Braa, J. and Sahay S. Health Information Systems Programme: Participatory Design within the HISP network.
Simonsen, J. and Robertson, T. (eds.) Routledge International Handbook of Participatory Design. Routledge,
New York, 2012, 235-56.
Braa, J., Monteiro, E. and Sahay, S. (2004). Networks of Action: Sustainability Health Information Systems
across Developing Countries. MIS Quarterly, ( 28:28), pp. 337-362.
Bratteteig, T., Bødker, K., Dittrich, Y., Mogensen, P. H., and Simonsen, J. (2012). Methods: organising
principles and general guidelines for Participatory Design projects. Simonsen, J. and Robertson, T. (eds.)
Routledge International Handbook of Participatory Design. Routledge, New York, 2012,117-44.
Brown, J.S. and Duguid, P. (1991). Organizational learning and communities-of-practice: Toward a unified view
of working, learning, and innovation. Organization science, 2(1), 40-57.
Bryman, A. (2008) Social research methods, 3rd Ed., Oxford University Press New York.
Busza J., Damilola W., Hairston, A., Gable, A., Pitter, C., Lee, S., Katirayi L., Simiyu R., Mpofu, D. (2012)
Community-based approaches for prevention of mother to child transmission in resource-poor settings: a
social
ecological
review.
Journal
of
the
International
AIDS
Society,
15,
2:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3499910/ .
Byrne, E. and Sahay, S. (2007). Participatory design for social development: A South African case study on
community-based health information systems. Information Technology for Development, 13(1), 71-94.
Cabitza, F., Locatelli, M. P.,and Simone, C. (2009). Promoting Process-Based Collaboration Awareness to
Integrate Care Teams. In Business Process Management Workshops (pp. 385-396). Springer Berlin
Heidelberg.
Cabitza, F., Sarini, M., Simone, C., and Telaro, M. (2006). Torres, a conceptual framework for articulation work
across boundaries. Frontiers in Artificial Intelligence and Applications, 137, 102.
Carroll, J. (2004). Completing design in use: closing the appropriation cycle. European Conference on
Information Systems, Finland, 44.
Carstensen, P. H., and Sørensen, C. (1996). From the social to the systematic. Computer Supported Cooperative
Work, 5(4), 387-413.
Cassens, J., and Kofod-Petersen, A. (2006). Using Activity Theory to Model Context Awareness: A Qualitative
Case Study. In FLAIRS Conference (pp. 619-624).
Chadambuka, A., Chimusoro, T.A., Tshimanga, M., Namusisi, O. and Luman E.T. (2012). The need for
innovative strategies to improve immunization services in rural Zimbabwe. Disasters, 36, 1, 161 -173.
Chawani, M.S. (2014a). A Cross-case Analysis of the Effects of EMR Deployment on Antenatal Care Services
in Rural Health Centres in Malawi. Journal of Health Informatics in Africa, 2(1).
Chawani M.S. (2014b). Development of Electronic Medical Record Systems for maternal health services in rural
settings an action research study from Malawi. PhD, Department of Informatics, University of Oslo, Norway.
128
Chawani, M. S., Kaasbøll, J., Finken, S., and Herstad, J. (2014a). Balancing Work Practices and Protocols in the
Design of EMR Systems: The Case of Developing an EMR System for Antenatal Care Services in
Malawi. The Electronic Journal of Information Systems in Developing Countries, 63.
Chawani, M.S., Kaasbøll, J., and Finken, S. (2014b). Stakeholder participation in the development of an
electronic medical record system in Malawi. In Proceedings of the 13th Participatory Design Conference:
Research Papers-Volume 1 (pp. 71-80). ACM.
Chen, Y., Liu, Z., and Vainio, J. (2013). Activity-based context-aware model. In Design, User Experience, and
Usability. Design Philosophy, Methods, and Tools (pp. 479-487). Springer Berlin Heidelberg.
Cheng, K.G., Ernesto, F., and Truong, K.N. (2008). Participant and interviewer attitudes toward handheld
computers in the context of HIV/AIDS programs in sub-Saharan Africa. In Proceedings of the SIGCHI
Conference on Human Factors in Computing Systems (pp. 763-766). ACM.
Chilundo,B., Sundby, J., and Aanestad, M. (2004). Analysing the quality of routine malaria data in
Mozambique. Malaria Journal, 3(1), 3.
Cimino, J. J. (2013). Improving the electronic health record—are clinicians getting what they wished
for?. Journal of American Medical Association, 309(10), 991-992.
Ciolfi, L., Fitzpatrick, G., and Bannon, L. (2008). Settings for collaboration: The role of place. Computer
Supported Cooperative Work, 17(2-3), 91-96.
Cluts, M.M. (2003). The evolution of artifacts in cooperative work: constructing meaning through activity.
In Proceedings of the 2003 international ACM SIGGROUP conference on Supporting group work (pp. 144152). ACM.
Cole, M. (1995). Socio-cultural-historical psychology: some general remarks and a proposal for a new kind of
cultural-genetic methodology. Sociocultural studies of mind, 187-214.
Cutts, F.T. (1991). Strategies to improve immunization services in urban Africa. Bulletin of the World Health
Organization, 69, 4, 407- 414: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2393244.
Dalsted, R. (2012). Patient trajectory as idea and practice. A sociological analysis of cancer treatment and care
in Denmark. PhD, Faculty of Health and Medical Sciences University of Copenhagen, Denmark.
Damtew, Z.A., and Moges, A.S. (2013). From Multiple Register to Family folder: The Transition of Data
Collection and Reporting Tools for Health Extension Workers in Ethiopia. Journal of Health Informatics in
Developing Countries, 7(2).
Darcy, N., Elias, M., Swai, A., Danford, H., Rulagirwa, H., and Perera, S. (2015). eHealth strategy development:
a case study in Tanzania. Journal of Health Informatics in Africa, 2(2).
DeRenzi, B., Borriello, G., Jackson, J., Kumar, V. S., Parikh, T. S., Virk, P., and Lesh, N. (2011). Mobile Phone
Tools for Field Based Health care Workers in Low-Income Countries. Mount Sinai Journal of Medicine: A
Journal of Translational and Personalized Medicine, 78(3), 406-418.
DeRenzi, B., Findlater, L., Payne, J., Birnbaum, B., Mangilima, J., Parikh, T., ... and Lesh, N. (2012). Improving
community health worker performance through automated SMS. In Proceedings of the Fifth International
Conference on Information and Communication Technologies and Development (pp. 25-34). ACM.
DeRenzi, B., Lesh, N., Parikh, T., Sims, C., Maokla, W., Chemba, M., ...andBorriello, G. (2008). E-IMCI:
Improving pediatric health care in low-income countries. In Proceedings of the SIGCHI conference on
human factors in computing systems, 753-762, ACM.
Dey, A.K., Abowd, G. D., and Salber, D. (2001). A conceptual framework and a toolkit for supporting the rapid
prototyping of context-aware applications. Human-computer interaction, 16(2), 97-166.
Dittrich, Y., Eriksén, S., and Wessels, B. (2014). Learning through Situated Innovation: Why the specific is
crucial for Participatory Design Research. Scandinavian Journal of Information Systems, 26(1), 29-56.
Dittrich, Y., Randall, D. W., and Singer, J. (2009). Software engineering as cooperative work. Computer
Supported Cooperative Work, 18(5-6), 393-399.
DMO (2013a). Bahari district council; comprehensive council health plan: 2013 – June 2014. Bahari: The
United Republic of Tanzania.
DMO (2013b). Bwawa district council; comprehensive council health plan: 2013 – June 2014. Bwawa: The
United Republic of Tanzania.
Doryab A. and Bardram J. (2011). Context- and Activity-awareness in Collaborative Environments.
In Proceedings of the 2011 international workshop on Situation activity and goal awareness (pp. 57-66).
ACM.
129
Dourish, P. and Button, G. (1998). On" technomethodology": Foundational relationships between
ethnomethodology and system design. Human-computer Interaction, 13(4), 395-432.
Dourish, P., Holmes, J., MacLean, A., Marqvardsen, P., and Zbyslaw, A. (1996). Freeflow: mediating between
representation and action in workflow systems. In Proceedings of the 1996 ACM conference on Computer
supported cooperative work (pp. 190-198). ACM.
Dwolatzky, B., Trengove, E., Struthers, H., McIntyre, J. A., and Martinson, N. A. (2006). Linking the global
positioning system (GPS) to a personal digital assistant (PDA) to support tuberculosis control in South
Africa: a pilot study. International Journal of Health Geographics, 5(1), 34.
Ehn, P. Participation in Design Things. Proc. PDC 2008, 92-101.
Elovaara, P., Igira, F. T., and Mörtberg, C. (2006). Whose participation? Whose knowledge?: exploring PD in
Tanzania-Zanzibar and Sweden. Proc. PDC 2006, 105-14.
Emerson, R.M., Fretz, R. I., and Shaw, L. L. (1995). Writing ethnographic fieldnotes. University of Chicago
Press.
Engeström, Y. (1987). Learning by expanding: An activity theoretical approach to developmental research.
Orienta-Konsultit Oy ,Helsinki, Finland.
Engestrom, Y. (2000). Activity theory as a framework for analyzing and redesigning work. Ergonomics, 43(7),
960-974.
Engeström, Y. (2001). Expansive Learning at Work: Toward an activity theoretical reconceptualization. Journal
of Education and Work, 14, 1, 133-156.
Engeström, Y. and Escalante, V. (1996). Mundane tool or object of affection? The rise and fall of the postal
buddy. Context and consciousness: Activity theory and human-computer interaction, 325-374.
Engeström, Y. and Mazzocco, D. W. (1995). Disturbance management and masking in a television production
team: An activity-theoretical study in organizational communication. In 45th Annual Conference of the
International Communication Association, Organizational Communication Division.
Engeström, Y., Puonti, A., and Seppänen, L. (2003). Spatial and temporal expansion of the object as a challenge
for reorganizing work. Knowing in organizations: A practice-based approach, 151-186.
Falzon, M. A. (Ed.). (2012). Multi-sited ethnography: theory, praxis and locality in contemporary research.
Ashgate Publishing, Ltd.
Farah, S.K.S. (2010). Challenges and Opportunities of Integrating Primary Health Care Information System:
Northern State, Sudan. Master Thesis, Faculty of Medicine, University of Oslo, Norway.
Faus, S. A. and Sujansky, W. (2008). Open-Source EHR Systems for Ambulatory Care: A Market Assessment.
California HealthCare Foundation.
Fitzpatrick, G. (2004) Integrated care and the working record. Health Informatics Journal, 10(4), 291-302.
Fitzpatrick, G. and Ellingsen, G. (2013). A review of 25 years of CSCW research in healthcare: Contributions,
challenges and future agendas. Computer Supported Cooperative Work, 22(4-6), 609-665.
Folcher, V. (2003). Appropriating artifacts as instruments: when design-for-use meets design-in-use. Interacting
with computers, 15(5), 647-663.
Forster, M., Bailey, C., Brinkhof M.W., Graber C., Boulle, A., Spohr, M., Balestre, E., May, M., Keiser, O.,
Jahn, A. and Egger, M. (2008) Electronic medical record systems, data quality and loss to follow-up: survey
of antiretroviral therapy programmes in resource-limited settings. Bulletin of the World Health Organization,
86, 12, 939-947: http://www.ncbi.nlm.nih.gov/pubmed/19142294.
Fraser, H.SF., Allen., C., Bailey, C., Douglas, G., Shin, S., Blaya, J. (2007). Information systems for patient
follow-up and chronic management of HIV and Tuberculosis: A life-saving Technology in resource poor
settings, (2007), J Med Internet Res, 9(4):e29
Garfinkel, H. and E. Bittner (1967). Good Organizational Reasons for “bad” Clinic Records. In Studies in
Ethnomethodology. H. Garfinkel, ed. Pp. 186–207. Englewood Cliffs: Prentice-Hall
Gasser, L. (1986). The integration of computing and routine work. ACM Transactions on Information Systems
(TOIS), 4(3), 205-225.
Geissbuhler, A., Ly, O., Lovis, C. and L’Haire, J.F. (2003). Telemedicine in Western Africa: lessons learned
from a pilot project in Mali, perspectives and recommendations. In AMIA Annual Symposium
Proceedings (Vol. 2003, p.249). American Medical Informatics Association.
Gerson, E.M. and Star, S. L. (1986). Analyzing due process in the workplace. ACM Transactions on Information
Systems (TOIS), 4(3), 257-270.
130
Ghadrshenasa, A., Amor, Y.B., Chang, J., Dale, H., Sherman, G., Vojnova, L., Young, P. and Yoge, R. (2013).
Improved access to early infant diagnosis is a critical part of a child-centric prevention of mother-to child
transmission agenda. AIDS, 27, 2, 197–220.
Gherardi, S. (2009). “Introduction: The Critical Power of the ‘Practice Lens’”, Management Learning Journal,
40(2), pp. 115-128.
Gherardi, S. (2012). How to conduct a practice-based study: Problems and methods. Edward Elgar Publishing.
Gilson, L., Alilio, M., and Heggenhougen, K. (1994). Community satisfaction with primary health care services:
an evaluation undertaken in the Morogoro region of Tanzania. Social science and medicine, 39(6), 767-780.
Gizaw A.A., Mukherjee, A., Lewis J. and Sahay, S. (2012). Understanding design-use interaction: the case of
name based health information system in India, In Proceedings of IADIS International Conference on
Information Systems.
Gizaw, A. A. (2014). Open Generification, PhD, PhD, Departments of Informatics, University of Oslo, Norway.
Gregory, J. (2003). Scandinavian approaches to Participatory design. International Journal of Engineering
Education, 19(1), 62-74.
Grinter, R.E. (2003). Recomposition: Coordinating a web of software dependencies. Computer Supported
Cooperative Work, 12(3), 297-327.
Grisot, M. (2008). Foregrounding Differences: A performative approach to the coordination of distributed work
and information infrastructures in use. PhD, Departments of Informatics, University of Oslo, Norway.
Grudin, J. (1994). Computer-supported cooperative work: History and focus. Computer, (5), 19-26.
Haberer, J. E., Kiwanuka, J., Nansera, D., Wilson, I. B., and Bangsberg, D. R. (2010). Challenges in using
mobile phones for collection of antiretroviral therapy adherence data in a resource-limited setting. AIDS and
Behavior, 14(6), 1294-1301.
Hage, G. (2005). A not so multi-sited ethnography of a not so imagined community. Anthropological
theory, 5(4), 463-475.
Hammersley, M. and Atkinson, P. (2007). Ethnography: Principles in practice. Routledge.
Harrison, M.I., Koppel, R., and Bar-Lev, S. (2007). Unintended consequences of information technologies in
health care—an interactive sociotechnical analysis. Journal of the American medical informatics
Association, 14(5), 542-549.
Harrison, S. and Dourish, P. (1996). Re-place-ing space: the roles of place and space in collaborative systems.
In Proceedings of the 1996 ACM conference on Computer supported cooperative work (pp. 67-76). ACM.
Hartswood, M., Procter, R., Rouncefield, M., and Slack, R. (2003). Making a case in medical work: implications
for the electronic medical record. Computer Supported Cooperative Work, 12(3), 241-266.
Hartswood, M., Procter, R., Slack, R., Vob, A., Buscher, M., Rouncefield, M., and Rouchy, P. (2002). Corealisation: Towards a principled synthesis of ethnomethodology and participatory design. Scandinavian
Journal of Information Systems, 14(2).
Hasu, M. (2001). Critical transition from developers to users: Activity-theoretical studies of interaction and
learning in the innovation process. PhD, Faculty of Education, University of Helsinki, Finiland.
Häyrinen, K., Saranto, K., and Nykänen, P. (2008). Definition, structure, content, use and impacts of electronic
health records: a review of the research literature. International journal of medical informatics, 77(5), 291304.
Heath, C. and Luff, P. (1996). Documents and professional practice:“bad” organisational reasons for “good”
clinical records. In Proceedings of the 1996 ACM conference on Computer supported cooperative work (pp.
354-363). ACM.
Heath, C., Svensson, M.S., Hindmarsh, J., Luff, P., and Vom Lehn, D. (2002). Configuring awareness. Computer
Supported Cooperative Work, 11(3-4), 317-34
Heeks, R. (2002). Information systems and developing countries: Failure, success and local improvisations. The
Information Society, 18(2), 101-112.
Heeks, R. (2006). Health information systems: Failure, success and improvisation. International journal of
medical informatics, 75(2), 125-137.
Heeks, R. (2008). ICT4D 2.0: The Next Phase of Applying ICT for International Development. Computer, 41(6),
26–33.
Helle, M. (2000). Disturbances and contradictions as tools for understanding work in the
newsroom. Scandinavian Journal of Information Systems, 12(1), 7.
131
Henderson, A. and Kyng, M. (1991). There’s no place like home: continuing design in use. Greenbaum, J. and
Kyng, M. (eds.) Design at Work: Cooperative Design of Computer Systems. Lawrence Erlbaum Associates,
New Jersey, USA., 219-40.
Herbsleb, J.D., Mockus, A., Finholt, T.A., and Grinter, R.E. (2000). Distance, dependencies, and delay in a
global collaboration. In Proceedings of the 2000 ACM conference on Computer supported cooperative
work (pp. 319-328). ACM.
Heunis, C., Wouters, E., Kigozi, G., Engelbrecht, M., Tsibolane, Y., van der Merwe, S., and Motlhanke, S.
(2011). Accuracy of tuberculosis routine data and nurses’ views of the TB-HIV information system in the
Free State, South Africa. Journal of the Association of Nurses in AIDS Care, 22(1), 67-73.
Heywood, A. and Rohde, J. (2001). Using information for action: A manual for health workers at facility level
South Africa: EQUITY Project.
Hindemark, F. (2013). E-Health at Outpatient Clinics in Uganda. Master thesis, Royal Institute of Technology,
Sweden.
Hinds, P.J. and Mortensen, M. (2005). Understanding conflict in geographically distributed teams: The
moderating effects of shared identity, shared context, and spontaneous communication. Organization
science, 16(3), 290-307.
Hine, C. (2007). Multi-sited ethnography as a middle range methodology for contemporary STS. Science,
Technology and Human Values, 32(6), 652-671.
Hoffman, J. A., Cunningham, J. R., Suleh, A. J., Sundsmo, A., Dekker, D., Vago, F., ... and Hunt-Glassman, J.
(2010). Mobile direct observation treatment for tuberculosis patients: a technical feasibility pilot using mobile
phones in Nairobi, Kenya. American journal of preventive medicine, 39(1), 78-80.
Huang, H. and Gartner, G. (2009). Using activity theory to identify relevant context parameters. In Location
Based Services and TeleCartography II (pp. 35-45). Springer Berlin Heidelberg.
Hughes, J., King, V., Rodden, T., and Andersen, H. (1994). Moving out from the control room: Ethnography in
system design. In Proceedings of the 1994 ACM conference on Computer supported cooperative work (pp.
429-439). ACM.
Huizing, A. and Cavanagh, M. (2011). Planting contemporary practice theory in the garden of information
science. Information Research Journal, 16(4), 497.
Husein, K., Adeyi, O., Bryant, J., and Cara, N.B. (1993). Developing a primary health care management
information system that supports the pursuit of equity, effectiveness and affordability. Social Science and
Medicine, 36(5), 585-596.
Hyysalo, S. and Lehenkari, J. (2000). Defining the crucial users-An activity theoretical for finding the missing
voices. In PDC (pp. 283-286).
Igira, F. (2012). Information Systems Deployment as an Activity System. In Information Systems Theory (pp.
397-411). Springer New York.
Igira, F.T. (2008a). The situatedness of work practices and organizational culture: implications for information
systems innovation uptake. Journal of Information Technology, 23(2), 79-88.
Igira, F.T. (2008b). The interplay between transformation in everyday work practices and IS design and
implementation processes: Empirical experiences from the health information system in Tanzania. PhD,
Department of Informatics, University of Oslo, Norway.
Igira, F. and Aanestad, M. (2009). Living with contradictions: Complementing activity theory with the notion of
“installed base” to address the historical dimension of transformation. Mind, Culture, and Activity, 16(3),
209-233.
Iivari, N. (2011). Participatory design in OSS development: interpretive case studies in company and community
OSS development contexts. Behaviour and Information Technology, 30(3), 309-323.
Ijadunola, K. T., Bello, M. O., Korpela, M., and Koskivirta, M. (1998). A strategy for the use of information
systems to support community-oriented primary healthcare in a deprived economy: A case study of Ife
Central LGA, Nigeria. Current Perspectives in Healthcare Computing, 139-145.
Issroff, K. and Scanlon, E. (2002). Using technology in higher education: activity theory perspective. Journal of
Computer assisted Learning, 18, 1, 77- 83.
Jarzabkowski, P. (2005). Strategy as practice: An activity based approach. Sage.
Jirotka, M., Lee, C. P., and Olson, G. M. (2013). Supporting scientific collaboration: Methods, tools and
concepts. Computer Supported Cooperative Work, 22(4-6), 667-715.
132
Johansson, M. and Linde, P. (2005). Playful collaborative exploration: New research practice in participatory
design. Journal of Research Practice, 1(1), 5.
Jones, M. (2003). “Computers can land people on Mars, why can't they get them to work in a hospital?"
Implementation of an Electronic Patient Record System in a UK Hospital. Methods of information in
medicine, 42(4), 410-415.
Kaasbøll, J., Chawani, M.S., Hamre, G.A, and Sandvand, J. (2010). Competencies and Learning for
Management Information Systems. Journal of Information, Information Technology and Organizations, 5,
85-100.
Kabuya, C., Wright, G., Odama, A., and O'Mahoney, D. (2014). Routine Data for Disease Surveillance in the
Undeveloped Region of the OR Tambo District of the Eastern Cape Province. In Cross-Border Challenges in
Informatics with a Focus on Disease Surveillance and Utilising Big Data: Proceedings of the EFMI Special
Topic Conference, 27-29 April 2014, Budapest, Hungary (Vol. 197, p. 103). IOS Press.
Kaenampornpan, M. and O’Neill, E. (2004). An integrated context model: Bringing activity to context. In Proc.
Workshop on Advanced Context Modelling, Reasoning and Management.
Kane, B. and Luz, S. (2008). Taking lessons from teleconference to improve same time, same place interaction.
In Computer-Based Medical Systems, 2008. CBMS'08. 21st IEEE International Symposium on (pp. 494-499).
IEEE.
Kanjo, C. (2011). Pragmatism or Policy: Implications on Health Information Systems Success. The Electronic
Journal of Information Systems in Developing Countries, 48.
Kanjo, C. (2012). In search of the missing data. The case of maternal and child health data in Malawi. PhD,
Faculty of Mathematics and Natural Sciences, University of Oslo, Norway.
Kaplan, B. and Duchon, D. (1988). Combining qualitative and quantitative methods in information systems
research: a case study. MIS quarterly, 571-586.
Kaplan, W.A. (2006). Can the ubiquitous power of mobile phones be used to improve health outcomes in
developing countries. Global Health, 2(9), 1-14.
Kaptelinin, V. and Nardi, B. A. (2006). Acting with Technology: Activity Theory and Interaction Design, MIT
Press.
Kaptelinin, V. and Nardi, B. (2012). Activity Theory in HCI: Fundamentals and Reflections. Morgan and
Claypool Publishers.
Kaptelinin, V., Kuutti, K., and Bannon, L. (1995). Activity theory: Basic concepts and applications. In Humancomputer interaction (pp. 189-201). Springer Berlin Heidelberg.
Kaptelinin, V., Nardi, B. A., and Macaulay, C. (1999). Methods and tools: The activity checklist: a tool for
representing the “space” of context. interactions, 6(4), 27-39.
Karanasios, S. (2014). Framing ICT4D research using activity theory: a match between the ICT4D field and
theory?. Information Technologies and International Development, 10(2), pp-1.
Karara, G., Verbeke, F., and Nyssen, M. (2013). Hospital Information Management Using Open Source
Software: Results of the MIDA Project in 3 Hospitals in Rwanda. Journal of Health Informatics in
Africa, 1(1).
Karasti, H. (2001a). Increasing sensitivity towards everyday work practice in system design. PhD. Department of
Information Processing Science, University of Oulu, Oulu, Finland.
Karasti, H. (2001b). Bridging work practice and System Design: Integrating Systemic Analysis, Appreciative
Intervention and Practitioner participation. Computer Supported Cooperative Work, 10(2), 211-246.
Karasti, H. (2010). Taking PD to Multiple Contexts. Scandinavian Journal of Information Systems, 22(1), 85-92.
Kensing, F. and Blomberg, J. (1998). Participatory design: Issues and concerns. Computer Supported
Cooperative Work, 7(3-4), 167-185.
Kensing, F. and Greenbaum, J. (2012). Heritage: having a say. Simonsen, J. and Robertson, T. (eds.) Routledge
International Handbook of Participatory Design. Routledge, New York, 2012, 21-36.
Kensing, F., and Madsen, K. H. (1992). Generating visions: Future workshops and metaphorical design (pp.
155-168). L. Erlbaum Associates Inc..
Kimaro, H.C. (2006). Strategies for developing human resource capacity to support sustainability of ICT based
health information systems: a case study from Tanzania. The Electronic Journal of Information Systems in
Developing Countries, 26.
133
Kimaro, H.C. and Titlestad, O.H. (2008). Challenges of user participation in the design of a computer based
system: The possibility of participatory customisation in low income countries. Journal of Health informatics
in developing countries, 2(1).
Kofod-Petersen, A. and Cassens, J. (2006). Using activity theory to model context awareness. In Modeling and
Retrieval of Context (pp. 1-17). Springer Berlin Heidelberg.
Koppel, R. and Lehmann, C.U. (2014). Implications of an emerging EHR monoculture for hospitals and
healthcare systems. Journal of the American Medical Informatics Association, amiajnl-2014.
Korpela, M., De La Harpe, R., and Luukkonen, I. (2008). Depicting the landscape around information flows:
Methodological propositions. GlobDev 2008, 5.
Korpela, M., Mursu, A. and Soriyan, H.A. (2002). Information systems development as an activity. Computer
Supported Cooperative Work, 11(1-2), 111-128.
Korpela, M., Soriyan, H.A., and Olufokunbi, K.C. (2001). Activity analysis as a method for information systems
development. Scandinavian Journal of Information Systems, 12(1-2), 191-210.
Kossi, E.K., Sæbø, J.I., Braa, J., Jalloh, M.M., and Manya, A. (2012). Developing decentralised health
information systems in developing countries–cases from Sierra Leone and Kenya. The Journal of Community
Informatics, 9(2).
Kovner, C., Schuchman, L., and Mallard, C. (1996). The application of pen-based computer technology to home
health care. Computers in nursing, 15(5), 237-244.
Kuutti, K. (1996). Activity theory as a potential framework for human-computer interaction research. Context
and consciousness: Activity theory and human-computer interaction, 17-44.
Kuutti, K. (2013). ‘Practice turn’and CSCW identity. ECSCW 2013 Adjunct Proceedings.
Kwesigabo, G., Mwangu, M.A., Kakoko, D.C., Warriner, I., Mkony, C.A., Killewo, J., ... and Freeman, P.
(2012). Tanzania's health system and workforce crisis. Journal of public health policy, S35-S44.
Kyng, M. (1991). Designing for cooperation: cooperating in design. Communications of the ACM, 34(12), 65-73.
Law, J. (1992). Notes on the theory of the actor-network: Ordering, strategy, and heterogeneity. Systems
practice, 5(4), 379-393.
Layer, E. H., Kennedy, C. E., Beckham, S. W., Mbwambo, J. K., Likindikoki, S., Davis, W. W., ... and LTC
Tanzania Collaborative Study Team. (2014). Multi-level factors affecting entry into and engagement in the
HIV continuum of care in Iringa, Tanzania.
Ledikwe, J.H., Grignon, J., Lebelonyane, R., Ludick, S., Matshediso, E., Sento, B. W., ... and Semo, B. W.
(2014). Improving the quality of health information: a qualitative assessment of data management and
reporting systems in Botswana. Health Res Policy Syst, 12(7), 10-1186.
Lee, A.S., Thomas, M. A., and Baskerville, R.L. (2013). Going Back to Basics in Design: From the IT artefact to
the IS artefact.
Leontev, A. (1981). The problem of activity in psychology. In James V. Wertsch, The Concept of Activity in
Soviet Psychology (pp 40-71). New York: M.E. Sharpe, Inc.
Leontev, A. N. (1978). Activity, Consciousness, and Personality, Prentice-Hall.
Lester, R. T., Ritvo, P., Mills, E. J., Kariri, A., Karanja, S., Chung, M. H., ... and Plummer, F. A. (2010). Effects
of a mobile phone short message service on antiretroviral treatment adherence in Kenya (WelTel Kenya1): a
randomised trial. The Lancet, 376(9755), 1838-1845.
Li, J. and Robertson, T. (2011). Physical space and information space: studies of collaboration in distributed
multi-disciplinary medical team meetings.Behaviour and Information Technology, 30(4), 443-454.
Li, J., Robertson, T., Hansen, S., Mansfield, T., and Kjeldskov, J. (2008). Multidisciplinary medical team
meetings: a field study of collaboration in health care. In Proceedings of the 20th Australasian Conference on
Computer-Human Interaction: Designing for Habitus and Habitat (pp. 73-80). ACM.
Light, A. and Akama, Y. (2012). The human touch: participatory practice and the role of facilitation in designing
with communities. In Proceedings of the 12th Participatory Design Conference: Research Papers-Volume
1 (pp. 61-70). ACM.
Lippeveld T., (2000). Routine data collection methods, In: T. Lippeveld, R. Sauerborn and C. Bodart (eds.).
Design and Implementation of Health Information Systems. Geneva, World Health Organisation, 88 – 113.
Lippeveld, T. and Sauerborn, R., (2000). A framework for designing health information systems. In: T.
Lippeveld, R. Sauerborn and C. Bodart (eds.). Design and Implementation of Health Information Systems.
Geneva: World Health Organisation, 13- 32 .
134
Ludwick, D. A., and Doucette, J. (2009). Adopting electronic medical records in primary care: lessons learned
from health information systems implementation experience in seven countries. International journal of
medical informatics, 78(1), 22-31.
Luff, P., Hindmarsh, J., and Heath, C. (2000). Workplace studies: Recovering work practice and informing
system design. Cambridge university press.
Luk, R., Ho, M., and Aoki, P.M. (2008) Asynchronous remote medical consultation for Ghana. In Proceedings
of the SIGCHI conference on human factors in computing systems, 743-752, ACM.
Lund, H.H., and Sutinen, E. (2010). Contextualised ICT4D: a bottom-up approach. In Proceedings of the 10th
International Conference on Applied Computer Science, WSEAS, Japan.
Lyytinen, K. and Robey, D. (1999). Learning failure in information systems development. Information Systems
Journal, 9(2), 85-101.
Macaulay, C., Benyon, D., and Crerar, A. (2000). Ethnography, theory and systems design: From intuition to
insight. International Journal of Human-Computer Studies, 53(1), 35-60.
MacLeod, B., Phillips, J., Stone, A. E., Walji, A., and Awoonor-Williams, J. K. (2012). The architecture of a
software system for supporting community-based primary health care with mobile technology: the Mobile
Technology for Community Health (MoTeCH) initiative in Ghana. Online Journal of Public Health
Informatics, 4(1).
Madden, R. (2010). Being ethnographic: A guide to the theory and practice of ethnography. Sage Publications.
Maquil, V., Psik, T., Wagner, I., and Wagner, M. (2007, November). Expressive interactions-supporting
collaboration in urban design. In Proceedings of the 2007 international ACM conference on Supporting
group work (pp. 69-78). ACM.
Marcos Y., Phelps, B.R. and Bachman, G. (2012) Community strategies that improve care and retention along
the prevention of mother-to-child transmission of HIV cascade: a review. Journal of the International AIDS
Society, 15, 2: http://www.jiasociety.org/index.php/jias/article/view/17394.
Marcus, G.E. (1995). Ethnography in/of the world system: The emergence of multi-sited ethnography. Annual
review of anthropology, 95-117.
Markussen, R. (1994). Dilemmas in Coopertive Design. In PDC (pp. 59-66).
Mars, M. (2010). Health capacity development through telemedicine in Africa. IMIA Yearbook, 5, 87-93.
Martin, D., Rouncefield, M., O'Neill, J., Hartswood, M., and Randall, D. (2005, November). Timing in the art of
integration:'that's how the bastille got stormed'. In Proceedings of the 2005 international ACM SIGGROUP
conference on Supporting group work (pp. 313-322). ACM.
Martínez-Alcalá, C.I., Muñoz, M., and Monguet-Fierro, J. (2013). Design and customization of telemedicine
systems. Computational and mathematical methods in medicine, 2013.
Mayo-Smith, M.F., and Agrawal, A. (2007). Factors associated with improved completion of computerized
clinical reminders across a large healthcare system. International journal of medical informatics, 76(10), 710716.
Mellini, B. (2013). The Materialization of Articulation Work and its Implications on Nursing Practices.
In CHItaly (Doctoral Consortium) (pp. 74-83).
Miettinen, R. and Hasu, M. (2002). Articulating user needs in collaborative design: Towards an activitytheoretical approach. Computer Supported Cooperative Work, 11(1-2), 129-151.
Milne, C., Scantlebury, K., Blonstein, J., and Gleason, S. (2011). Coteaching and disturbances: Building a better
system for learning to teach science. Research in Science Education, 41(3), 413-440.
Ministry of Health and Social Welfare Tanzania (2006). A manual of the national tuberculosis and leprosy
programme in Tanzania, 5th Edition, Ministry of Health and social welfare, Dar es Salaam, Tanzania.
Ministry of Health and Social Welfare Tanzania (2007a). Assessment of the Country Health Information System
in Tanzania, http://www.who.int/healthmetrics/library/countries/HMN_TZA_Assess_Draft_2007_05_en.pdf
Ministry of Health and Social Welfare Tanzania (2007b). Primary Health Services Development Programme
(PHSDP) 2007 – 2017. Ministry of Health and social welfare, Dar es Salaam, Tanzania.
Ministry of Health and Social Welfare Tanzania (2007c). National Health Policy. Ministry of Health and social
welfare, Dar es Salaam, Tanzania.
Ministry of Health and Social Welfare Tanzania (2008). The National Road Map Strategic Plan To Accelerate
Reduction of Maternal, Newborn and Child Deaths in Tanzania 2008 – 2015. Ministry of Health and social
welfare, Dar es Salaam, Tanzania.
135
Ministry of Health and Social Welfare Tanzania (2010). National Guidelines For Home Based Care Services,
Ministry of Health and social welfare, Dar es Salaam, Tanzania.
Ministry of Health and Social Welfare Tanzania (2013a). Implementation of HIV AIDS Care and Treatment
Services in Tanzania, Report Number 3. Abbreviated Edition. Dar es Salam, Tanzania: MOHSW.
Ministry of Health and Social Welfare Tanzania (2013b). National Guidelines for Comprehensive Services for
Prevention of Mother-to-Child Transmission of HIV, Dar es Salam, Tanzania.
Ministry of Health and Social Welfare Tanzania (2014). National guideline for comprehensive package of HIV
interventions for key populations. Dar es Salam, Tanzania: MOHSW.
Mlambo, M. G., Peltzer, K., & Koivu, A. (2014). Prevention of Mother-to-Child Transmission of HIV data
completeness and accuracy assessment in health facilities of the Nkangala District. Health SA Gesondheid
(Online), 19(1), 1-8.
Mønsted, T., Reddy, M. C., & Bansler, J. P. (2011, January). The use of narratives in medical work: A field
study of physician-patient consultations. In ECSCW 2011: Proceedings of the 12th European Conference on
Computer Supported Cooperative Work, 24-28 September 2011, Aarhus Denmark (pp. 81-100). Springer
London.
Monteiro, E., Pollock, N., Hanseth, O., and Williams, R. (2013). From artefacts to infrastructures. Computer
Supported Cooperative Work, 22(4-6), 575-607.
Mörtberg, C., Bratteteig, T., Wagner, I., Stuedahl, D. and Morrison, A. (2010). Methods that Matter in Digital
Design Research. Exploring Digital Design: Multi-Disciplinary Design Practices. I. Wagner, T. Bratteteig
and D. Stuedahl. London, Springer 105-144.
Moyo, C.M. (2005). An Assessment of the Quality of Health Management Information System Data in Selected
Health Facilities in Lilongwe District. Master Thesis. Department of Public Health, College of Medicine,
University of Malawi.
Mphatswe, W., Mate, K. S., Bennett, B., Ngidi, H., Reddy, J., Barker, P. M., and Rollins, N. (2012). Improving
public health information: a data quality intervention in KwaZulu-Natal, South Africa. Bulletin of the World
Health Organization, 90(3), 176-182.
Murphy, E. and Rodriguez-Manzanares, M.A. (2008). Using activity theory and its principle of contradictions to
guide research in educational technology. Australasian Journal of Educational Technology, 24(4), 442-457
Mursu, A. S., Luukkonen, I., Toivanen, M., and Korpela, M. J. (2006). Activity Theory in information systems
research and practice: theoretical underpinnings for an information systems development model. Information
Research, 12(3), 3.
Mushi, R.T., and Maharaj, M. (2014). Health Information Systems in Tanzania. African Journal of Library,
Archives and Information Science, 23(2), 151-156.
Mwatawala, S., Sando, D., Malele, R., Moshiro, C., Senyael, B.R., Somi, G., Maselle, A., Chalker, J. and RossDegnan, D. (2012). Strengthening the appointment and tracking systems for patients on antiretroviral therapy
in Tanzania: A optimizing adherence to ART as part of people-centered public health. International Journal
of Person Centered Medicine, 2(4), 825-836.
Myers, M.D. (1997). Qualitative research in information systems. Management Information Systems
Quarterly, 21, 241-242.
Myers, M. D., and Avison, D. (2002). An introduction to qualitative research in information systems. Qualitative
research in information systems, 4, 3-12.
Nardi, B.A. (1996). Context and consciousness: Activity theory and human-computer interaction. Cambridge
MA: MIT Press.
NBS and ICF (2011). Tanzania Demographic and Health Survey 2010. Dar es salaam, National Bureau of
Statistics Tanzania and ICF Macro.
NBS and Macro International Inc.. (2007). Tanzania Service Provision Assessment Survey 2006. Dar es Salaam,
Tanzania: National Bureau of Statistics and Macro International Inc.
NBS and OCGS (2014). The 2012 Population and Housing Census: Basic Demographic and Socio- Economic
Profile; Key Findings. Dar es Salaam, Tanzania: National Bureau of Statistics and Office of Chief
Government Statistician (Zanzibar).
NBS and ORC (2005). Tanzania Demographic and Health Survey 2004-2005. Dar es salaam, National Bureau of
Statistics Tanzania and ORC Macro
136
Ngabo, F., Nguimfack, J., Nwaigwe, F., Mugeni, C., Muhoza, D., Wilson, D. R., ... and Binagwaho, A. (2012).
Designing and Implementing an Innovative SMS-based alert system (RapidSMS-MCH) to monitor
pregnancy and reduce maternal and child deaths in Rwanda. The Pan African Medical Journal, 13.
Nglazi, M.D, Kaplan, R., Wood, R., Bekker, L. and Lawn, S.D. (2010) Identification of losses to follow-up in a
community-based antiretroviral therapy clinic in South Africa using a computerized pharmacy tracking
system. BMC Infectious Diseases. http://www.biomedcentral.com/1471-2334/10/329
Ngoma, C. (2007). Cultivation Strategies in the Implementation of Health Management Information System in
Zanzibar: An Action Research Study. Master Degree, Department of Informatics, University of Oslo.
Ngoma, C. (2014). Approaches for Improving the Quality and Accessibility of Maternal and Child Health Data
from Rural Communities: Action-Case Studies from Tanzania, PhD, Faculty of Mathematics and Natural
Sciences, University of Oslo, Norway.
Ngoma, C. and Igira, F. T. (Forthcoming). Re-organising Collection and Recording of Maternal and Child
Health Data through an Intervention of Resolving Contradictions.
Ngoma, C. and Igira, F.T. (2012) Contradictions in Collecting and Recording Maternal Health Data at the
Community Level: A Case study from two Districts in Tanzania. Journal of Health Informatics in
Developing Countries, 6, 2, 509-520.
Ngoma, C., Chawani, M.S. and Herstad, J. (2010) Adaptation of Mobile Application to Improve Flow of Birth
Information from the Community to the District Level. R. Popescu-Zeletin et al. (Eds.) 2nd International
ICST Conference on e-Infrastructure and e-Services for Developing Countries, 79-92.
Ngoma, C., Kaasbøll, J. and Aanestand, M. (2008). From Training to Inservice Support. IST-Africa 2008
Conference and Exhibition, Windhoek, Namibia.
Ngoma, C., Kimaro, H., Igira, F., Mukulu, J., Kaasbøll, J., Yona, A., Mkabara, N., Sarikoki, S., Sangu, I., Njau,
C., Mnaya, A., Mpoyo, S., Mlongakweli, Z., Mwaga, W. and Pilyimo, R. (2012) Challenges in Implementing
a Computerized Namebased Information Tracking System: Practical Experiences from Maternal Health Care.
IST-Africa 2012 Conference Proceedings, P. Cunningham and M. Cunningham (eds). International
Information Management Corporation.
Nguyen T. and Nyella E, (2010) Health information systems implementation in developing countries: A
translation process. AMCIS 2010 Proceedings Paper 63; 2010.
Nhampossa, J.L., Kaasbøll, J., and Braa, J. (2004). Participation in the information system adaptation process in
the public sector in Mozambique. In PDC-04 Proceedings of the Participatory Design Conference, Vol 2.
ACM, New York, 84-88.NBS and ICF (2011).
Nicolini, D. (2009). Zooming In and out: Studying Practices by Switching Theoretical Lenses and Trailing
Connections. Organizational studies, 30, 1391-1418.
Nicolini, D. (2013). Practice Theory, Work, and Organization: An Introduction Oxford University Press United
Kingdom.
Nicolini, D., Gherardi, S., and Yanow, D. (2003). Introduction: Toward a practice-based view of knowing and
learning in organizations. In D. Nicolini, S. Gherardi and D. Yanow (Eds.), Knowing in organizations: A
practice-based approach. ME Sharpe.
Norman, D.A. (1993). Things that make us smart: Defending human attributes in the age of the machine. Basic
Books.
Nsubuga, P., 2001. Surveillance: What is it? How does it relate to routine information gathering and use? Do we
need it? In: The RHINO workshop on issues and innovation in routine health information in developing
countries., The Bolger Center, Protomac, MD, USA 14-16 March 2001. Arlington, VA 22209, USA:
MEASURE Evaluation, JSI Research and Training Institute, 59-63.
Nyella, E. (2009). Challenges in health information systems integration: Zanzibar Experience. In Information
and Communication Technologies and Development (ICTD), 2009 International Conference on (pp. 243251). IEEE.
O’Day, V., Adler, A., Kuchinsky, A., and Bouch, A. (2001). When worlds collide: Molecular biology as
interdisciplinary collaboration. In ECSCW 2001(pp. 399-418). Springer Netherlands.
Oborn, E., Barrett, M., and Davidson, E. (2011). Unity in diversity: electronic patient record use in
multidisciplinary practice. Information Systems Research,22(3), 547-564.
Omary, Z., Lupiana, D., Mtenzi, F., and Wu, B. (2010). Analysis of the challenges affecting e-healthcare
adoption in developing countries: A case of Tanzania. International Journal of Information Studies, 2(1), 3850.
137
Orlikowski, W. J., and Baroudi, J. J. (1991). Studying information technology in organizations: Research
approaches and assumptions. Information systems research, 2(1), 1-28.
Orr, J.E. (1996). Talking about machines: An ethnography of a modern job. Cornell University Press.
Pan, G., Hackney, R., and Pan, S. L. (2008). Information Systems implementation failure: Insights from
prism. International Journal of Information Management, 28(4), 259-269.
Patterson, E. S., Doebbeling, B. N., Fung, C. H., Militello, L., Anders, S., and Asch, S. M. (2005). Identifying
barriers to the effective use of clinical reminders: bootstrapping multiple methods. Journal of biomedical
informatics,38(3), 189-199.
Pedersen, R. and Ellingsen, G. (2011). The electronic patient record-sufficient quality for clinical research?.
In ECIS.
Pilemalm, S., Hallberg, N., and Timpka, T. (2000). From Utopia to DLK: Management of External Voices in
Large Participatory Design Projects. In 6th Biennial Participatory Design Conference 2000, November 28December 1, 2000, New York, USA (pp. 156-165).
Piras, E.M. and Zanutto, A. (2010). Prescriptions, x-rays and grocery lists. Designing a Personal Health Record
to support (the invisible work of) health information management in the household. Computer Supported
Cooperative Work, 19(6), 585-613.
Pitula, K. and Radhakrishnan, T. (2011). On eliciting requirements from end-users in the ICT4D
domain. Requirements Engineering, 16(4), 323-351.
Pollock, N., Williams, R., and Procter, R. (2003). Fitting standard software packages to non-standard
organizations: the ‘biography’ of an enterprise-wide system. Technology Analysis and Strategic
Management, 15(3), 317-332.
Pop-Eleches, C., Thirumurthy, H., Habyarimana, J. P., Zivin, J. G., Goldstein, M. P., De Walque, D., ... and
Bangsberg, D. R. (2011). Mobile phone technologies improve adherence to antiretroviral treatment in a
resource-limited setting: a randomized controlled trial of text message reminders. AIDS (London,
England), 25(6), 825.
Poppe, O., Jolliffe, B., Adaletey, D. L., Braa, J., and Manya, A. S. (2013). Cloud computing for health
information in Africa? Comparing the case of Ghana to Kenya. Journal of Health Informatics in Africa, 1(1).
Pries-Heje, L. and Dittrich, Y. (2009). ERP implementation as design: Looking at participatory design for means
to facilitate knowledge integration. Scandinavian Journal of Information Systems, 21(2), 4.
Puri, S. K., Byrne, E., Nhampossa, J. L. and Quraishi, Z. B. (2004). Contextuality of Participation in IS Design:
A Developing Country Perspective. Proceedings of the Participatory Design Conference 2004, ACM Press,
42-52.
Puri, S., Sahay, S., and Lewis, J. Building Participatory HIS Networks: A Case Study from Kerala India,
Information and Organization, 2009, 19, 63-83.
Ramadhani, H. O., Thielman, N. M., Landman, K. Z., Ndosi, E. M., Gao, F., Kirchherr, J. L., ... and Crump, J.
A. (2007). Predictors of incomplete adherence, virologic failure, and antiviral drug resistance among HIVinfected adults receiving antiretroviral therapy in Tanzania. Clinical Infectious Diseases,45(11), 1492-1498.
Ramduny-Ellis, D., Dix, A., Rayson, P., Onditi, V., Sommerville, I., and Ransom, J. (2005). Artifacts as
designed, artifacts as used: resources for uncovering activity dynamics. Cognition, Technology and
Work, 7(2), 76-87.
Reddy, M.C., Dourish, P., and Pratt, W. (2001). Coordinating heterogeneous work: information and
representation in medical care. InECSCW 2001 (pp. 239-258). Springer Netherlands.
Reddy, M., Bardram, J., and Gorman, P. (2010) CSCW research in healthcare: Past, present, and
future.In Proceedings of the Conference on Computer Supported Cooperative Work. Retrieved from
http://sites.google. com/site/cscwinhc.
Ren, Y., Kiesler, S., and Fussell, S. R. (2008). Multiple group coordination in complex and dynamic task
environments: Interruptions, coping mechanisms, and technology recommendations. Journal of management
information systems, 25(1), 105-130.
Rispel, L. C., Peltzer, K., Phaswana-Mafuya, N., Metcalf, C. A., and Treger, L. (2009). Assessing missed
opportunities for the prevention of mother-to-child HIV transmission in an Eastern Cape local service
area. SAMJ: South African Medical Journal, 99(3), 174-179.
RMO (2013). Dodoma Regional Health Management Team Annual Plan and Budget, 2013/2014. Dodoma: The
United Republic of Tanzania.
138
Robertson, T. and Simonsen, J. (2012). Challenges and opportunities in contemporary participatory
design. Design Issues, 28(3), 3-9.
Robertson, T., Li, J., O’Hara, K., and Hansen, S. (2010). Collaboration within different settings: A study of colocated and distributed multidisciplinary medical team meetings. Computer Supported Cooperative
Work, 19(5), 483-513.
Rothstein, M.A. (2010). Is deidentification sufficient to protect health privacy in research?. The American
Journal of Bioethics, 10(9), 3-11.
Rubona, J.J. (2001). Routine health information systems that operate in Tanzania. In The RHINO workshop on
issues and innovation in routine health information in developing countries (pp. 183-193).
Sæbø, J. I., Kossi, E. K., Titlestad, O. H., Tohouri, R. R., and Braa, J. (2011). Comparing strategies to integrate
health information systems following a data warehouse approach in four countries. Information Technology
for Development, 17(1), 42-60.
Safadi, H. and Faraj, S. (2010) The role of workarounds during an opensource electronic medical record system
implementation (2010). ICIS 2010 Proceedings. Paper 47.
Sangiorgi, D. and Clark, B. (2004). Toward a participatory design approach to service design. In PDC (pp. 148151).
Sapirie, S. 2000. Assessing health information systems. In: T. Lippeveld, R. Sauerborn and C. Bodart (eds.).
Design and Implementation of Health Information Systems. Geneva: World Health Organisation, 73 – 87.
Sauerborn, R. (2000). Using information to make decisions. Design and Implementation of Health Information
Systems. Geneva: World Health Organisation, 33-48.
Sauerborn, R. and Lippeveld, T. (2000) Introduction. In: T. Lippeveld, R. Sauerborn, and C. Bodart (eds.)
Design and Implementation of Health Information System World Health Organisation Geneva, 1-14.
Saugene, Z. B. (2013). Customization of Generic Open Source Software for Health Sector in Developing
Countries. PhD, Department of Informatics, University of Oslo, Norway.
Saugene, Z. and Kaasbøll, J. (2014) Foregrounding invisible work practices: the customization of generic
software systems.
Schmidt, K. (1991). Riding a tiger, or computer supported cooperative work. In Proceedings of the Second
European Conference on Computer-Supported Cooperative Work ECSCW’91 (pp. 1-16). Springer
Netherlands.
Schmidt, K. (2011). Cooperative work and coordinative practices (pp. 3-27). Springer London.
Schmidt, K. and Bannon, L. (1992). Taking CSCW seriously. Computer Supported Cooperative Work, 1(1-2), 740.
Schmidt, K. and Simonee, C. (1996). Coordination mechanisms: Towards a conceptual foundation of CSCW
systems design. Computer Supported Cooperative Work, 5(2-3), 155-200.
Schmidt, K. and Wagner, I. (2004). Ordering systems: Coordinative practices and artifacts in architectural design
and planning. Computer Supported Cooperative Work, 13(5-6), 349-408.
Schmidt, K., Wagner, I., and Tolar, M. (2007). Permutations of cooperative work practices: a study of two
oncology clinics. In Proceedings of the 2007 international ACM conference on Supporting group work (pp.
1-10). ACM.
Seebregts, C. J., Mamlin, B. W., Biondich, P. G., Fraser, H. S., Wolfe, B. A., Jazayeri, D., ... and Network, T. O.
I. (2009). The OpenMRS implementers network. International journal of medical informatics, 78(11), 711720.
Shaw, V., and Braa, J. (2014). Approaches to participatory design in Africa in the age of cloud computing.
In Proceedings of the 13th Participatory Design Conference: Short Papers, Industry Cases, Workshop
Descriptions, Doctoral Consortium papers, and Keynote abstracts-Volume 2 (pp. 45-48). ACM.
Shidende, N.H. 2005. Challenges And Approaches to the Integration of HIS: Case Studies from Tanzania.
Master thesis, University of Oslo.
Shrestha, L B. and Bodart, C., 2000. Data transmission, data processing, and data quality. In: T. Lippeveld, R.
Sauerborn and C. Bodart (eds.). Design and Implementation of Health Information Systems. Geneva: World
Health Organisation, 128 – 145.
Silverman, D. (2005). Doing qualitative research. 2nd edition. Sage Publications
139
Simba, D., Kamwela, J., Mpembeni, R., and Msamanga, G. (2010). The impact of scaling-up prevention of
mother-to-child transmission (PMTCT) of HIV infection on the human resource requirement: the need to go
beyond numbers. The International Journal of Health Planning and Management, 25(1), 17-29.
Simba, D.O. and Mwangu, M.A. (2005) Quality of a routine data collection system for health: case of Kinondoni
district in the Dar es Salaam region, Tanzania, South African Journal of Information Management, 7, 2.
Simone, C., and Schmidt, K. (1998). Taking the distributed nature of cooperative work seriously. In euromicropdp (p. 0295). IEEE.
Sinnig, D., Pitula, K., Becker, R., Radhakrishnan, T., and Forbrig, P. (2010). Structured Digital Storytelling for
Eliciting Software Requirements in the ICT4D Domain. In Human-Computer Interaction (pp. 58-69).
Springer Berlin Heidelberg.
Smit, D. and Venter, E. (2009). Initial experiences from designing and implementing a clinical health
information system in the primary healthcare sector of rural South Africa. IST Africa.
Smith, M., Madon, S., Anifalaje, A., Lazarro-Malecela, M., and Michael, E. (2007). Integrated health
information systems in Tanzania: experience and challenges. The Electronic journal of information systems
in developing countries, 33.
Smith, S.W., and Koppel, R. (2013). Healthcare information technology's relativity problems: a typology of how
patients’ physical reality, clinicians’ mental models, and healthcare information technology differ. Journal of
the American Medical Informatics Association, amiajnl-2012.
Somi, G., Matee, M., Makene, C. L., Van Den Hombergh, J., Kilama, B., Yahya-Malima, K. I., ... and Swai, R.
O. (2009). Three years of HIV/AIDS care and treatment services in Tanzania: achievements and
challenges. Tanzania journal of health research, 11(3), 136-143.
Spinuzzi, C. (2005). The methodology of participatory design. Technical Communication, 52(2), 163-174.
Star, S.L. and Ruhleder, K. (1994). Steps towards an ecology of infrastructure: complex problems in design and
access for large-scale collaborative systems. In Proceedings of the 1994 ACM conference on Computer
supported cooperative work (pp. 253-264). ACM.
Star, S.L. and Strauss, A. (1999). Layers of silence, arenas of voice: The ecology of visible and invisible
work. Computer supported cooperative work, 8(1-2), 9-30.
Star, S. L., Bowker, G. C., and Neumann, L. J. (2003). Transparency beyond the individual level of scale:
Convergence between information artifacts and communities of practice. Digital library use: Social practice
in design and evaluation, 241-269.
Starring, K. (2011). Organizational open source in the Global South. PhD Thesis, University of Oslo,
Department of Informatics.
Strauss A.L. (1991). Creating Sociological Awareness. Collective Images and Symbolic Representations. New
Brunswick: Transaction Publishers.
Suchman, L. (2002). Located accountabilities in technology production. Scandinavian journal of information
systems, 14(2), 7.
Sychareun, V., Hansana, V., Phengsavanh, A., Chaleunvong, K., Eunyoung, K., and Durham, J. (2014). Data
verification at health centers and district health offices in Xiengkhouang and Houaphanh Provinces, Lao
PDR. BMC Health Services Research, 14(1), 255.
Tanzania Commission for AIDS (TACAIDS), Zanzibar AIDS Commission (ZAC), National Bureau of Statistics
(NBS), Office of the Chief Government Statistician (OCGS), and ICF International 2013. Tanzania
HIV/AIDS and Malaria Indicator Survey 2011-12. Dar es Salaam, Tanzania: TACAIDS, ZAC, NBS, OCGS,
and ICF International.
Thai, S., Koole, O., Un, P., Ros, S., De Munter, P., Van Damme, W., ... and Lynen, L. (2009). Five-year
experience with scaling-up access to antiretroviral treatment in an HIV care programme in
Cambodia. Tropical Medicine and International Health, 14(9), 1048-1058.
Thompson, J.D. (1967). Organization in Actions, McGraw Hill.
Thomson, K.A., Cheti, E.O. and Reid, T. (2011) Implementation and outcomes of an active defaulter tracing
system for HIV, prevention of mother to child transmission of HIV (PMTCT), and TB patients in Kibera,
Nairobi, Kenya. Transaction of the Royal Society of Tropical Medicine and Hygiene, 105, 320-326.
Tiihonen, T. (2009). Mapping the Information System Context in Organizations. Defining the ‘D’in ICT4D, 27.
Tiihonen, T. (2011). Information systems in context. University of Eastern Finland.
140
Tiihonen, T., Korpela, M., and Mursu, A. (2006). Creating a framework to recognize context-originated factors
in IS in organizations. In Social Informatics: An Information Society for all? In Remembrance of Rob
Kling (pp. 367-379). Springer US.
Tiihonen, T., Luukkonen, I., and Korpela, M. (2010). Modelling the context of information systems. In ISTAfrica, 2010 (pp. 1-10). IEEE.
Titlestad, O., Staring, K., and Braa, J. (2009). Distributed Development to Enable User Participation: Multilevel
design in the HISP network. Scandinavian Journal of Information Systems, 21(1), 3.
Tongia, R., and Subrahmanian, E. (2006). Information and Communications Technology for Development
(ICT4D)-A design challenge?. In Information and Communication Technologies and Development, 2006.
ICTD'06. International Conference on (pp. 243-255). IEEE.
Tswane, S., Korpela, M., M’Rithaa, D. K., and Kabaso, B. (2015). The maternal healthcare landscape around
Grabouw, South Africa: Setting the stage for information systems development. Journal of Health
Informatics in Africa, 2(2).
Van Bemmel J.H. and Musen M.A. Handbook of medical informatics. Heidelberg: Springer; 1997.
Van Maanen, J. (1988). Tales of the field: On writing ethnography. The university of Chacago Press, Chicago
Vikkelsø, S. (2005). Subtle redistribution of work, attention and risks: Electronic patient records and
organisational consequences. Scandinavian Journal of Information Systems, 17(1), 10.
Vygotsky, L.S. (1978). Mind and Society: The Development of Higher Psychological Process. Cambridge MA,
Wagenaar, B. H., Gimbel, S., Hoek, R., Pfeiffer, J., Michel, C., Manuel, J. L., ... and Sherr, K. (2015). Effects of
a health information system data quality intervention on concordance in Mozambique: time-series analyses
from 2009–2012. Population health metrics, 13(1), 9.
Walsham, G. (1995). Interpretive case studies in IS research: nature and method. European Journal of
information systems, 4(2), 74-81.
Walsham, G. (1997). Actor-network theory and IS research: current status and future prospects. In Information
systems and qualitative research (pp. 466-480). Springer US.
Walsham, G. (2006). Doing interpretive research. European journal of information systems, 15(3), 320-330.
Watson-Jones, D., Balira, R., Ross, D. A., Weiss, H. A., and Mabey, D. (2012). Missed opportunities: poor
linkage into ongoing care for HIV-positive pregnant women in Mwanza, Tanzania. PLoS One, 7(7), e40091.
Williams, R., Stewart, J., and Slack, R. (2005). Social learning in technological innovation: Experimenting with
information and communication technologies. Edward Elgar Publishing.
Winschiers-Theophilus, H., Bidwell, N. J., and Blake, E. (2012). Altering participation through interactions and
reflections in design. CoDesign, 8(2-3), 163-182.
Winschiers-Theophilus, H., Chivuno-Kuria, S., Kapuire, G. K., Bidwell, N. J., and Blake, E. (2010). Being
participated: a community approach. In Proceedings of the 11th Biennial Participatory Design
Conference (pp. 1-10). ACM.
Wischweh, J.D. and Bade, D. (2012). Activity-Oriented Context Adaptation in Mobile Applications. In Mobile
and Ubiquitous Systems: Computing, Networking, and Services (pp. 298-313). Springer Berlin Heidelberg.
World Health Organization. (1988). From alma Ata to the year 2000: reflections at the midpoint.
World Health Organization. (2008). Framework and Standards for Country Health Information systems/ Health
Metrics Network, World Health Organization.
World Health Organization/UNICEF: (1978). Primary Health Care. Health for all. Geneva, World Health
Organization.
Wyatt, J.C. and Wright, P. (1998). Design should help use of patients' data. The Lancet, 352(9137), 1375-1378.
Xue, Y., Liang, H., Boulton, W. R., and Snyder, C. A. (2005). ERP implementation failures in China: Case
studies with implications for ERP vendors. International journal of production economics, 97(3), 279-295.
141
Appendices
142
Appendix 1: Paper I
Shidende, N.H., Grisot, M., Aanestad M., and Igira, F.T. (2014). Coordination challenges in
Collaborative Practices in the Prevention of Mother to Child Transmission of HIV in
Tanzania, Journal of Health Informatics in Africa, 2(1).
I
Coordination Challenges in Collaborative Practices in the Prevention
of Mother to Child Transmission of HIV in Tanzania
Nima Herman Shidende, Miria Grisot, Margunn Aanestad
University of Oslo, Norway
[email protected], [email protected], [email protected]
Faraja Teddy Igira
Institute of Finance Management, Tanzania
[email protected]
Background and Purpose: This paper describes some of the complexities that face health service
provision in developing countries, with the motivation to inform design of appropriate information
systems. In particular, we are interested in a better understanding of the challenges to coordination
and collaboration between health staff that are located in different facilities and employed to work in
different health programs.
Methods: The study reported and analyzed in this paper was conducted in two districts in Tanzania.
Using ethnographic data collection methods, we studied health workers’ practices of coordination and
collaboration in Prevention of Mother to Child Transmission (PMTCT) services in Tanzania.
Results: In our study we describe the collaboration required when managing patient trajectories of
PMTCT patients across facilities and programs, and how contingencies may change the course of a
patient trajectory. We provide a rich empirical description of coordination work in a resource constrained setting and we propose improvements to the design of both computer and paper-based information systems.
Conclusions: The rich empirical description of coordination work in a resource constrained setting
and our analysis of coordination challenges contribute to a better understanding that can strengthen
collaboration and thus also improve health care provision.
Keywords: Patient-Care Information Systems, Collaborative Practices, Coordination Mechanisms,
Coordination Artefacts
1
Introduction
While management information systems have received much attention in the health informatics literature
of the developing countries, recently there is a focus on designing and implementing computerized patient
information systems that are intended for tracking individual health problems and treatment over time
(WHO, 2012). For instance clinical information systems for HIV care in outpatients, and inpatients clinics
(Fraser et al., 2007; Kamadjeu, Tapang, & Moluh, 2005; Oluoch et al., 2012; Rotich et al., 2003). However, the design and implementation of computerized record systems is a complicated endeavour with
many challenges (e.g. Berg, 1999b; Jones, 2003; Robertson et al., 2010; Safadi & Faraj, 2010; Vikkelsø,
2005). These studies for instance indicate that there is a main tension between existing work practices and
new technologies. In this paper we contribute to understanding the challenge of designing patient information system by examining the existing work practices of tracking individual health problems and
treatment over time. These practices require collaboration and coordination across facilities, health programs and in time along patient trajectories.
*Corresponding author: University of Oslo, Norway. Email: [email protected]
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
2
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
Collaboration and coordination are key characteristics of health service provision (Berg, 1999a, 1999b;
Reddy, Bardram, & Gorman, 2010; Schmidt, Wagner, & Tolar, 2007). Healthcare provision involves
many professionals and often different specialties treat a single patient at the same time and location, or
across time and locations (Reddy et al., 2010). Patient care develops as a trajectory in time that is partly
shaped by the work of health professionals and partly emerges in a contingent way (Fagerhaugh &
Strauss, 1997). Thus, the management of patient trajectory is a collective and cooperative effort of health
providers where health practitioners not only have to inter-relate their work tasks, but also to know where
the process of managing a patient’s trajectory is (Berg, 1999b). When many health practitioners are involved in treating patients, patients’ records are an important resource used to assess the overview of care
provided (Berg, 1999b; Fitzpatrick, 2004). Patient records are repositories of patient information within
an institution (Mønsted, Reddy, & Bansler, 2011). They perform a role of communicative artefacts between health providers and configure health care provision services, for instance by organizing consultations between providers and clients (Engestrom, Engestrom, & Saarelma, 1988; Heath & Luff, 1996).
They also support administrative tasks (Garfinkel & Bittner, 1967). Moreover digitized medical records
may have reminders which can prompt providers to perform due health care tasks (Mayo-Smith &
Agrawal, 2007; Patterson et al., 2005). Further, Berg (1999b) asserts that both formal and informal aspects of patient records together with the work of clinicians are important for providing healthcare and
supporting the collaborative work it entails.
Empirical studies conducted in a developing country context, have address the issue of collaboration in
relation to the use of health information systems. For instance, telemedicine supports collaborative practices by facilitating communication between providers (sometimes with patients), the sharing of medical
information, and also educational purposes (Bagayoko, Anne, Geissbuhler, & Fieschi, 2008; Bath, 2006;
Geissbuhler, Ly, Lovis, & L’Haire, 2003; Mars, 2010; Martínez-Alcalá, Muñoz, & Monguet-Fierro,
2013). Geissbuhler et al. 2003 describe a telemedicine project in Western Africa that facilitated collaboration between Malian practitioners and practitioners at Geneva University (Geissbuhler et al., 2003). Other
studies examine collaboration in the context of software development activities for health systems between different ICT specialists located in different settings (Korpela, Mursu, & Soriyan, 2002; Saugene,
2013; Saugene & Kaasbøll, 2013). However, despite the focus on collaboration, researchers in health
informatics have given less attention to the understanding of collaborative practices during ordinary
health service provision. Health practices in a developing country context often cut across multiple health
programs provided in a dispersed and distributed setting and with limited resources, and require complex
coordinative and collaborative work in order to be performed.
In this paper we aim to empirically describe and examine the complexities of collaborative work practices. Our study focuses on the work practices of healthcare providers in a cross-setting collaborative
practice, and their use of patient-oriented artifacts. Specifically, our empirical material comes from research in the context of prevention of mother to child transmission of HIV (PMTCT) program in Tanzania. In this context we have empirically studied the collaborative practices of healthcare providers and the
trajectories of HIV-positive pregnant women. These trajectories develop in time, across facilities and
often across other health programs. In our analysis we pay particular attention both to information artefacts used in the facilities to track patients, and to those carried by patients themselves to bring information from one provider to another or to keep track of their visits. Our argument, based on CSCW research, is that the design of patient-care information systems should be based on understanding the collaborative practices of those actually working with patient information during ordinary health service
provision and ‘doing’ the coordination work of patient trajectories. Our aim is to improve the understanding of coordination in collaborative practices and its associated challenges in healthcare provision in a
resource constrained context, and to draw design implications for patient-oriented information systems.
The following question is addressed in the study: What are the challenges to coordination in collaborative
practices and what are the implications for the design of patient information systems as coordinative artefacts?
The rest of the paper is structured as follows. First, we review the relevant literature from CSCW and
we present key concepts that are used in this study. Then, in section 3, we present the research context
and research methodology. This is followed by the case description in section 4. Section 5 presents our
findings and implications for design, and section 6 our conclusions.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
2
3
Literature Review and Theory
Our theoretical approach is based on concepts from studies in Computer-Supported Collaborative Work
(CSCW). The aim of the CSCW field is to understand the nature of cooperative work in its natural settings with the objective of designing computer based technologies to support work practices (Schmidt &
Bannon, 1992). People engage in cooperative work when they are mutually dependent in their tasks and
therefore are required to cooperate in order to get the work done (Schmidt, 1991). However, tasks interdependencies are tractable only if appropriate coordination mechanisms are in place (Schmidt & Simonee, 1996). A coordination mechanism is “a coordinative protocol with an accompanying artifact such as
for instance a standard operating procedure supported by a certain form” (Simone & Schmidt, 1998, p.
295). Actors participating in cooperative work are assisted by coordination mechanisms in managing an
otherwise overwhelming complexity (Simone & Schmidt, 1998).
In healthcare, collaborative practices are often mediated and supported by artefacts and coordinative
protocols (Berg, 1999b). For example, coordinative artifacts support activities such as marking, checking,
reading and writing and contribute to coordinate activities of practitioners in time (Berg, 1999a). Many
studies in the CSCW tradition have investigated empirically the use of coordinative artefacts in order to
draw implications for the design of digital artifact for supporting collaborative work (Berg, 1999a; Bjørn
& Hertzum, 2011; Reddy et al., 2010). Bjørn and Hertzum (2011) for example, examined the use of
whiteboards in collaborative work practices of health providers, and draw implications for improving the
design of digital whiteboards. Likewise, Bardram (1998) reports a study in a large Danish hospital where
he studied coordination and planning of patient care, and informed the design of a computer system for
planning and scheduling surgical operations, and for requesting and booking diagnostic examination at
other departments. These studies show the relevance of understanding the actual work practices of
healthcare providers in their work settings in order to draw insights for designing improved collaborative
artefacts.
Research in the field of CSCW recognizes that the design and implementation of coordination artefacts
is challenging. For example, Cabitza et al. assert that artefacts could be not rich enough to facilitate communication between participants (Cabitza, Sarini, Simone, & Telaro, 2006). The deficit in coordinative
artifacts may lead to the development of informal artifacts (Mellini, 2013), or the development of workarounds (Gasser, 1986) or unintended practices (Fitzpatrick, 2004). For instance, Fitzpatrick (2004) describes how clinicians tailored and augmented patient records in order to support their own role and practices. Safadi and Faraj (2010) show how workarounds emerged after the implementation of an EMR system in an ambulatory clinic in Canada. For instance since the software did not allow adding new disease
to a list of favorite diseases, clinicians recorded diseases in the medical history of the patient. The study
shows how workarounds carry rich knowledge about the needs of the users and the required customizations of artefacts.
While healthcare work has received considerable attention in CSCW studies less attention has been
given to studying healthcare work in resource constrained settings. In a recent article reviewing 25 years
of publications in CSCW in healthcare, Fitzpatrick and Ellingsen assert that “the developing world presents a very different context for CSCW research and design - reliable infrastructure cannot be depended
on, markets are less vendor dominated, and there is significantly lower access to resources than in the
developed world” (Fitzpatrick & Ellingsen, 2013, p. 33). However, in CSCW few studies reports from
empirical cases conducted in such ‘different’ context. Specifically, in their review Fitzpatrick and Ellingsen have found only three publications reporting studies conducted in Africa (Fitzpatrick & Ellingsen,
2013). In the first study, Cheng and colleagues assess residents’ attitudes toward the use of handheld
computers in the collection of HIV/AIDS survey data in Angola (Cheng, Ernesto, & Truong, 2008). They
compare residents’ attitudes toward handheld computers with their attitudes toward the standard mode of
data collection, paper surveys. The authors conclude that computerized data collection in Sub-Saharan
Africa may lead to biased reports of HIV/AIDS-related risk behaviors, when compared to the traditional
method of paper-and-pencil surveys due to social-economic factors. In the second study, DeRenzi and
colleagues discusses how to facilitate management of child health through the design of software, specifically a Personal Digital Assistant (PDA) system which would help clinicians make decision and adhere to
child treatment protocol (DeRenzi et al., 2008). During their study in Tanzania which focused on humancomputer collaboration, some features from PDA software had to be redesigned in order to match the
observed human-PDA collaborative practices. In the third study, Luk and colleagues describe how pro© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
4
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
viders collaborate with the use of telemedicine in Ghana and provide principles to inform the design of
telemedicine systems in developing regions (Luk, Ho, & Aoki, 2008). The remote system enabled doctors
to enter medical case information into a distributed repository and then requesting a consultation from a
pool of medical specialists. As illustrated in these three studies, in health care service provision, coordination is important and can be improved by designing and implementing digital information systems. Hence
challenges to collaboration and coordination in context with less resources and infrastructures need more
attention, empirical research and understanding. In this study, we aim to contribute to understanding challenges to collaboration based on a study of maternal and child health care in Tanzania.
3
Research Setting and Methodology
This study is part of the Health Information System Program (HISP) which is run by the Department of
Informatics, University of Oslo. The primary goal of HISP is to enhance the information use behavior of
health managers, planners and workers, in the health sector of developing countries, by strengthening
local professionals’ capacity, for the development of a sustainable HIS (Braa, Monteiro, & Sahay, 2004).
HISP has developed software known as DHIS (District Health Information Software), which is used to
manage data in various levels of the health sector. Recently, HISP has focused on developing patientoriented information systems (known as the DHIS Tracker module) which could be used for managing
name-based patient data and also for providing reminders and alerts to practitioners and clients (Gizaw,
Mukherjee, Lewis, & Sahay, 2012; Saugene, 2013). Our study contributes to HISP by providing an empirical understanding of the existing information practices in order to design patient-oriented information
systems.
The study reported and analyzed in this paper was conducted in Tanzania, in Kitangili and Singidani
districts1 in Dodoma Region. The HIV prevalence among pregnant mothers in Kitangili and Singidani
districts are 2% and 8% respectively. In Tanzania, primary health facilities (dispensaries and health centers) are responsible for health care provision such as curative, preventive, promotive, rehabilitative and
palliative care to the population. Tanzania, is striving to reduce the impact of HIV pandemics by using
various interventions, such as, prevention of mother to child transmission of HIV (PMTCT). Within the
PMTCT program the aim is to prevent transmission of HIV/AIDS from the HIV positive pregnant women
to the children. Based on guidelines from the World Health Organization (WHO), the program offers
interventions during pregnancy, birth and breastfeeding period. One of the operational targets to be
achieved by the year 2015 by the Tanzanian Government is to have PMTCT services provided to at least
80% of pregnant women, their babies and families. Comprehensive intervention for PMTCT services
involves multiple clinics such as maternity care clinics, HIV/AIDS clinics (CTC), tuberculosis (TB) clinics (in case pregnant women develop TB infection), and clinics for children up to 5 years of age.
The data collection for the study was conducted during three periods: August 2011-January 2012, July
– August 2012, and November 2013-April 20132. Eleven primary health facilities offering maternal and
child healthcare services were researched. The study design was inspired by multi-site ethnographic (see
for example Madden, 2010; Marcus, 1995; Blomberg & Karasti, 2013; Nicolini, 2013). This approach
was useful for gaining a deeper understanding of the work practices in different places, albeit being difficult to travel to geographically dispersed facilities in the two districts. The main data collection technique
employed was participant observation by the first author with her role altering between observer as participant, and participant as observer (Bryman, 2012). In total, 101 sessions of observations were conducted
in health facilities with each lasting between 3 to 7 hours. Observations were conducted in antenatal sections, maternity care clinics, children clinics, HIV clinics and one TB clinic. We performed in-situ interviews (Jordan and Henderson, 1995) in Kiswahili language to health providers in facility and Community
Health Workers (CHW). In total we encountered 42 nurses, 13 doctors, 5 data clerks, 2 patient-tracing
coordinators3, and 6 CHW who were involved in PMTCT services.
We also studied and reviewed artifacts which were used to organize the collaborative work of service
providers during health service provision. Artifacts reviewed included clients’ cards, clients’ folders,
1The
actual district names are not disclosed.
fieldwork was conducted as part of PhD work and it required spending at least one year in the field settings.
3Community health workers who coordinate defaulter tracing activities in certain facilities
2The
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
5
referral forms, laboratory forms and facility registers. We examined the artifacts to understand how they
facilitated collaborative work by making visible work of others and coordinating one’s own work. We
also conducted informal interviews with 5 (4 women and 1 man) clients4 who gave oral consent first to
health providers and then to us. The theme of the interviews was their experience of going to multiple
settings carrying coordinative artifacts, as for instance patient cards and referral forms. The efforts to
involve more clients did not bear fruits since in some cases nurses disapproved or we sensed that clients
were not in good state or unwilling to cooperate. For instance, we arranged some appointments where
clients who had agreed to be involved did not show up. Thus, most of our data are based on observation
of meetings between nurses and clients in which we listened to their conversations, took notes, and observed information practices when clients’ information was registered or retrieved in different programs
and settings. In addition, we reviewed several policy documents such as Tanzanian Health Policy, Reproductive and Child Health (RCH) Strategic Plan, PMTCT guidelines, Home Based Care (HBC) service
guidelines, national guidelines for the management of HIV/AIDS and HIV Act and policy. These documents gave us background information about PMTCT healthcare provision protocols.
To get a better understanding of the observed coordination practices, we also interviewed coordinators
for the health services supporting maternal and child health and HIV/AIDS services at district, regional
and zonal5 levels. A summary of the interviewees in relation to their program of affiliation and their work
places is presented in Table 1.
Table 1. Summary of interviews at district, regional and zone levels
Location
Type of
informant
Number of
informants
Kitangili
RCH
HIV/AIDS
3
2
Total number of
interviews (with
repetition)
9
4
Singidani
RCH
HIV/AIDS
5
3
11
8
Regional
RCH
HIV/AIDS
2
1
3
1
RCH
1
17
1
37
Zone
Total
The empirical study was approved by University of Dodoma6’s Research and Publication ethical
committee. Ethical concerns were taken into consideration during fieldwork, for example, through gaining written permission prior to the fieldwork from both districts managers, and obtaining oral consent
from health workers and clients prior to conducting interviews or observations. We have also changed the
name of clients as well as districts and facilities in order to provide anonymity.
The study employed an interpretive approach for data analysis (Walsham, 1995). Notes were taken
during fieldworks and later transcribed. Some interviews with nurses were recorded and transcribed. Field
notes were read several times in order to obtain an understanding of the data, and concepts of cooperative
work, patient trajectory, coordination mechanisms, and coordinative artefacts were used to make sense of
the practices observed. Our interest was directed towards identifying and understanding the challenges
faced by maternal and child health provision in coordinating their collaborative work. For instance, nurses
during care encounters would interrogate or remind clients about attendance in other visits to check their
compliance to health programs, but they would not have any formal way to check that attendance actually
took place.
In order to organize and analyze the material, we employed both inductive and deductive approaches.
For the first level of analyses, we prepared “ethnographic descriptions” (Emerson et al, 1995) of work
practices within and across health facilities in order to foreground collaborative practices. We then creat4
The terms clients, pregnant women and patients have been used interchangeably
Intermediary administrative level for several regions which is under the national level
6 First Author is employee of the university.
5
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
6
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
ed narrative descriptions (Emerson et al, 1995) of work practices in maternal and child health settings that
focused on how information artifacts in different units were used in the collaborative process. To prepare
ethnographic description, we have worked with a characterization strategy (Van Maanen, 1988). We have
combined the stories of multiple clients into the character of Joyce, obtained from our interviews with
clients and our observation of interactions of nurses and clients for successful cases of clients’ compliance. In addition, we also used stories from interviews and observations to construct the characters of
Prisca and MdalaMatonya. These two women’s stories showed different trajectories and exemplify unsuccessful cases of coordination. Van Maanen (1988) argues that this strategy makes it easier to tell
fieldwork tales (p. 104). Thus we believe that the narrative descriptions and the way they are constructed
improve readers’ understanding about the interactions between different units involved in PMTCT care,
the use of coordinative artifacts and how they facilitate collaborative work.
4
Case Description
This section describes how PMTCT service delivery is organized in practice and which coordinative artifacts are in use. Tanzania’s policies both for HIV and RCH indicate how clients should be attended by
different health providers, and health authorities have provided specific coordination protocols. Yet, our
description shows how protocols and coordinative artefacts are not always sufficient to coordinate work.
PMTCT interventions include testing and counseling for HIV, antiretroviral prophylaxis for HIVinfected pregnant women (and their partners) in antenatal care and their exposed children in maternity
wards, treatment of eligible women, counseling and support for infant feeding, safer obstetric practices
and family planning to prevent unintended pregnancies in HIV-infected women. The actual treatment of
HIV/AIDS is conducted in cooperation with the general HIV/AIDS Care and Treatment Clinics (CTC).
Not all health facilities would offer ANC, PMCTC and CTC services, and patient will then be referred to
other facilities for one or more of these services.
In this context, PMTCT providers have three main information and coordination needs. First, as the
PMTCT program covers the pregnancy and the first 18 months of a children’s life, PMTCT providers
need to keep track of information over time. Second, during pregnancy HIV-positive women may also
develop other diseases, for instance TB and may need to enroll in and comply with other health services
providing testing, visits and counseling. In such cases PMTCT providers need to be informed about these
services in order to adjust their treatments accordingly or to check that clients attend them. Third, attending PMTCT services often requires women to travel because many facilities do not host all sections. For
instance a woman may attend services in a facility that has a maternity care clinic but not a CTC clinic. In
addition a woman may prefer to attend a CTC clinic that is not in her residential area to avoid meeting
familiar people. In these cases, health providers need to be informed also of the services delivered in other
facilities. Thus PMTCT services are not delivered at the same point of care but across facilities and over a
period of some years (from pregnancy and until the child is 18 months of age, and can continue to family
planning services in the family planning section). The information and coordination needs are addressed
by using various coordination artefacts, protocols and procedures. However, breakdowns in how information is recorded, updated, transmitted undermine coordination.
To show the complexity of PMTCT service delivery we have organized our case material according to
the chronological order of events (HIV diagnosis, pregnancy, delivery and postnatal care) and we have
used the stories of three women enrolled in the PMTCT program: Joyce, Prisca, and MdalaMatonya. Our
main character is Joyce, a woman who is tested HIV positive and is enrolled in CTC services. Afterwards
she becomes pregnant and is enrolled in ANC and PMTCT services. Next Joyce develops tuberculosis
and is enrolled in TB services. In addition we use vignettes of other characters - Prisca and MdalaMatonya –two HIV-positive pregnant women who are enrolled in the PMTCT program but have a different
trajectory than Joyce. In our description we focus on movements across programs and facilities, on instances of coordination work needed to deliver PMTCT services, and on information artefacts used to
coordinate work. We also include in our description instances of breakdowns that we have observed during fieldwork. Our aim is to make visible the complexity of clients’ trajectories, the structure and use of
coordinative artefacts, the coordinative work performed by health providers and the coordination challenges leading to breakdowns.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
4.1
7
Diagnosis: from Outpatient to Voluntary Counseling and Testing (VCT) to Care and
Treatment Clinics (CTC)
Joyce is a housewife living in the catchment area of H facility. In the year 2006 her health condition deteriorated and she decided to seek help. Joyce decided to go to the outpatient clinic at facility D, which was
not located in her residential area but offered higher level services. The doctor at facility D suspected that
she could be infected with HIV and referred her to the voluntary counseling and testing (VCT) section
within the same facility. After a counseling session, Joyce was tested and diagnosed HIV positive and
referred to the CTC unit (i.e. the HIV clinic) within the same facility. The providers suspected that she
might not be well prepared psychologically to deal with the diagnosis and that she could opt out of the
service. Thus, one nurse personally escorted her from the VCT room to the CTC section.
At the CTC section she was given a CTC client retained card (CTC card) with a unique identification
number. Then a patient folder (CTC folder) was created for her, with her name and the unique CTC number written on top of the folder. The folder was stored in the facility’s archive. The folder contained
Joyce’s residential tracking form with contact information such as her treatment supporter name, community health workers name and phone contact, residential area, household leader, facility name and district
name. Over time, whenever Joyce attended her visits, information about specific HIV care provided was
recorded chronologically on the CTC2 patient record form, and doctors’ notes about diagnosis and treatment (usually written on plain paper) were added in the CTC folder. In the folders, the clients would have
multiple CTC2 forms and it is likely the clients will have multiple client retained cards since HIV requires
life long service provision. This folder stored also results from lab test that were prescribed when required, for instance TB or malaria tests (one paper sheet per test conducted).
The CTC folder is a central artefact for coordination on various levels. First, some information is used
for coordinating activities across services: some forms are used to record– and thus make visible –
information about services from other sections involved in treating the client. For example, in the CTC2
form client’s status with respect to pregnancy and TB services are recorded. Second, the patient folders’
information is used for tracing defaulters (i.e. clients who do not show up for their appointments): the
folders contain forms that trace attendance of clients in the CTC clinic and they can be browsed in order
to identify clients who had missed their appointment. Once the clients are identified, the folders can be
sorted according to reasons for missed appointments.
Fig. 1. Example of patients' folders grouped according to their status.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
8
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
For instance, figure 1 shows how patient folders are stored on a shelf in one of HIV clinic section. The
folders are organized in three groups: deceased, transferred, and missed appointments (called “lost to
follow up”). This sorting of the folders makes visible the load of patients not coming to appointments and
that need to be traced. Nurses can trace them by using the contact information form. They may call mobile phones or ask the community HBC provider to contact them.
Third, information contained in the clients’ folders is also used as a resource for reporting both locally
and across levels. For instance, aggregated reports are created on quarterly basis and sent to district coordinators for monitoring, evaluation and planning purposes. In addition, facility reports are used locally for
coordinating health service provision within the facility. For instance Figure 2 shows a report with the
overview on clients’ appointments and registered attendance in one facility. The local reports are used by
facility management for getting overview of defaulter tracing activities and coordinating appointment and
attendance of clients.
Fig. 2. Sample page of a report used locally in one facility
The above description illustrates how the patient folder functions as a coordinating artefact for treatment,
as well as a resource for tracking and reporting. The description also shows how providers may need to
engage in additional coordination work in the form of extra activities to take into account contextual issues and support the collaboration between services. For instance, Joyce was not sent alone with the referral form to the CTC section, rather a nurse chose to escort her so that she should not abscond. The act of
escorting implies extra use of resources: the nurse escorts the client instead of attending clients sitting in
the waiting area.
A different trajectory to CTC services was that of Prisca. Prisca is a client who attended facility S for
her antenatal care. She was diagnosed HIV positive and since S facility did not host a CTC unit, she was
referred to R for CTC care (5 kilometers away). She was given a referral form. However, Prisca did not
accept to be labeled as HIV positive. She stopped going to antenatal care at S and she did not go to R.
Unfortunately, her health condition worsened and she decided to seek medical services. Remembering her
case at S, Prisca decided to go to R to get treated but acting as a new client, not referred from S. The doctor on duty suspected that she might have HIV, and decided to conduct provider-initiated HIV testing.
Prisca then confessed that she had been diagnosed positive earlier at S antenatal care. Through more questions, doctor learned that Prisca had torn the referral form. He further counseled and directed her to CTC
unit escorted by a nurse. However, Prisca informed the providers that she would have preferred a distant
CTC where she would not meet familiar people. Unfortunately, in that area like in most rural area there
was only one CTC and it meant that using other CTC would cost more in transport and time. Later Prisca
attended PMTCT unit at S facility after providers at R linked her with provider at S.
This vignette illustrates not only the challenge of tracking patients due to the distributed and fragmented organization of the service, but also how service delivery relies on clients as carriers of information.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
9
Clients take actively part in the coordination of their care trajectories by carrying referral forms between
facilities. However, if they do not comply with their schedules and treatments they also hinder coordination practices.
4.2
Pregnancy: from Care and Treatment Clinics (CTC) to Antenatal care and PMTCT
In 2009 Joyce, who had two children, became pregnant again. During one of her visits to CTC in the D
facility, the provider realized that she was pregnant and inquired about her adherence to the antenatal care
program. The provider learned that Joyce had not been to any antenatal checkups. The D facility had no
antenatal care unit, and he advised her to attend antenatal care at her residential area, facility H, in order
to facilitate frequent visits with low costs for transportation. The CTC provider then filled a referral form
and gave it to her so that she could bring it to the antenatal clinic in her residential area (H). The referral
form consisted of two parts on a single page. The first upper part was filled by the provider at CTC: the
doctor wrote which health services Joyce needed. He also recorded on the form that Joyce had been on
anti-retroviral therapy (ARV).
Afterwards Joyce went to H facility bringing with her the referral form and her CTC card. At H facility
she gave the form to the health provider at the antenatal care unit, who, by reading the information on the
form, enrolled her into the antenatal care (ANC) program. Enrollment consisted of creating a patientrecord in the antenatal register which was kept locally in the unit, and issuing an antenatal card which was
given to Joyce to keep it herself. Information from the referral form and CTC card were transcribed in the
antenatal register and card, and specifically, her CTC number and information about ARV use. Joyce was
also enrolled into the PMTCT program and her information registered into the PMTCT care register
which was kept in the PMTCT unit. The unique number from her CTC1 card was transcribed into her
record in the PMTCT care register. The alpha-numeric code generated by enrollment into PMTCT
(PMTCT code) was also noted on the antenatal card. The presence of the PMTCT code on this card will
indicate her positive HIV status to ANC staff. At the end of the visit, the health provider filled in the second part of the referral form and gave the form to Joyce to bring back to the CTC unit at the D facility.
Back at H facility, Joyce then had to undergo some tests related to antenatal care such as urine and
blood check ups. Facility H has no laboratory and their clients are usually sent to perform blood and urine
checks up at facility J. Joyce was given a laboratory request form and told to go to J facility or any nearby
hospital that had a laboratory for analysis. Joyce opted to go to J facility and took the tests. The results
were recorded on the laboratory form that Joyce passed back to facility H.
At one time during Joyce’s pregnancy, providers at the CTC clinic in facility D suspected her to have
tuberculosis (TB) symptoms. She was then referred, using a laboratory request form for TB testing, to
another facility which had laboratory and X-ray units for testing TB infection. Joyce went to the TB testing facility, the test was performed and she was diagnosed TB positive. She brought back the form with
her results to the CTC unit together with her X-ray picture. Providers at CTC then filled another referral
form for TB care, and gave it to her to take it to TB clinic in another facility (the only one with TB services in the area). She then went to the facility with TB services and presented the referral form to the TB
clinic providers. There at the TB clinic, she was enrolled into TB care. She was provided with TB1 card
(tuberculosis identification card) and a record on her name was created in the TB register. In the TB register was also recorded information about her HIV status, ARV drug dose and pregnancy status. As part of
the TB treatment regime, she had also to bring with her a treatment supporter who was given a treatment
supporter card for monitoring Joyce’s drug adherence during the treatment at home. Lastly, the care provider at the TB clinic completed the second part of a referral form which Joyce took back to the CTC
clinic in facility D.
The above description illustrates how Joyce moves between facilities and health programs and related
services. For instance when the practitioner saw that Joyce was pregnant, he started the process for enrolling her in both antenatal services and PMTCT services. First of all he interrogated Joyce about her status
and her attendance to antenatal services. Then, he used the referral form to inform the antenatal service
providers of the need to enroll Joyce, and he used Joyce to carry the form to its intended receiver. Also,
clients take actively part in the coordination of their care trajectories by carrying referral forms and laboratory results between facilities. Both referral and laboratory forms are examples of inter-facility coordination artifacts that are used to coordinate service provision across multiple facilities. These forms are particularly important in a context where facilities do not provide all services and do not frequently have
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
10
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
laboratories for analysis. For example, a referral may be sent from PMTCT to CTC or from CTC to TB
unit. These forms provide visibility of the services which has been done from the referring units and what
needs to be done in the recipient unit. In addition, the forms provide a feedback channel. The second part
of the form, which clients return to the referring facility, makes the provider aware that visits and tests
have been performed in the recipient units. Thus, some information from the referral feedback form is
used to update the client status, for instance regarding pregnancy or TB, in the facility register at the referring unit. Importantly, we saw the crucial role of Joyce for enabling the actual flow of these coordination artefacts. The client’s compliance with the requirements and directives are fundamental if the coordinative artefacts shall work.
4.3
From Antenatal to Delivery to Postnatal Care and Enrolling the Infant in the PMTCT and
Vaccination Programs
Antenatal care is scheduled with four visits during the course of a pregnancy. Whenever Joyce attended
antenatal care at H facility, providers were alerted of her HIV status through the PMTCT code written on
her antenatal card and they would remind her to attend the PMTCT clinic, located in the same facility.
Also, whenever Joyce attended PMTCT services, the provider would inquire about her attendance to CTC
at facility D. And whenever she attended visits at the CTC unit in facility D, the care provider by reading
her pregnancy status from CTC2 form in the patient folder, would inquire about her attendance to antenatal clinic at H clinic. At one time, the care provider even asked Joyce to show her antenatal card in order
to ascertain that she was attending antenatal visits as required.
At Facility H, the Reproductive and Child Health (RCH) service had no delivery unit. Thus, Joyce had
to deliver at facility E. The nurse at E facility, by recognizing her PMTCT code on her antenatal card
asked about ARV drug treatment. Joyce showed her the CTC card and the nurse read on her CTC card
about her visits and ARV drug administration. The nurse concluded that according to the Tanzanian
PMTCT protocol, Joyce would not need to take ARV during labor since she was already on ARV treatment. Joyce delivered safely a baby boy who was provided with ARV syrup. The nurse recorded into the
delivery register the following information (but not limited to): ARVs dispensed during labor, infant’s
ARV doses, newborn feeding practice, and linkage to CTC. Joyce’s antenatal card which had a section to
record information about the delivery and child conditions was filled in. Finally, the nurse filled a referral
form with information about the delivery and gave it to Joyce to bring it back to the CTC unit at facility
D.
After the delivery, Joyce attended postnatal care at her residential facility H. The nurses used her antenatal card to copy information and fill her postnatal register including her and child’s PMTCT and delivery information. Following the protocol, after six weeks the child was enrolled into the child health clinic
in facility H. The child was registered into the children register and issued with a child health card. The
providers on duty after recognizing PMTCT code in Joyce’s ANC card also filled information related to
PMTCT services to children. The child was also enrolled into PMTCT child follow-up health services
and a record created in the PMTCT child follow-up register. Some information from Joyce’s ANC card
delivery section was transcribed into the PMTCT register. The child health card also contained a code
(generated in children clinic) which communicate that the child is in PMTCT services (HIV exposed
child).
The services of the children health clinic are offered for five years and include services such as vaccination, vitamin A, growth monitoring, and mosquito net distribution. Whenever Joyce attended the clinic as scheduled, providers by seeing the child’s PMTCT code would remind her to attend also the
PMTCT clinic. In PMTCT services, infants are tested for HIV infection at age of 4-6 weeks, 9 month and
18 month. Children are supposed also to continue with ARV drugs during breastfeeding time. In the
PMTCT room, providers would also enquire about Joyce’s attendance in CTC clinic. Sometimes they
would ask her to show her CTC card so that nurses could verify what Joyce told them.
Joyce’s baby was tested and diagnosed HIV negative at the first test. The PMTCT services were also
challenged by resource shortage including shortage of referral forms which is one of coordinative artifacts. However, during the time when the second test was supposed to be performed for Joyce’s baby at
facility H, there was neither HIV testing kits nor ARV drugs. This implied that Joyce had to be referred to
another facility for her treatment. This was combined with shortage of referral forms. The breakdown was
resolved by the nurse who instructed Joyce to tell the provider what services she needed at the other facil© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
11
ity. Thus, Joyce went to the other facility where she was attended, and brought back the feedback to facility H.
The description above illustrates how providers articulate the collaborative work through various coordinative artifacts (e.g. PMTCT codes and CTC cards). The health workers practice of interrogating and
reminding is an example of how coordinative work can be supported through coordinative artifacts. For
example, PMTCT codes provide awareness of the HIV status and the linkage with CTC and PMTCT
services. The patient card offers an overview of activities that have to be performed or that have already
been performed in either this provider’s section or in another section. The PMTCT code also facilitated
nurses’ awareness of tasks to be done in their location, e.g. in the delivery ward, and trigger coordinative
practices such as reminding or inquiring clients. We also see how PMTCT information is recorded into
different facilities’ registers: antenatal, delivery, postnatal and children. Information recording in registers
provides an overview of clients’ usage of other health services: nurses can know that clients are attending
PMTCT services or vaccination/antenatal services by using facility registers.
A different trajectory about reminding practices was that of MdalaMatonya. MdalaMatonya is a mother of twin boys who are PMTCT clients. She was diagnosed HIV positive during her pregnancy in the
antenatal section of facility P and was enrolled into their PMTCT program. However, facility P offered
limited PMTCT services due to lack of training to providers and does not have an HIV clinic, hence referring their clients to facility Q. MdalaMatonya was referred to Q and was enrolled into their HIV clinic,
meanwhile she was attending PMTCT services at antenatal section of Q. After delivery, her boys were
enrolled into the children clinic program in facility P and into PMTCT program in facility Q. In the meantime since pregnancy, MdalaMatonya had been attending facility Q for CTC services. One day when
MdalaMatonya was attending CTC, a nurse (who also was responsible for PMTCT) inquired about the
boys’ attendance in PMTCT services. The nurse learned that the boys had not done a third HIV test and
instructed the mother to bring the boys to PMTCT services the next day. The boys’ children’s card
showed that they had attended all children visits in facility P. However, at the age of 20 months the boys
had had only two HIV tests, while according to the PMTCT protocol they should have had three HIV
tests. MdalaMatonya explained that no one had told her to bring the children for a third test. Through our
studies in four children clinics (including facility P) and HIV clinics, we learned that nurses might forget
to remind clients about their visits and tests due to high workload, and this may create breakdowns.
4.4
Enrolling the Infants in the HIV Program
Joyce’s baby was due for a third test and was diagnosed HIV negative. Because by the third test Joyce’s
child was 18 month, the nurse recorded ‘discharged’ on the remark column in the register. This meant that
the child would no longer need to visit PMTCT. The nurse also informed Joyce that whenever nurses in
children’s clinic would ask her about PMTCT service, she was supposed to inform them that the child had
been discharged from PMTCT services and that he resulted negative in all three tests. Nurses at the children’s clinic used the child card to record the results.
The different trajectory was observed on the same day where another child (Baraka) of nine months
tested HIV positive. This information was recorded in the child register and in the PMTCT child followup register. In addition, the nurse filled an infant referral form from PMTCT to CTC services. In the same
town area there were four CTC units whereby one was just half a kilometer from H facility. Nurses advised the mother of the child who tested HIV positive to attend CTC in the clinic nearby. In this way she
could conveniently combine children and CTC clinics on the same day. However, the mother disapproved
of using the nearby facility since she would probably meet people she may be familiar with. She preferred
to attend a more distant facility where she would also receive her CTC services. She proposed to ask the
provider to combine her and child CTC clinics dates.
However, there was a problem regarding the infant referral form due to a mismatch between its design
and existing work practices. The infant referral form had no feedback part, such as what the general referral form had. Figure 3 shows a form with a feedback part (on the left), and the infant referral form without
a feedback part (on the right).
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
12
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
Fig. 3. General HIV form (left) and infant referral form (right).
The infant referral form has specific information about PMTCT services. The absence of a second part for
feedback was not supporting collaborative practices. Despite the form not supporting it, feedback was
assumed to happen; the registers had space for recording CTC number which would make visible that a
client had been enrolled into HIV care. In Baraka’s case, the providers told the mother to come back with
the CTC card so that information could be recorded in the register. The mother passed the infant referral
form to the CTC unit and she came back with the child CTC card. Information such as CTC unique number was transcribed into the PMTCT child follow up register. From that time nurses in PMTCT were also
inquiring about child attendance in CTC unit by seeing her records in PMTCT register.
As a result of this lack of feedback section in children’s referral form, in one urban facility with high
number of clients, new information practices emerged. The health workers started to record all referred
clients and their enrollment into the HIV/AIDS clinic even though these were recorded on papers. One
health worker explained: “Some mothers do not come back to ascertain that they have joined the
HIV/AIDS clinic. We have started to record their names so that we can make a follow up of their enrollment to HIV/AIDS clinic”. Figure 4 shows a page of referred clients in one register.
Fig. 4. A local registry to support tracking of children from PMTCT to the HIV clinic.
In other two facilities (located in the outskirts of the town), the practice was to combine in one register
information about home based care services (for HIV/AIDS clients) and PMTCT information so that all
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
13
children clients born from HIV positive mothers would be tracked in the community across time, along
schedules of PMTCT services in case they do not show up to the original facility.
5
Discussion
In the previous sections we have described several trajectories of HIV-positive pregnant women and how
information is collected, organized, stored and eventually retrieved by health providers. In the different
stages of the trajectories, whenever a woman contacts or is contacted by a health provider, information is
produced and correlated with previous collected information. For health providers is critical to be able to
track women and monitor how their pregnancies and their HIV conditions are developing, and how well
they comply with the PMTCT program. For instance, health providers check if women are attending their
counseling sessions and periodical visits, if they are regularly tested, and if they are taking the prescribed
drugs.
We have also described challenges in tracking women along their trajectories. Specifically we have focused on two main challenges. First, we have described how women visit several facilities when attending
PMTCT services. This happens for different reasons. For instance in the case of Prisca, her home facility
did not have a CTC unit and she was referred to another facility 5 kilometers away. In the case of Joyce
her home facility did not have an antenatal care unit and again she was referred to another facility. In
another case, MdalaMatanoya was referred to another facility because her home facility did not have a
HIV clinic and trained personnel. Second, we have described how women receive care services from
different programs. It is rare, especially in rural areas, that care facilities provide comprehensive services
for CTC, TB care, antenatal and postnatal care, vaccinations and dispensation of ARV drugs. A pregnant
woman who is tested HIV positive would need to be enrolled in both PMTCT and ANC, and perhaps
other programs, such as TB. For instance Joyce was already enrolled in PMTCT services but needed also
TB care. Her case illustrates how the work of tracking women’s trajectories is usually performed within
programs, and that there is little coordination across programs. In addition once the child is born he would
also need care which belongs yet to another care program that may be offered in few facilities. This implies that women’s information gets distributed across facilities and programs, and tracking becomes
difficult. In this situation, health providers need to keep an overview of how services are interrelated, and
where are they offered. For instance when possible they escort women to other parts of the facility offering another service. The case of Joyce describes how she was escorted from the VCT to the CTC section.
Or they also give the option to choose a facility that is closer to the home area of the woman. In addition,
women themselves collaborate in coordinating and supporting the communication across facilities and
programs. As our analysis illustrates women carry referral forms, test results, and personal cards with
them. They participate actively in the coordination of their own trajectories.
Our analysis also shows the different artefacts used to manage information in PMCTC services. Codes
and unique patient numbers are important coordination artefacts that identify women and their diagnosis,
and are copied from cards to referral forms to registers. For instance when Joyce went to delivery unit, the
nurse was alerted by her PMTCT code on her card and inquired about her ARV drugs treatment. Other
artefacts are referral letters which are used for ordering exams and visits in other facilities or units. These
forms may also have a feedback part that is brought back to the health provider. For instance when Joyce
was referred to a facility for antenatal care, she brought back the second half of the referral to her home
facility. However, while artefacts are usually designed for tracking patients within programs and within a
single facility, the design and use of cross-facility and cross-program artefacts meets challenges.
We have also described how breakdowns may happen and health providers may lose track of their clients. Women may intentionally not collaborate. For example, Prisca tore the referral form she was supposed to bring to a facility for PMTCT care. In another example Baraka’s referral form did not have a
feedback part. Thus the health provider had to ask the mother to come back again with the CTC card in
order to register her CTC code and relevant information into the local register.
In the following section we discuss two key set of findings from our analysis of the case. First we discuss the complexities of coordinating work in a resource constrained setting and PMTCT care, and then
we formulate implications for the design of patient information systems.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
14
5.1
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
Complexities of Coordinating Work
As Schmidt states, people engage in cooperative work when they are mutually dependent in their tasks
and therefore are required to cooperate in order to get the work done (Schmidt, 1991). In our study we
have shown the collaboration required when managing patient trajectories of PMTCT patients across
facilities and programs, and how contingencies may change the course of a trajectory. In the descriptions
reported in our study it is evident that coordinating work across time (along clients’ trajectories), across
facilities (different locations), and across health programs (type of services) is crucial to PMTCT service
delivery. However, patient trajectories are emergent and contingent, and in a resource scarce setting like
in ordinary health provision in developing countries, coordination is challenging and prone to breakdowns. Furthermore, the nature of the disease contributes to the complexity of care delivery in PMTCT
services, and to forming more intricate task interdependencies. HIV is a disease burdened with social
stigma and some clients would rather travel to a distant facility than receiving care from their home health
services. In addition other diseases are likely to develop, and requiring additional care, and the involvement of additional health providers. Thus, in our study we show how the complexity of cooperative work
is increased not only by involving additional stakeholders from different health programs (for instance
specialists in TB care), but also by stretching the trajectory to other facilities where clients would not be
recognized. While the literature recognizes that the involvement of multiple stakeholders and diverse
professional groups increases the complexity of cooperative work (Fitzpatrick & Ellingsen, 2013; Hartswood, Procter, Rouncefield, & Slack, 2003), and has examined the complexity of distributed settings (e.g.
Hinds & McGrath, 2006; Olson & Olson, 2000), it has not addressed the challenge of unbounded distributed cooperative work. For instance the case of Baraka’s mother, being in an urban setting, shows how
clients can select one HIV clinic among many. However, the case shows also how it was complicated for
the health provider to coordinate Baraka’s mother’s connection with her home HIV facility.
Coordination work relies on and it is mediated by artifacts and protocols (Bardram, 1998; Berg, 1999a,
1999b; Reddy et al., 2010). Our findings are in line with research showing the shortcomings of coordinative artifacts in complex healthcare settings(Cabitza et al., 2006; Gulube & Wynchank, 2001).In cases
when coordinative artefacts fail to support cooperation, this might lead to the development of informal
artifacts (Mellini, 2013). Differently, in our case additional coordinative work was needed due to the
scarcity of resources (e.g. absence of referral forms, of test kit, of labs). There are cases when coordinative artifacts are just not available, and health providers have to come up with ways to deal with the situation. For example, in Joyce’s case the health provider sent her to another facility for PMTCT testing services with oral instructions but without referral letter. Many studies point to similar challenges when discussing collaboration in developing countries (e.g. Geissbuhler et al., 2003).
5.2
Implications for Design
In this study we have foregrounded the complexities of collaboration and coordination. In this section we
argue that based on our study we can draw important implications for the design of paper based and digital artefacts, and ultimately for improving health care provision. Specifically, we argue that some coordination breakdowns that we have observed could be avoided by improving those artefacts that are used in
PMTCT health delivery.
First, the design of patient information systems should consider how patient artifacts support coordination work between different health programs. For instance, it is important that referral forms have a feedback part in order to support a two way communication between health providers and to ensure visibility
between providers in involved facilities. Second, coordination work should be supported by improving
coordination protocols and procedures. For instance, reminding practices should be enforced both within
facilities and across health programs. This could be procedures where clients are required to show their
different health services patients’ cards where attendance is registered. Third, in the case of PMTCT services, we have showed how the coordination mechanism of clients carrying information is vulnerable. In
this case the coordination breakdowns observed in RCH, can not only be resolved through redesign of
patient information systems tools (e.g. infant referral forms) or enforcing protocols (e.g. reminding practices for health workers). We suggest the need for creating complimentary coordination mechanisms
based on a computerized patient system supporting horizontal coordination between health facilities. For
example, the design of patient information systems would need to afford visibility and communication
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
15
between groups involved in treating clients. In the case of the HISP program, the DHIS tracker, which is
patient-oriented could facilitate users’ communications through mailing lists i.e. queries for patients are
sent to a list of all involved health workers in PMTCT care in different health facilities. The DHIS tracker
displays (in a list) the email addresses of other users in other facilities so that a worker can find the address of someone and write to ask if the patient has come or not. The same strategy can be implemented at
the district and regional levels where health facilities communicate to each other about clients’ status on
attending maternal and child healthcare.
6
Conclusion
Our study has addressed the following research question: what are the challenges to coordination in collaborative practices and what are the implications for the design of patient information systems as coordinative artefacts? Taking a CSCW approach we have described and analyzed several challenges. We have
provided a rich empirical description of coordination work in a resource constrained setting and we have
proposed improvements to the design of both computer and paper-based information systems. Our findings have contributed to understanding the coordination work to facilitate collaboration around health
care provision. Future research could investigate more concrete solutions to the needs for appropriate
coordination mechanisms through participatory design with the end users, in order to generate wellworking solutions to the very practical and significant challenges we have started to illuminate in this
paper.
Acknowledgment.
We thank: health workers, clients, and other informants, the Global Infrastructure research group at the
University of Oslo, Eric Monteiro and the anonymous reviewers for their constructive comments.
References
Bagayoko, C., Anne, A., Geissbuhler, A., & Fieschi, M. (2008). Can ICTs Contribute to the Efficiency and provide
Equitable Access to the Health Care System in Sub-Saharan Africa? Thromb Haemost, 99(6), 1112-1115.
Bardram, J. (1998). Designing for the dynamics of cooperative work activities. Paper presented at the Proceedings of
the 1998 ACM conference on Computer supported cooperative work.
Bath, G. (2006). Telemedicine in Africa. Journal of the South African Veterinary Association, 77(3), p. 107-p. 107.
Berg, M. (1999a). Accumulating and coordinating: occasions for information technologies in medical work.
Computer Supported Cooperative Work (CSCW), 8(4), 373-401.
Berg, M. (1999b). Patient care information systems and health care work: a sociotechnical approach. International
journal of medical informatics, 55(2), 87-101.
Bjørn, P., & Hertzum, M. (2011). Artefactual multiplicity: A study of emergency-department whiteboards. Computer
Supported Cooperative Work (CSCW), 20(1-2), 93-121.
Blomberg, J., & Karasti, H. (2013). Reflections on 25 years of ethnography in CSCW. Computer Supported
Cooperative Work (CSCW), 22(4-6), 373-423.
Braa, J., Monteiro, E., & Sahay, S. (2004). Networks of action: sustainable health information systems across
developing countries. Mis Quarterly, 337-362.
Bryman, A. (2012). Social research methods: Oxford university press.
Cabitza, F., Sarini, M., Simone, C., & Telaro, M. (2006). Torres, a conceptual framework for articulation work across
boundaries. Frontiers in Artificial Intelligence and Applications, 137, 102.
Cheng, K. G., Ernesto, F., & Truong, K. N. (2008). Participant and interviewer attitudes toward handheld computers
in the context of HIV/AIDS programs in sub-Saharan Africa. Paper presented at the Proceedings of the
SIGCHI Conference on Human Factors in Computing Systems.
DeRenzi, B., Lesh, N., Parikh, T., Sims, C., Maokla, W., Chemba, M., . . . Borriello, G. (2008). E-IMCI: Improving
pediatric health care in low-income countries. Paper presented at the Proceedings of the SIGCHI
conference on human factors in computing systems.
Engestrom, Y., Engestrom, R., & Saarelma, O. (1988). Computerized medical records, production pressure and
compartmentalization in the work activity of health center physicians. Paper presented at the Proceedings of
the 1988 ACM conference on Computer-supported cooperative work.
Fagerhaugh, S. Y., & Strauss, A. L. (1997). Social organization of medical work: Transaction Publishers.
Fitzpatrick, G. (2004). Integrated care and the working record. Health Informatics Journal, 10(4), 291-302.
Fitzpatrick, G., & Ellingsen, G. (2013). A review of 25 years of CSCW research in healthcare: contributions,
challenges and future agendas. Computer Supported Cooperative Work (CSCW), 22(4-6), 609-665.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
16
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of Mother to Child
Transmission of HIV in Tanzania
Fraser, H. S., Allen, C., Bailey, C., Douglas, G., Shin, S., & Blaya, J. (2007). Information systems for patient followup and chronic management of HIV and tuberculosis: a life-saving technology in resource-poor areas.
Journal of medical Internet research, 9(4).
Garfinkel, H., & Bittner, E. (1967). Good organizational reasons for “bad” clinic records. Englewood Cliffs.
Gasser, L. (1986). The integration of computing and routine work. ACM Transactions on Information Systems
(TOIS), 4(3), 205-225.
Geissbuhler, A., Ly, O., Lovis, C., & L’Haire, J.-F. (2003). Telemedicine in Western Africa: lessons learned from a
pilot project in Mali, perspectives and recommendations. Paper presented at the AMIA Annual Symposium
Proceedings.
Gizaw, A., Mukherjee, A., Lewis, J., & Sahay, S. (2012). Understanding design-use interaction: the case of name
based health information system in India. Paper presented at the Proceedings of IADIS International
Conference on Information Systems.
Gulube, S., & Wynchank, S. (2001). Telemedicine in South Africa: success or failure? Journal of telemedicine and
telecare, 7(suppl 2), 47-49.
Hartswood, M., Procter, R., Rouncefield, M., & Slack, R. (2003). Making a case in medical work: implications for
the electronic medical record. Computer Supported Cooperative Work (CSCW), 12(3), 241-266.
Heath, C., & Luff, P. (1996). Documents and professional practice:“bad” organisational reasons for “good”
clinical records. Paper presented at the Proceedings of the 1996 ACM conference on Computer supported
cooperative work.
Hinds, P., & McGrath, C. (2006). Structures that work: social structure, work structure and coordination ease in
geographically distributed teams. Paper presented at the Proceedings of the 2006 20th anniversary
conference on Computer supported cooperative work.
Jones, M. (2003). " Computers can land people on Mars, why can't they get them to work in a hospital?"
Implementation of an Electronic Patient Record System in a UK Hospital. Methods of information in
medicine, 42(4), 410-415.
Kamadjeu, R. M., Tapang, E. M., & Moluh, R. N. (2005). Designing and implementing an electronic health record
system in primary care practice in sub-Saharan Africa: a case study from Cameroon. Informatics in primary
care, 13(3), 179-186.
Korpela, M., Mursu, A., & Soriyan, H. A. (2002). Information systems development as an activity. Computer
Supported Cooperative Work (CSCW), 11(1-2), 111-128.
Luk, R., Ho, M., & Aoki, P. M. (2008). Asynchronous remote medical consultation for Ghana. Paper presented at the
Proceedings of the SIGCHI conference on human factors in computing systems.
Madden, R. (2010). Being ethnographic: A guide to the theory and practice of ethnography: Sage Publications.
Marcus, G. E. (1995). Ethnography in/of the world system: the emergence of multi-sited ethnography. Annual review
of anthropology, 95-117.
Mars, M. (2010). Health capacity development through telemedicine in Africa. Yearb Med Inform, 87, 93.
Martínez-Alcalá, C. I., Muñoz, M., & Monguet-Fierro, J. (2013). Design and customization of telemedicine systems.
Computational and mathematical methods in medicine, 2013.
Mayo-Smith, M. F., & Agrawal, A. (2007). Factors associated with improved completion of computerized clinical
reminders across a large healthcare system. International journal of medical informatics, 76(10), 710-716.
Mellini, B. (2013). The Materialization of Articulation Work and its Implications on Nursing Practices. Paper
presented at the CHItaly (Doctoral Consortium).
Mønsted, T., Reddy, M. C., & Bansler, J. P. (2011). The use of narratives in medical work: A field study of physicianpatient consultations. Paper presented at the ECSCW 2011: Proceedings of the 12th European Conference
on Computer Supported Cooperative Work, 24-28 September 2011, Aarhus Denmark.
Olson, G. M., & Olson, J. S. (2000). Distance matters. Human-computer interaction, 15(2), 139-178.
Oluoch, T., Santas, X., Kwaro, D., Were, M., Biondich, P., Bailey, C., . . . de Keizer, N. (2012). The effect of
electronic medical record-based clinical decision support on HIV care in resource-constrained settings: A
systematic review. International journal of medical informatics, 81(10), e83-e92.
Patterson, E. S., Doebbeling, B. N., Fung, C. H., Militello, L., Anders, S., & Asch, S. M. (2005). Identifying barriers
to the effective use of clinical reminders: bootstrapping multiple methods. Journal of biomedical
informatics, 38(3), 189-199.
Reddy, M., Bardram, J., & Gorman, P. (2010). CSCW research in healthcare: Past, present, and future. Paper
presented at the Proceedings of the Conference on Computer Supported Cooperative Work. Retrieved from
http://sites.google. com/site/cscwinhc.
Robertson, A., Cresswell, K., Takian, A., Petrakaki, D., Crowe, S., Cornford, T., . . . Jacklin, A. (2010).
Implementation and adoption of nationwide electronic health records in secondary care in England:
qualitative analysis of interim results from a prospective national evaluation. BMJ, 341.
Rotich, J. K., Hannan, T. J., Smith, F. E., Bii, J., Odero, W. W., Vu, N., . . . Tierney, W. M. (2003). Installing and
implementing a computer-based patient record system in sub-Saharan Africa: the Mosoriot Medical Record
System. Journal of the American Medical Informatics Association, 10(4), 295-303.
Safadi, H., & Faraj, S. (2010). The role of workarounds during an opensource electronic medical record system
implementation.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. DOI: 10.12856/JHIA-2014-v2-i1-88
Shidende et al. / Coordination Challenges in Collaborative Practices in the Prevention of
Mother to Child Transmission of HIV in Tanzania
17
Saugene, Z. (2013). Customization of Generic Open Source Software for Health Sector in Developing Countries.
University of Oslo, Norway.
Saugene, Z., & Kaasbøll, J. (2013). Foregrounding invisible work practices: the customization of generic software
systems.
Schmidt, K. (1991). Riding a tiger, or computer supported cooperative work. Paper presented at the Proceedings of
the Second European Conference on Computer-Supported Cooperative Work ECSCW’91.
Schmidt, K., & Bannon, L. (1992). Taking CSCW seriously. Computer Supported Cooperative Work (CSCW), 1(12), 7-40.
Schmidt, K., & Simonee, C. (1996). Coordination mechanisms: Towards a conceptual foundation of CSCW systems
design. Computer Supported Cooperative Work (CSCW), 5(2-3), 155-200.
Schmidt, K., Wagner, I., & Tolar, M. (2007). Permutations of cooperative work practices: a study of two oncology
clinics. Paper presented at the Proceedings of the 2007 international ACM conference on Supporting group
work.
Simone, C., & Schmidt, K. (1998). Taking the distributed nature of cooperative work seriously. Paper presented at
the 16th Euromicro Conference on Parallel, Distributed and Network-Based Processing (PDP 2008).
Vikkelsø, S. (2005). Subtle redistribution of work, attention and risks: Electronic patient records and organisational
consequences. Scandinavian Journal of Information Systems, 17(1), 10.
Walsham, G. (1995). Interpretive case studies in IS research: nature and method. European Journal of information
systems, 4(2), 74-81.
© 2014 JHIA. This is an Open Access article published online by JHIA and distributed under the terms of the Creative Commons Attribution NonCommercial License. 10.12856/JHIA-2014-v2-i1-88
Appendix 2: Paper II
Shidende, N.H. (2014). Challenges in Implementing Patient-Centred Information Systems in
Tanzania: An Activity Theory Perspective. The Electronic Journal of Information Systems in
Developing Countries, 64.
II
EJISDC (2014) 64, 2, 1-20
1
CHALLENGES IN IMPLEMENTING PATIENT-CENTRED INFORMATION SYSTEMS
IN TANZANIA: AN ACTIVITY THEORY PERSPECTIVE
Nima Herman Shidende
University of Oslo, Norway
[email protected]
ABSTRACT
Defaulter tracing systems are patient centred information systems which are used to enhance
monitoring of patients who have dropped out of health programs. The aim of this study is to
explore the operation of defaulter tracing systems in practice. The study has been undertaken
in Tanzanian health facilities by using qualitative data collection methods. Activity theory is
employed as a conceptual framework to analyze the findings. Findings show that the
implementation of defaulter tracing system is challenged by issues of the existing information
systems, such as absence of clear guidelines and tools, and information recording
incompleteness, as well as contextual issues. The paper contributes to Healthcare Information
System literature (HIS) which has elaborated how patient centred information systems are
utilized in practice at the level of primary health care. This study concludes by providing
implications for policy, practice, design and implementation.
Keywords: Health Information Systems, Patient-centered Information Systems, Defaulter
Tracing System, Activity Theory, Maternal and Child Healthcare
1.
INTRODUCTION
Defaulter tracing systems can be described as procedures and tools that can enhance
monitoring of patients who have dropped out of health programs. There are several reasons
which can make clients opt out of the health programs, including social stigma attached to the
HIV disease, the need to prioritize income-generating work, and too high transport costs
(Cutts, 1991; Fraser et al., 2007). Defaulting health programs is not beneficial for the patients
themselves, and it also undermines WHO (World Health Organization) and national health
authorities’ initiatives and efforts to improve the health status and reduce morbidity and
mortality rates. To counter the defaulting problem, defaulter tracing systems become
important in healthcare provision (ibid).
Defaulter tracing systems are common in HIV related programs (see for example,
Fraser et al., 2007; Mwatawala et al., 2012). These systems mainly support two tasks:
identifying and tracing defaulters. Thomson et al. (2011) present an example of a defaulter
tracing system in a HIV/TB program with the following steps: patient is enrolled into the
program; patient misses the scheduled appointment; patient is identified as a potential
defaulter; health worker traces patient, health worker documents outcome and enters data in
the computer system which could be statistically analyzed. In the literature, several
approaches for tracing defaulters are suggested, including using CHW (Community Health
Workers), home visits, phone calls, written reminders, collaboration with community-based
organizations and outreach teams (Cutts, 1991; Marsh, 2000; Fraser et al., 2007; Forster et al.,
2008; Thomson et al., 2011; Chadambuka et al., 2012).
There are also other strategies for addressing client defaulting problem. One approach
is to improve clients attendance in health programs through informing the whole population
about health activities, e.g. immunization activities, by disseminating messages through
schools, religious organization and influential leaders (Cutts, 1991; Chadambuka et al., 2012).
Another approach is to use both facility and community based approaches (Busza et al., 2012;
Marcos et al., 2012). Busza et al. (2012, p.3) define community based approaches to PMTCT
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
2
(Prevention of Mother to Child Transmission of HIV) as “strategies and interventions to
improve health behavior and outcome that are delivered outside the formal health settings”.
The community groups that could be involved are diverse. For example CHW, peer
counselors, volunteers, mentor mothers, traditional chiefs and religious leaders could be
involved (Marcos et al., 2012).
Another strategy to support defaulter tracing activities is to use an information system
which will keep records of different health services provided, provide continual updates about
treatment status and track all patients within health programs (Fraser et al., 2007;
Ghadrshenas et al., 2013). However, standardized information artefacts and adequate training
for health workers are needed, in order to enable healthcare workers to collect accurate,
timely and consistent data (Ghadrshenas et al., 2013).
The above mentioned studies illustrate some experiences of how defaulter tracing
systems are expected to work. However, we do not know much about how they are utilized in
practice and their challenges. Therefore, the aim of this study is to explore the operation of
defaulter tracing systems in practice. By practice, I mean how people act in organizational
context when they are performing their daily work (Feldman and March, 2010; Nicolini,
2013). The study was conducted within the maternal and child health care program in
Tanzania. Activity theory was selected as the analytical framework because it is useful for
understanding practice (Kuutti, 1996; Nardi, 1996). With that insight, the following research
question is addressed: What are the challenges faced by defaulter tracing systems in practice?
This paper contributes to health information systems (HIS) literature around defaulter
tracing systems. Specifically, the paper contributes to HIS literature which has elaborated how
designed tools are utilized in practice at the level of primary health care of developing
countries’ context. In this literature there are few studies which have elaborated defaulter
tracing practices in maternal and child health settings (see for example DeRenzi et al., 2012;
MacLeod et al., 2012); since most studies on defaulter tracing practices are in vertical
programs such as tuberculosis, HIV/AIDS and PMTCT (see for example Jones et al., 2005;
Fraser et al., 2007; Nglazi et al., 2010; Busza et al., 2012; Thomson et al., 2011; Kalembo and
Zgambo, 2012; Marcos et al., 2012).
The structure of this paper is as follows: following the introduction, a literature review
and conceptual framework is presented in section 2. The literature review centers on studies
of health information systems in developing countries with a focus on the design of patient
centered information systems. Sections 3 and 4 present the research context and methodology
respectively. Section 5 presents the findings of the study, while sections 6 present discussion,
conclusion and study implications.
2.
LITERATURE REVIEW AND CONCEPTUAL FRAMEWORK
2.1
Patient Centered HIS in Developing Countries
Health facilities provide premises where curative, preventive, rehabilitative, promotive and
palliative health services are offered. Health facilities are also the source of patient data which
is collected during health service provision (WHO, 2008). At this level, information and
health service provision activities are intertwined. Patient centred information systems
whether paper based or digital, are designed at higher administrative levels such as province
or district (Puri et al., 2004) or at national levels (Chilundo et al., 2004; Gizaw et al., 2012).
It has been long recognized that HIS is crucial for effective health care and improved
health management can be attributed to improved HIS (Mahler, 1986 cited in Sauerborn and
Lippeveld, 2000). Unfortunately, there are problems related to the data generated from the
health facilities. The issue of poor data quality problem is among the most prominent issues
discussed in the HIS literature. The problem of poor data quality is manifested through
irrelevance of information gathered, poor use of information and lack of timely, correct and
comprehensive reporting (Sauerborn and Lippeveld, 2000; Simba and Mwangu, 2005). From
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
3
the literature, reasons for low data quality are attributed to issues like health workers’ being
unaware of health information system importance, lack of information systems skills, lack of
feedback from higher levels, and poor motivation among health workers (Lippeveld, 2000;
Heywood and Rohde, 2001; Sauerborn and Lippeveld, 2000; WHO, 2008).
Patient centered information systems literature of developing countries has been
centered on the following themes: design of computerized artifacts or implementation of
designed artifacts. HIS studies focusing on the design of artifacts have provided an
understanding on what functionality or information technologies could be implemented to
support healthcare delivery. In Malawi, a mobile phone-based system was developed to guide
HIV/AIDS counsellors in decision making and data capture, when consulting with clients
(Manda and Msosa, 2011). In Kenya, electronic medical record was implemented in rural
Kenya health centre for capturing data during patient visits in order to improve information
management (Rotich et al., 2003). Similar efforts have been made in Cameroon where locally
developed electronic health record system was implemented in urban primary health care
(Kamadjeu, 2005). Some of these studies have discussed design of artifacts for vulnerable
groups such as HIV and Tuberculosis in order to improve information management in the
respective health programs.
There are also studies which have provided an understanding of how to involve users
in design of artifacts. For example, Puri et al. (2004) describe the process of designing and
development of a child health community-based Information System in South African
context. They elaborate how the design process is shaped by aspects of socio-economic,
cultural and political situations. Elovaara et al. (2006) discuss the design of hospital based
information system in Zanzibar where the participation of different health workers was not
uniform and the design process was dominated by information technologists.
Many studies on computerized patient information systems implementation in
developing countries have reported on macro-level challenges such as on electricity,
communication, low literacy, data quality, technical aspect of EMR, system acceptance,
human resource and management issues, system training, inadequate budget and system
sustainability (see for example, Fraser et al., 2004; Kamadjeu et al., 2005; Rotich et al., 2003).
There is a paucity of HIS literature which discusses actual usage of designed artifacts by
primary health level workers (Kanjo, 2011; Ngoma et al., 2011; 2012) and its associated
challenges. For example, Kanjo (2011) described a situation where data from community is
not part of the official health information system in Malawi and she argues that dichotomy
between policy and practice have resulted in data not flowing from community level to higher
levels. Similarly, Ngoma et al. (2011) presented a model showing how the use of mobile
phones could improve the flow of maternal and child delivery information from community to
district level. Ngoma et al. (2012) show the challenges of designing computerized information
system due to differences between designed tools and work practices in the primary health
care facilities. These studies discuss utilization of paper based-information artefacts, with the
purpose of improving information flow between different levels. Fraser et al. (2007) asserts
that there are few studies which had evaluated the effects of designed and/or implemented IT
on patient outcomes.
The understanding of actual usage of patient centred information systems is crucial for
future design and implementation of computerized patient cantered information systems. The
decrease of computer hardware cost, increased computer awareness and literacy, and
availability of free and open source software in developing countries, could lead to
computerization of clinical level in both public and private health facilities as shown by
implementation examples above. The implementation of Electronic Medical Records (EMR)
have been challenging even in the developed world (Safadi and Faraj, 2010) and one of the
obstacles being the incompatibility between established work practice and the new
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
4
technology. Hence, the understanding of actual usage of designed tools might be useful for
future design of EMR. In the next section, activity theory is presented, as a framework to
analyze the use of designed patient centred artefacts.
2.2
Conceptual Framework
Activity Theory (AT) is a theoretical framework which can be used to understand practice and
also for informing the design process (Kaptelinin et al., 1995, Korpela, et al., 2001; Igira,
2008). For instance, Korpela et al. (2001) used AT to understand work practices and used the
insight to inform the design of the information system. Activity theory has been used in
information systems studies to understand the design process or its expansion (Cole, 1995;
Igira, 2008; Igira, 2012). In AT, the minimal unit of analysis is the activity system (Kuutti,
1996). Figure 1 shows a pictorial representation of the activity system.
Tools and Artifacts
Subject
Rules and Procedures
Object
Community
Outcome
Division of Labour
Figure 1: The structure of human activity system (Engeström, 2001)
The elements of an activity system consist of subject, tools/ artefacts, object, division
of labour, community and rules and procedures. The subject represents actor(s) who are
motivated by the object to produce the outcome of the activity which is under consideration.
The object can be a material or problem space, which the subject aims to accomplish by using
tools and artefacts. The tools can be material or conceptual, and are used to transform the
object into the outcome. The object of certain activity is also shared by people in the
community making up the activity system. The division of labour refers to explicit and
implicit allocation of tasks between community and subject elements so that the
transformation process of the object into the outcome can take place. The activity’s rules
cover both explicit and implicit norms, expectations, conventions, and social relations within
a community (Kuutti, 1996).
Relation between elements of activity systems is not direct but mediated (Kuutti, 1996,
p.26). For example, the rules and procedures mediate the relationship between the subject and
community elements. The systemic nature of activity system implies that if there is a change
in one element, it would cause changes in other elements as well (Nicolini, 2013). For
instance, if there is a change in community elements, it will affect the rules elements and
eventually, the instruments which mediate the relationship between subject and object.
Activity theorists understand all practices as being under continuous development,
since change is inevitable in activity systems (Kaptelinin et al., 1995). Change of activity
systems occurs through contradictions which manifest themselves as problems, disturbances
or inconsistencies between the components of an activity system (Engestrom, 2001; Issroff
and Scanlon, 2002). Contradictions point to misfits occurring within or between elements of a
single activity system (Kuutti, 1996). Participation in connected activities with different
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
5
objects can also cause contradictions (Kuutti, 1996) since real life situations involve
connected web of activities that can be distinguished according to their objects.
Activity systems realize and reproduce themselves through actions and operations
(Nicolini, 2013, p.108). The actions and operations of activity systems are executed in
accordance with, and adjusted to, the specific conditions of the context in which the
operations and actions take place (Bardram and Doryab, 2011). The change of contextual
issues (such as time, place or material kind) can also result in change of elements of activity
systems (ibid).
To summarize, in my analysis, a crucial notion is change in activity systems, due to
contradictions or change in elements of activity systems that are caused by changes in the
contextual issues. In the next section, the research context, which the fieldwork was done is
presented.
3.
RESEARCH CONTEXT
The study is affiliated with the Health Information System Program (HISP) which is based at
the Department of Informatics, University of Oslo. The primary goal of the HISP program is
to enhance the information use behaviour of health managers, planners and workers in the
health sector of developing countries, by strengthening local professional capacity, for the
development of a sustainable HIS (Braa et al., 2004). This study was conducted in Tanzania.
Primary health facilities are responsible for the provision of curative, preventive, promotive,
rehabilitative and palliative care to the population. The leaders in the communities can also be
involved in the health service provision through mobilizing, educating and sensitizing
community members to receive facility heath services. In rural areas, every hamlet has a
village health worker (VHW) who is a link between the community and the health facility.
The health management system in Tanzania is organized into national, zonal, regional, district
and facility levels. The national level is responsible for health policy and guidelines
formulation. The zone and regional levels are the extended arms of the Ministry of Health and
their main roles is to co-ordinate and support the regions and districts in planning,
implementation, monitoring and evaluation. The district is the focal point for the
implementation of the health policy and interventions. Its major role is formulating, carrying
out and monitoring health intervention packages and quality-assurance programmes and
supervising all health facilities. At the health facility level, a health facility management team
will implement the health policy and the interventions. Their major role is to carry out and
monitor health intervention and quality assurance programmes within the health facility. The
team will also supervise health related activities in the community within their catchment
area.
4.
RESEARCH METHODOLOGY
This study was conducted in the Dodoma region, in two districts namely Kisima and Ziwa1.
Eleven health facilities offering maternal and child healthcare services were researched, and
the study was mainly conducted in children clinics. The study was conducted during the
following periods: from August 2011-January 2012; July-August 2012 and November 2013April 2013. I employed a qualitative research approach, because the research was concerned
with exploring defaulter tracing practices (Silverman, 2005). I employed ethnomethodological ethnographic (Blomberg and Karasti, 2012) inspired data collection methods.
Ethno-methodologically informed ethnography helps researchers to understand and describe
practice by foreground how practitioners accomplish practices through situated resources
(Nicolini, 2013, p.221).
1
The names used in this document for the districts are not the real names.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
6
I participated in children clinics in health facilities during health service provision,
where I observed information practices of recording, identification, and tracing of defaulters.
In addition, I reviewed information artifacts such as children cards, antenatal cards, defaulter
lists, children registers and other tools which were used for interaction between facility and
community, such as village registration system and VHW’s notebooks. I also attended facility
meetings where defaulter tracing practices were discussed among health workers. I also
participated in tracing defaulters in homes and attended community gatherings such as
initiation of new vaccines. Bryman (2008) categorizes a participant observer’s roles into
complete participant, participant-as-observer, observer-as-participant and complete observer.
A complete participant is a fully functioning member of a social setting and other members
are not aware of the researcher’s identity. A participant-as-observer is similar to complete
participant but other members are aware of researcher’s status. An observer-as-participant
researcher performs some observation with very little participation while in complete
observation the researcher does not interact with people. In both facility and community
levels, my roles altered between observer as participant, and participant as observer. In total,
101 sessions of observations made in facilities with each lasting between 3 to 7 hours, were
included in the data collection.
I also conducted some unstructured and semi-structured interviews to get an overview
of work practices around clients’ retention. In some cases I performed group interviews for a
group of health workers doing a certain activity. In most cases I performed in-situ interviews
(Jordan and Henderson, 1995) in Kiswahili language in health facilities while performing
observation of interaction of information practices and health service provision. I had
encounters with eight medical doctors, 14 nurses, 13 community leaders, two patient tracing
coordinators, and 12 CHW (Community Health Workers). A summary of the interviewees in
is presented in Table 1.
Table 1: Summary of Interviews at Facility and Community Levels
Type of Informant
Number of
Informants
CHW
Community leaders
Nurses
Medical doctors
Patient tracker coordinator
Total
12
13
14
8
2
49
Total Number of Interviews (with
repetition)
Individual
Group
7
4
10
2
13
2
9
1
5
44
9
In addition, all coordinators for health services supporting maternal and child health
services at district, regional and zonal levels were interviewed. The theme of my interviews,
with zonal/regional/district officials, was the essence of defaulter tracing practices, and its
relations to information work practices at the facility level. A summary of the interviewees in
relation to their working places is presented in Table 2.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
7
Table 2: Summary of Interviews at District, Regional and Zone Levels
Location
Type of
Informant
Number of
Informants
Ziwa
RCH2
HIV/AIDS
HMIS
Health officer
3
2
1
1
Total number of
interviews (with
repetition)
9
4
2
2
Kisima
RCH
HIV/AIDS
HMIS3
Health officer
5
2
1
1
11
6
1
2
RCH
HIV/AIDS
Health officer
Total
3
1
1
21
4
1
2
44
Regional and
Zone Levels
I also reviewed several policy documents such as Tanzanian Health Policy,
Reproductive and Child Healthy Strategic Plan, PMTCT (Prevention of Mother to Child
Transmission of HIV) guidelines, Home Based Care (HBC) service guidelines, national
guidelines for the management of HIV/AIDS and HIV Act and policy. These documents gave
me background information about maternal and child healthcare provision.
A research proposal for conducting the study was submitted and approved by
University of Dodoma’s Research and Publication Ethical Committee. Ethical issues were
addressed during fieldwork, for example, through gaining written permission prior to the
fieldwork from both districts and obtaining prior oral consent from health workers to conduct
interviews or observation.
My main approach for analysing data was qualitative (Miles and Huberman, 1994)
with thematic analysis being the main technique. During fieldwork, I took notes which later
were transcribed into field notes. The unit of my analysis is a defaulter tracing activity
system. Kuutti (1996: 32) state that the border between action and activity could be blurred
and is dependent on what the subject or object is in that particular situation. That is, some
actions can be broken down into series of successive acts or a goal can be broken down into
sub-goals depending on the situation. While the actions around general HIS consists of
recording, processing, analyzing, presenting, reporting and using information (Heywood and
Rohde, 2001) and tracking (Gizaw et al., 2012), I have decided to focus on a more granular
level where I break HIS system into sub-information systems such as health care HIS system
(recording) and tracking (defaulter tracing system). By analysing a defaulter tracing system as
an activity system, instead of an action or sub-goal around a general HIS system, I have been
able to follow it across space (health facilities, health programs and geographical contexts).
The decision is according to Nicolini (2009; 2013) who asserts that when researchers want to
understand work activities, they should choose and follow a particular practice.
Through reading and reflecting field notes over a period of time, the theme of change
emerged. Activity theory emphasizes how practitioners develop collective competencies
(Nicolini et al., 2003). Kaasboll et al. (2010) argue that users develop information systems
competence around three domains: information, technology and work tasks. I have used the
2
3
Reproductive and Child Healthcare
Health Management Information Systems
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
8
categorization of information systems competence by Kaasboll et al. (2010) to determine
change in defaulter activity system. By coding-procedure I found that the changes in defaulter
tracing activity systems were the result of contradiction. I also identified other interacting
activity systems which are health management activity system, client activity system,
community based IS activity systems and identified contradictions between defaulter tracing
system and these activity systems. I then compared defaulter tracing practices between rural
and urban geography and across health programs and found that the context has influence on
defaulter tracing practices.
5.
DEFAULTER TRACING PRACTICES
5.1
Prologue
During the time of the research, Tanzania National Health Policy entailed that, all children
under five years of age are supposed to be enrolled into different health programs when they
are between four and six weeks. There were many programs including immunization, weight
growth monitoring, nutritional supplement distribution and PMTCT (Prevention of Mother to
Child Transmission of HIV). All children born in Tanzania were supposed to be enrolled into
one of the children clinics where immunization services would be performed. Children born
from HIV positive mothers would also be enrolled into the PMTCT program. In some
facilities, children who were underweight or stunted would be enrolled into nutritional
supplement distribution program. In these programs, children were supposed to attend the
clinic according to a schedule as stipulated in the health program guidelines.
However, clients did not follow their clinic schedules as directed by the health
workers. There were several reasons which were given to account for the missed
appointments. The following comments emerged in conversations with three health workers
in different health facilities:
“We have a peak in drop out numbers during farming and harvesting
seasons, since some clients could move temporarily to their distant
farms or to Bahari district to work in farms as migrant worker”.
“Some clients move between facilities and we have no way to know
about their status”.
“Some clients do not want to be labeled as HIV positive due to social
stigma”.
These comments show that clients also have other concerns; social-economic (e.g.
moving between facilities for social reasons, moving temporarily to improve their economic
status, etc.) or psychological and social, such as avoiding the social stigma attached to the
HIV disease. These concerns contradict the health care provision activity system, which aims
for the well being of the whole population.
Defaulter tracing practices included collective efforts by different actors. For instance,
in April 2013, there was a campaign all over the country, where a whole week was designated
to make sure that the children, who were behind their vaccines schedule, were immunized4.
For the whole week, facilities and community leaders collaborated in making sure that all
children had been immunized according to their clinic schedules. I also observed in February
2013 in Kisima district, a campaign, coordinated by government and various donors, calling
all people who have dropped out of HIV care to return. These examples elaborate that
4
http://www.afro.who.int/en/tanzania/pressͲmaterials/item/5529ͲtanzaniaͲcommemoratesͲworldͲ
immunizationͲweekͲwithͲaͲcallͲtoͲensureͲallͲchildrenͲareͲvaccinated.html
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
9
defaulter tracing systems are recognized and realized by various actors from regional, district,
facility and community levels. These examples have also shown that there are expectations
and efforts around improving retention of patients in HIV care and immunization services.
Also, the existing information systems were potential tools for facilitating
identification of defaulters. For example, one regional officer had the following to say
regarding the use of HIS reports at regional level in immunization defaulter activities: “We
know the problem of defaulters through immunization monthly reports. Every district/facility
is given immunization targets. When the figures are low we will direct the unit about the
problem. We leave it up to them for the strategies to raise immunization coverage.”
In the next section, I am going to present the operation of the defaulter tracing system
in health facilities.
5.1 Defaulter Tracing Activity System
Three main actions comprised the defaulter tracing activity system: identifying defaulters,
tracing them, and documenting the outcome. Regularly, nurses would conduct identification
of defaulters (by searching through the register(s)) to determine clients who are overdue in
their health program(s) schedules. Afterwards, the tracing defaulter action would be
performed with the aim of reminding or informing the defaulting clients. Lastly, people who
have been involved in tracing activities were supposed to provide feedback to the health
facility workers about the outcome of the tracing activity.
Table 3 presents the illustration of the defaulter tracing’s activity system. The
components of an activity system are presented to the left. To the right, I present additional
description in order to elaborate the components.
Table 3: Defaulter Tracing Activity System
Component
Object
Subject
Description
Clients completing health service schedules
Facility health workers
Health information system
Mobile phones
Tools
Community based information systems
Information about defaulting clients
Finance
Medical ethics; improve retention of clients and coverage of health
Rules
services
Facility health workers identify defaulters
Tracing defaulters directly or indirectly through other stakeholders
Division of
(mobile phone callers, catchment population leaders or CHW)
Labor
Stakeholders documenting the outcome and giving feedback to health
facilities
Community Facility health workers
Catchment population
Table 3 provides the basic structure of defaulter tracing system in health facilities. The
community involved included many participants: ranging from catchment population leaders
and CHW) to facility health workers. The main object is that clients would complete health
service schedules. There are many tools employed for the activity. The health programs’
register books are used to identify defaulters, and the outcome is the list of names of clients
who have defaulted. During tracing activities, mobile phones could be used for
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
10
informing/reminding the clients. Sometimes financial support could also be needed in order to
cover transport costs. Information in the community based information systems (village
registration information systems), would be used to provide defaulting clients’ location details
(in case the house of a client is unknown). CHW notebooks could be used to obtain
information about the status of the client or previous follow up activities. The outcome of
tracing activities is an updated list of defaulters with their status (e.g. travelled, died,
emigrated, client had attended services in other facility or client was found/not found). The
overall outcome of tracing activity could be that either the client comes back to health care or
that his/her status is known. For instance, if the outcome is that the client had attended another
facility, has migrated, or is deceased, the HIS registers would be updated. If client had not
been to any other facility, health services would be provided and eventually information
recorded in HIS tools. However, during participant observation and informal conversation
with health workers, I realized that defaulter tracing system is facing some contradictions as
presented in the next section.
5.2 Contradictions within Defaulter Tracing Systems
There were no formal tools where information about defaulters could be registered after the
identification action. As a result, facilities visited have developed local information practices
for dealing with defaulter tracing practices. Defaulter lists were prepared in any available
plain paper as shown in Figure 2. The common data attributes for identified defaulters were
child’s name, mothers’ name, street/hamlet, and type of service defaulted.
Figure 2: Sample defaulter lists in immunization services
There was a problem of lack of documentation of defaulter lists since the lists were
thrown away after tracing activity. However, not all cases of defaulters could be resolved
immediately and by throwing the lists, documentation about unresolved defaulter cases, and
their feedback/outcome, would be lost. This further led to another problem; duplication of
work in identifying and tracing defaulters, which is exemplified by the following comment
from a CHW during follow up services: “Again child Z! I have already made follow up
previously and gave feedback to facility that she is known to no one. It seems that this lady
was just a visitor to the village. I am surprised to see that her name is brought back for me to
follow her up again”. CHW noted this problem because she kept records of her follow up
activities in a notebook which included defaulter tracing activities notes. She had learned such
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
11
information management from a non government project she had been working with
previously.
There was another problem with developed tools and practices which emanated from
lack of clear guidelines about the interval for tracing defaulters. The interval for tracing
defaulters differed between the facilities, and I perceived that this difference led to a
difference in the efficiency of defaulter tracing systems among facilities. Some facilities could
identify and trace defaulters on a daily basis while other facilities could identify and trace
only once per month or even at longer intervals. For instance, the right picture in figure 2
shows a defaulter list generated for the period of March to April (two months), and it shows
some children who have escaped all vaccines (‘zote’ word, underlined in the picture).
Meanwhile, in the left picture the list was generated on 3rd January 2013, and it included
information for informing defaulters and reminders to some clients. For example, the list
included reminders for clients whose due date had not yet passed (such as on 11th to 25th dates
in January as shown by arrows in the diagram).
There was also a contradiction between the actions of identifying defaulters and
tracing defaulters. In identifying defaulters, health workers would go to each client’s record in
registers, and then compare date of birth and current date. Manually, they would figure out in
which health service stage, the child was supposed to be in, according to health service
guidelines. In most cases nurses did a manual calculation or used calendars from the wall or
from mobile phones. However, the process was tedious as there were many children
registered in programs. I observed several occasions where health workers calculated wrongly
(i.e. that the proper due date not yet arrived) and tracing activities were conducted to clients
who actually were not defaulters.
This section have shown contradictions which exists in the defaulter tracing system in
health facilities. The source of the contradiction is a lack of formal tools and guidelines to
guide the activities. The contradictions have led to some changes such as development of
informal practices to guide the activities. The next section present contradictions which arise
from the interaction of defaulter tracing activity system with other activity systems.
5.3 Contradictions in Defaulter Tracing System caused by interaction with Other
Activity Systems
Defaulter tracing system in operation was interacting with other activity systems such as the
health management activity system, the client activity system, and the community based IS
activity systems. The interaction resulted in contradictions which are elaborated in this
section.
A health management activity system has the object of managing healthcare services
for the whole population in the country by developing policy and guidelines, doing
supervision and coordination to lower levels. The community component in the health
management activity system consists of doctors, nurses and health managers (at facility,
district, regional and national levels). Tools used are health care data, budget, policy and
guidelines. However, in some facilities visited, even though defaulters have been identified or
known, tracing activities could not be performed due to financial constraints as explicated by
one facility manager: “We are supposed to do outreach services and trace defaulters in B
area. But we do not have money for transport”.
The financial problem was also related to village health workers (VHW), a cadre
considered as volunteers in Tanzania. One VHW had the following to say: “You see we just
visited six houses and we spent about three hours moving from one part of the village to
another since defaulter houses were in different locations of the village. This is a volunteering
job! My fellows are farming while I am tracing clients. How will I survive?” Lack of
incentives to VHW had consequences for the involvement of VHW in maternal and child
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
12
health services. In four villages visited, only one or two VHW were active even though there
were designated VHWs as stipulated in national health policy. This resulted in active VHWs
serving more than one street and hence increasing their workload.
Hence, there was a misfit between defaulter tracing and health management activity
system due to limited budget. The misfit between the defaulter tracing activity system and the
health management activity system resulted in changes in approaches for informing clients.
Defaulters were not only informed/reminded through facility health workers or community
leader/health workers, but also through other clients, such as neighbours, that were present in
children clinics. Nurses would on a daily basis ask clients who lived in the same street as the
defaulters’ mother about how to get in touch with the defaulter or ask them to convey
messages to them. One nurse had the following to say: “We use mothers to convey messages
to other mothers. Some of clients are staying in the border of the village which takes
approximately one hour to walk. We use this approach since we have no financial means to
visit clients”. Figure 3 shows different defaulter tracing approaches: left picture shows
immunization session in a facility, where a nurse performs health service provision, record
information while searching for defaulters and inquire feedback from clients and sending
message to defaulters. The right picture shows a VHW tracing client at home.
Figure 3: Different defaulter tracing approaches
Client activity system: The community element around client activity system include
the family, friends, and neighbors. Client activity system’s tool element consists of patient
cards, finance and time to go to health facility to continue with health care. Contradictions
between defaulter tracing and client activity system arose in tracing activities in two ways.
Firstly, people who traced clients might not find client(s) at home, since there was no prior
communication between tracers and clients about the time of visits. Secondly, some clients
would not come to facilities immediately when informed about their defaulter status. This
kind of disturbance could result in the tracing activity having to be repeated several times.
Health care IS activity system: Identification of defaulters is done through HIS tools
such as children and PMTCT registers. Health workers might not have recorded data properly
during health service provision, for instance immunizing a child without recording any
information in the child registry. One story was narrated in one facility by a nurse and a
VHW: “One time we had problems in tracing defaulters activities. We searched for
immunization defaulters in the registers and sent a message to the village authority. The
village authority sent community guards to bring mothers to the facility so that children could
be immunized. When mothers were brought to the facility, we realized through their cards
that they had done the services required but data was not recorded into facility registers. We
had to apologize to village authority and mothers. We also had to stop doing follow up and
updated registers before we could generate another defaulter list”.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
13
The excerpt shows that where there is incomplete recording of information, the result
is incorrect defaulter identification and eventually tracing a wrong person.
Community based IS: Information systems which operate at community level include
records about certain activities which occur at community levels; e.g. data on births, death
and on education (Marsh, 2000). The people (subjects) involved in community-based
information systems are CHW and community council members. There was a contradiction
between defaulter tracing system and community based HIS arising from absence of tools to
some CHW which resulted into duplication of work as shown in previous paragraphs in the
current section. I found that some CHWs during tracing activities, either didn’t have
information tools or used improvised ones. Among 6 village health workers I had interaction
with, only 1 had a notebook. Nevertheless, they would document on informal tools (paper
slips) which were thrown away after the activity or give oral feedback which resulted in
duplication of identifying and tracing defaulters as elaborated in section 5.2.
This section has elaborated on how defaulter tracing activity system interacts with
multiple systems which influenced defaulter tracing actions. Findings have shown that the
interaction between defaulter tracing system and its neighboring activity systems resulted in
contradictions and also changes. The changes were related to tracing activities whereby some
clients could be used to trace their fellow clients in their home vicinity. The section has also
shown other challenges such as information recording incompleteness during health care
provision, financial limitation and absence of CHWs’ information tools. I also found that
defaulter tracing system actions were also influenced by being in different situations, as
elaborated in the next sections.
5.4 Defaulter Tracing Activity System in Different Health Programs
I also compared defaulter tracing activity systems in immunization and PMTCT health
programs. In PMTCT program, as in other HIV related services, privacy of individuals is of
great concern. The need for protecting the patients’ privacy in PMTCT services resulted in
different approaches to defaulters’ tracing and different involvement of catchment population
leaders.
The need for privacy in PMTCT led to limitations for informing clients and
documenting the outcome and involvement of community leaders during the tracing activities.
Two CHW working with children under PMTCT services (female and male) had the
following to say regarding importance of privacy in tracing PMTCT clients: “I can not visit
the client if she is not willing to be visited at home; even if she is willing, I can not
remind/inform her in presence of other family members unless I am sure that the family
members are aware of her health status. I usually have some notebook in my handbag which I
use to document the outcome shortly after I part from client”. “If the client is not willing for
me to visit her at home, I will wait for an opportunity such as a religious, burial or wedding
ceremony or open day markets or on the road path. I usually document the outcome when I
come back home”.
Defaulter information management was also different in different programs. While the
defaulter list in immunization explicate clearly for which service client had defaulted, in
PMTCT the list does not mention the service(s) which clients have defaulted. The next figure
shows two pictures of defaulter lists, in PMTCT and immunization services showing the
differences.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
14
Figure 4: PMTCT and Immunization defaulter lists. Note that the service is mentioned
on top of immunization list (encircled) while for PMTCT no service is mentioned.
There are also differences in tracing activities between the two programs. First,
defaulter tracing activities in the two health programs drew the support of two different
groups. For immunization services, the local government leadership backed up the attempt to
retain children in the program. For example, some catchment population leaderships had set
bylaws5 which provided provision where catchment population could be punished if children
did not attend vaccination services, e.g. by setting amount to be paid. In PMTCT services,
defaulter tracing practices were backed up by HIV guidelines and donors. Donor support
allowed tracing of clients through mobile phone by calling the clients, and CHWs working
with HIV related health services, were provided with notebooks and other information tools.
In community group, there was also involvement of other stakeholders such as
PLWHIV (People Living With HIV) associations and at the same time limitation for full
involvement of community members. The following comments were extracted from
conversations with two patient tracing coordinators: “If clients change residential areas, I can
ask about their whereabouts from community leaders in new residential areas. Nevertheless, I
would not explain why I need them. Sometimes I lie and say they are my old friends from
primary school….if clients have registered in one PLWHIV association, we could ask them to
help to inform/remind clients”.
To summarize, differences in the context (nature of health services) where defaulter
tracing system operates is reflected in the difference in action, tools and community element
of the defaulter tracing activity system. Next section presents differences of the defaulter
tracing system which could occur due to different geographical settings.
5.5 Defaulter Tracing Activity System in Rural and Urban Geographical Contexts
Changing the context from urban to rural could result in change of defaulter tracing practices
due to different defaulter tracing approaches and involvement of catchment population
leaders. For instance, Kisima district is characterized by a mix of urban and rural areas and
one district coordinator had the following to say “A rural dweller is more willing to volunteer
than an urban dweller. We find more volunteering by community members in rural than
urban wards”. Likewise, one nurse, who worked in a rural district and later moved to an
urban district, had the following to say: “I see differences in tracing defaulters’ activities
between Maporomoko facility and this facility. Firstly, clients lived in permanent houses
compared to this ward where many clients live in rented houses and change residential areas
5
The bylaws needed to be approved by the respective district administrations
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
15
frequently. Secondly, we could involve village authority or even attend village meetings which
were held at least four times a year and could talk about the health facility activities and the
challenges we were facing, while in this ward there are rare occasion where all population
will meet. We could even conduct meetings with hamlet leaders and their ten cell leaders, and
discuss about their habitants behavior towards health services attendances.
These comments show how health workers needed to change their approaches of
tracing defaulters based on whether it was a rural or urban contexts. Therefore, change in the
context (rural and urban contexts) where defaulter tracing system operates could lead to
change of elements of the defaulter tracing activity system.
6.
DISCUSSION AND CONCLUSION
This study aimed to describe the operation of defaulter tracing systems in practice. The study
had identified defaulters’ information practices of identifying, tracing and documenting the
outcome similar to the literature (e.g. Nglazi et al., 2010; Thomson et al., 2011; Mwatawala et
al., 2012). Nevertheless, in maternal and child health settings, I did not observe similar
statistics of defaulter practices similar to Thomson et al. (2011). The findings have elaborated
that in defaulter tracing practices different information systems are used, both facility based
(health care IS system) and community based information systems (village registration
system, and CHW notebooks).
There are several approaches for tracing defaulters such as using CHW, home visits,
phone calls, written reminders, collaboration with community-based organizations and
outreach teams (Cutts, 1991; Marsh, 2000; Fraser et al., 2007; Forster et al., 2008; Thomson
et al., 2011; Chadambuka et al., 2012). My findings showed similar approaches and in
addition, I have elaborated on a new approach, where clients could be used to inform other
clients as a way of countering limitations arising from managerial and clients’ perspectives.
My findings have shown that several factors challenge the implementation of defaulter
tracing system and the challenges can be categorized into information systems and contextual
related. Information systems can be a challenge to defaulter tracing practices due to undesired
practices such as poor information collection practices (Chadambuka et al., 2012; Forster et
al., 2008; Fraser et al., 2007; Thomson et al.; et al., 2012; Ghadrshenas et al., 2013). The
findings concur with those studies. Findings have shown how information systems challenge
defaulter tracing practices from both design (absence of clear guidelines and tools) and
implementation (information recording incompleteness and wrong calculation of due date).
Further, findings have shown that defaulter tracing system could also be challenged by
contextual issues such as socio-economic (financial constraints, clients’ perspectives), nature
of health services and geographical context. The study has also shown the influence of the
context on defaulter tracing systems, which led to change in information behavior such as
recording and using information. For instance, rules between the programs of immunization
and PMTCT differ, and this makes change in information behavior. The rural and urban
contextual influence found in the study is similar to Cutts (1991) who showed that urban-rural
disparities have a influence on health service provision.
The findings have shown how activity theory could inform the understanding of
defaulter tracing system in practices and its associated challenges. By using activity theory,
the challenges facing defaulters tracing system have been elaborated. For instance the
concepts of contradiction have helped the study to uncover the challenges which emanates
from the paper based register, people using the tools and the context through which use of
tools occur. Furthermore, activity theory provides a conceptual tool to determine specific
contextual elements which challenges the implementation of patient centred information
systems. The findings have shown how change in the context (nature of health services,
geographical settings) results change in the use of information tools and action-operation of
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
16
certain health care activities such as defaulter tracing system. The actions of tracing defaulters
were adjusted according to specific conditions of rural-urban or immunization-PMTCT
services contexts. The findings were similar to Bardram and Doryab (2011) who showed that
context have influence on actions of a particular activity system. Further studies can also be
conducted in order to provide understanding on how the nature of activity system unfolds as
its context changes. In the next section, I present study recommendations and implications for
HIS design and implementation.
6.1 Recommendations and Implications for HIS Design and Implementation
Many studies on defaulter tracing systems have shown that EMR offer potential for retention
of clients since they provide automatic identification of defaulters and other benefits obtained
through electronic management of data (Fraser et al., 2007; Forster et al., 2008; Nglazi et al.,
2010; Thomson et al., 2011; Oluoch et al., 2012; Derenzi et al., 2012). Ghadrshenas et al.
(2013) argue that databases linking information between facilities and laboratories would
facilitate follow-up of clients. Nglazi et al. (2010) argue that computerized defaulter tracing
systems can be used to track patients efficiently and its utility depends on user friendliness,
affordability, sustainability, stability, security and data ownership. I argue that to track clients
efficiently, its utility also depends on policy, practice and design of health information
systems.
On a policy level, the institutionalizing of defaulter tracing practices in maternal and
child healthcare is important, for example by developing guidelines and tools to facilitate the
practice. Development of guidelines and information tools for village health workers which
could enhance tracing of defaulter and other similar activities are highly recommended. There
is also a need for providing incentives to people who are involved in tracing activities such as
CHW.
Both Fraser et al. (2007) and Forster et al. (2008) have echoed their concern about the
impact of data quality problem in defaulter tracing systems. This study recommends that
facility health workers/ management try to devise workable strategies within the facilities
which would solve the problem so that design and implementation of patient information
systems would be of benefit to health care provision.
The use of computer based systems to support tracing activities is common in chronic
disease services. Both studies done by Fraser et al. (2007) and Thomson et al. (2011) have
shown the use of technical infrastructure such as mobile phones and computer systems to
improve the performance of defaulter tracing systems. However, designers and implementers
of computer based systems need to be aware that defaulter tracing activities could be
influenced by contextual rural-urban disparity and the nature of health program. Future
research can be conducted to understand deeper aspects of the workflow of defaulter tracing
systems, tools and reporting structure and its linkage with aggregate (HIS) and community
based information systems.
ACKNOWLEDGEMENTS
I wish to thank all people who have given me insights for writing this paper. My special
thanks go to IRIS36 participants for instructive comments and engaged discussions. The
anonymous reviewers’ insights are very much appreciated.
7.
REFERENCES
Bardram, J. and Doryab, A. (2011) Activity Analysis – Applying Activity Theory to Analyze
Complex Work in Hospitals. Proceedings of the ACM Conference on Computer
Supported Cooperative Work, 455-464.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
17
Blaya, J.A., Fraser, H.S.F. and Holt, B. (2010) E-health Technologies Show Promise in
Developing Countries. Health Affairs, 29, 2, 244-251.
Blomberg, J. and Karasti, H. (2012) Ethnography: Positioning Ethnography within
Participatory Design. Simonsen, J. & Robertson, T. (eds.) Routledge International
Handbook of Participatory Design Routledge New York, 86–116.
Braa, J., Monteiro, E. and Sahay, S. (2004) Networks of Action: Sustainability Health
Information Systems across Developing Countries. MIS Quarterly, 28, 3, 337-362.
Bryman, A. (2008) Social Research Methods, 3rd Ed., Oxford University Press New York.
Busza J., Damilola W., Hairston, A., Gable, A., Pitter, C., Lee, S., Katirayi, L., Simiyu, R.
and Mpofu, D. (2012) Community-based Approaches for Prevention of Mother to
Child Transmission in Resource-Poor Settings: A Social Ecological Review. Journal
of the International AIDS Society, 15, 2: http://www.ncbi.nlm.nih.gov/pmc/
articles/PMC3499910/ .
Chadambuka, A., Chimusoro, T.A., Tshimanga, M., Namusisi, O. and Luman, E.T. (2012)
The Need for Innovative Strategies to Improve Immunization Services in Rural
Zimbabwe. Disasters, 36, 1, 161 -173.
Chilundo, B., Sundby, J. and Aanestad, M. (2004) Analysing the Quality of Routine Malaria
Data in Mozambique. Malaria Journal, 3, 3: http://www.malariajournal.com/
content/3/1/3.
Cole, M. (1995) Socio-Cultural-Historical Psychology: Some General Remarks and a
Proposal for a New Kind of Cultural Genetic Methodology. In Wertsch, J.V., Río,
P.D. and Alvarez, A. (eds). Sociocultural Studies of Mind: Cambridge University
Press New York.
Cutts, F.T. (1991) Strategies to Improve Immunization Services in Urban Africa. Bulletin of
the World Health Organization, 69, 4, 407-414: http://www.ncbi.nlm.nih.gov/
pmc/articles/PMC2393244.
DeRenzi, B., Findlater, L., Payne, J., Birnbaum, B., Mangilima, J., Parikh, T. and Borriello,
G. (2012) Improving Community Health Worker Performance through Automated
SMS Reminders. Proceedings of the Fifth International Conference on Information
and Communication Technologies and Development, 25-34.
Elovaara, P., Igira, F. and Mörtberg, C. (2006) Whose Participation? Whose Knowledge?:
Exploring PD in Tanzania-Zanzibar and Sweden. Proceedings of the Ninth
Conference on Participatory design, 1, 105-114.
Engeström, Y. (2001) Expansive Learning at Work: Toward an Activity Theoretical
Reconceptualization. Journal of Education and Work, 14, 1, 133-156.
Feldman, M.S. and Orlikowski, W.J. (2011) Theorizing Practice and Practicing Theory.
Organization Science, 22, 5, 1240-1253.
Forster, M., Bailey, C., Brinkhof, M.W., Graber, C., Boulle, A., Spohr, M., Balestre, E., May,
M., Keiser, O., Jahn, A. and Egger, M. (2008) Electronic Medical Record Systems,
Data Quality and Loss to Follow-up: Survey of Antiretroviral Therapy Programmes in
Resource-Limited Settings. Bulletin of the World Health Organization, 86, 12, 939947: http://www.ncbi.nlm.nih.gov/pubmed/19142294.
Fraser, H.S., Jazayeri, D., Nevil, P., Karacaoglu, Y., Farmer, P.E., Lyon, E., Fawzi, M.K.,
Leandre, F., Choi S.S. and Mukherjee J.S. (2004) An Information System and Medical
Record to Support HIV Treatment in Rural Haiti, BMJ Journal, 329, 1142-1146.
Fraser, H.S.F., Allen, C., Bailey, C., Douglas, G., Shin, S. and Blaya, J. (2007) Information
Systems for Patient Follow-up and Chronic Management of HIV and Tuberculosis: A
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
18
Life-saving Technology in Resource Poor Settings. Journal of Medical Internet
Research, 9, 4, e29: http://www.jmir.org/2007/4/e29/.
Ghadrshenasa, A., Amorb, Y.B., Changc, J., Dalec, H., Shermand, G., Vojnova, L., Youngc,
P. and Yoge, R. (2013) Improved Access to Early Infant Diagnosis is a Critical Part of
a Child-Centric Prevention of Mother-to-Child Transmission Agenda. AIDS, 27, 2,
197–220.
Gizaw A.A., Mukherjee, A., Lewis J. and Sahay, S. (2012) Understanding Design-Use
Interaction: The Case of Name Based Health Information System in India,
Proceedings of IADIS International Conference on Information Systems, Berlin.
Heywood, A. and Rohde, J. (2001) Using Information for Action: A Manual for Health
Workers at Facility Level Equity Project.
Igira, F.T. (2008) The Situatedness of Work Practices and Organizational Culture:
Implications for Information Systems Innovation Uptake. Journal of Information
Technology, 23, 2, 79-88.
Igira, F.T. (2012) Information Systems Deployment as an Activity System. In: Dwivedi, Y.K.,
Schneberger, S.L. and Wade M.R. (eds.) Information Systems Theory: Explaining and
Predicting our Digital Society, Integrated series in Information Systems, 29, 2, 397411.
Issroff, K. and Scanlon, E. (2002) Using Technology in Higher Education: Activity Theory
Perspective, Journal of Computer Assisted Learning, 18, 1, 77- 83.
Jones, S.A, Sherman, G.G. and Varga C.A. (2005) Exploring Socio-economic Conditions and
Poor Follow-up Rates of HIV Exposed Infants in Johannesburg, South Africa. AIDS
Care, 17, 4, 466 - 470.
Jordan, B. and Henderson, A. (1995) Interaction Analysis: Foundations and Practice. The
Journal of the Learning Sciences, 4, 39-103.
Kaasbøll, J., Chawani, M.S., Hamre, G.A. and Sandvand, J. (2010) Competencies and
Learning for Management Information Systems. Journal of Information, Information
Technology and Organizations, 5, 85-100.
Kalembo, F.W. and Zgambo, M. (2012) Loss to Follow up: A Major Challenge to Successful
Implementation of Prevention of Mother-to-Child Transmission of HIV-1 Programs in
Sub-Saharan Africa. ISRN AIDS. http://dx.doi.org/10.5402/2012/589817.
Kamadjeu R.M., Tapang, E.M. and Moluh R.N. (2005) Designing and Implementing
Electronic Health Record System in Primary Care Practice in sub-Saharan Africa: A
Case Study of Cameroon. Informatics in Primary Care, 13, 179-186.
Kanjo, C. (2011) Pragmatism or Policy: Implications on Health Information Systems Success.
The Electronic Journal on Information Systems in Developing Countries, 48, 1, 1-20.
Kaptelinin, V., Kuutti, K. and Bannon, L. (1995) Activity Theory: Basic Concepts and
Applications: A Summary of a Tutorial Given at the East West HCI95 Conference.
Selected papers from the 5th International Conference on Human-Computer
Interaction, 189-201.
Korpela, M., Soriyan, H.A. and Olufokunbi, K.C. (2001) Activity Analysis as a Method for
Information Systems Development. Scandinavian Journal of Information Systems, 12,
1–2, 191–210.
Kuutti, K. (1996) Activity Theory as a Potential Framework for Human–Computer Interaction
Research. In Nardi, B.A. (ed.) Context and Consciousness: Activity Theory and
Human–Computer Interaction MIT Press Cambridge, 17–44.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
19
Manda T.M. and Msosa Y. (2011) Socio-technical Arrangements for mHealth: Extending the
Mobile Device Use and Adoption Framework, R. Popescu-Zeletin et al. (Eds.) Third
International ICST Conference, AFRICOMM 2011, 208-217.
Marcos Y., Phelps, B.R. and Bachman, G. (2012) Community Strategies that Improve Care
and Retention along the Prevention of Mother-to-Child Transmission of HIV Cascade:
A Review. Journal of the International AIDS Society, 15, 2:
http://www.jiasociety.org/index.php/jias/article/view/17394.
Marsh, D.R. (2000) Population-based Community Health Information Systems. In: Lippeveld,
T., Sauerborn, R. and Bodart, C. (Eds.) Design and Implementation of Health
Information System World Health Organisation Geneva, 146-175.
McLeod, B., Phillips, J., Stone, A.E., Walji, A. and Awoonor-Williams, J.K. (2012) The
Architecture of a Software System for Supporting Community-based Primary Health
Care with Mobile Technology: The Mobile Technology for Community Health
(MoTeCH) Initiative in Ghana. Online Journal of Public Health Informatics, 4, 1:
http://www.ncbi.nlm.nih.gov/pubmed/23569631
Miles M. and Huberman, M. (1994) Qualitative Data Analysis, Sage.
Mwatawala, S., Sando, D., Malele, R., Moshiro, C., Senyael, B.R., Somi, G., Maselle, A.,
Chalker, J. and Ross-Degnan, D. (2012) Strengthening the Appointment and Tracking
Systems for Patients on Antiretroviral Therapy in Tanzania: A Strategy to Optimize
Adherence to ART. International Journal of Personal Centred Medicine, 2, 4, 825836.
Nardi, B.A. (1996) Studying Context: A Comparison of Activity Theory, Situated Action
Models and Distributed Cognition. In Nardi, B.A. (Ed.) Context and Consciousness:
Activity Theory and Human-Computer Interaction MIT Press, Cambridge, 69-102.
Nglazi, M.D, Kaplan, R., Wood, R., Bekker, L. and Lawn, S.D. (2010) Identification of
Losses to Follow-up in a Community-Based Antiretroviral Rherapy Clinic in South
Africa Using a Computerized Pharmacy Tracking System. BMC Infectious Diseases.
http://www.biomedcentral.com/1471-2334/10/329
Ngoma, C. and Igira, F.T. (2012) Contradictions in Collecting and Recording Maternal Health
Data at the Community Level: A Case Study from two Districts in Tanzania. Journal
of Health Informatics in Developing Countries, 6, 2, 509-520.
Ngoma, C., Chawani, M.S. and Herstad, J. (2010) Adaptation of Mobile Application to
Improve Flow of Birth Information from the Community to the District Level. 2nd
International ICST Conference on e-Infrastructure and e-Services for Developing
Countries, 79-92.
Ngoma, C., Kimaro, H., Igira, F., Mukulu, J., Kaasbøll, J., Yona, A., Mkabara, N., Sarikoki,
S., Sangu, I., Njau, C., Mnaya, A., Mpoyo, S., Mlongakweli, Z., Mwaga, W. and
Pilyimo, R. (2012) Challenges in Implementing a Computerized Namebased
Information Tracking System: Practical Experiences from Maternal Health Care. ISTAfrica 2012 Conference Proceedings, International Information Management
Corporation.
Nicolini, D. (2009) Zooming In and Out: Studying Practices by Switching Theoretical Lenses
and Trailing Connections. Organizational Studies, 30, 1391-1418.
Nicolini, D. (2013) Practice Theory, Work, and Organization: An Introduction Oxford
University Press United Kingdom.
Nicolini, D., Gherardi S. and Yanow, D. (eds.) (2003) Introduction: Toward a Practice-Based
View of Knowing and Learning in Organizations. In Knowing in Organizations: A
Practice Based Approach, M.E. Sharpe Inc. New York.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
EJISDC (2014) 64, 2, 1-20
20
Oluoch, T., Santas, X., Kwaro, D., Were, M., Biondich, P., Bailey, C., Abu-Hanna, A. and De
Keizer N. (2012) The Effect of Electronic Medical Record-Based Clinical Decision
Support on HIV Care in Resource-Constrained Settings: A Systematic Review.
International Journal of Medical Informatics, 81, 10, e83–e92.
Puri, S.K., Byrne, E., Nhampossa, J.L. and Quraishi, Z.B. (2004) Contextuality of
Participation in IS Design: A Developing Country Perspective. Proceedings of the
Participatory Design Conference, ACM Press, 42-52.
Rotich, J.K., Hannan, T.J., Smith, F.E., Bii, J., Odero, W.W., Vu, N., Mamlin, B.W.,
Mamlin, J.J., Einterz, R.M. and Tierney, W.M. (2003) Installing and Implementing a
Computer-Based Patient Record System in Sub-Saharan Africa: The Mosoriot
Medical Record System. Journal of the American Medical Informatics Association,
10, 4, 295 -303.
Safadi, H. and Faraj, S. (2010) The Role of Workarounds During an Opensource Electronic
Medical Record System Implementation. Proceedings of the Thirty First International
Conference on Information Systems, Paper 47.
Sauerborn, R. and Lippeveld, T. (2000) Introduction. In: Lippeveld, T., Sauerborn, R. and
Bodart, C. (eds.) Design and Implementation of Health Information System World
Health Organisation Geneva, 1-14.
Silverman, D. (2005) Interpreting Qualitative Data: Methods for Analysing Talk, Text and
Interaction, 2nd Edition, Sage London.
Simba, D.O. and Mwangu, M.A. (2005) Quality of a Routine Data Collection System for
Health: Case of Kinondoni District in the Dar es Salaam Region, Tanzania, South
African Journal of Information Management, 7, 2, 8 pages: http://www.sajim.co.za
/index.php/SAJIM/article/viewFile/262/253.
Thomson, K.A., Cheti, E.O. and Reid, T. (2011) Implementation and Outcomes of an Active
Defaulter Tracing System for HIV, Prevention of Mother to Child Transmission of
HIV (PMTCT), and TB patients in Kibera, Nairobi, Kenya. Transaction of the Royal
Society of Tropical Medicine and Hygiene, 105, 320-326.
The Electronic Journal of Information Systems in Developing Countries
www.ejisdc.org
Appendix 3: Paper III
Shidende, N.H., Aanestad, M., and Mork, B.E. (2014). The Role of Information artefacts in
Organizing: an Empirical Study from Tanzanian Health Care, Paper presented at the
Conference of Organizational Learning, Knowledge and Capabilities, Oslo, Norway, 2014.
III
THE ROLE OF INFORMATION ARTEFACTS IN ORGANIZING:
AN EMPIRICAL STUDY FROM TANZANIAN HEALTH CARE
Nima Shidende1, Department of Informatics, University of Oslo, Norway,
[email protected]
Margunn Aanestad, Department of Informatics, University of Oslo, Norway,
[email protected]
Bjørn Erik Mørk, Department of Informatics, University of Oslo, Norway and IKON,
Warwick Business School, United Kingdom, [email protected]
Abstract: This paper discusses the role of information artefacts in organizing practices.
Empirically we draw upon ethnographic studies of information-related practices of
Tanzanian health workers. We examine the layout and usage of artefacts and describe
how the health workers modified the artefacts to facilitate the identification of children
not receiving immunization. Theoretically we position ourselves within practice-based
studies and thematize organizing as performative and emphasize the co-constitutive
nature of artefacts, practices and organizational forms. We contribute by shedding light
on how the organization of work can be crucially dependent on information artefacts
and the multiple and temporally shifting roles they play in situated practices.
Keywords: practice-based studies, artefacts, organizing, situated practices, healthcare
1. INTRODUCTION
The crucial role of objects and material arrangements for organizing has for a long time
been perceived by scholars, however, in organizational studies the predominant frame
for studying objects has been as “boundary objects”, since boundaries have been a longstanding and fundamental area of interest in social and organizational theorizing (see for
instance Ancona and Caldwell, 1977; Fennel and Alexander, 1987; Hernes, 2004). The
notion of “boundary objects” emerged in the science and technology studies (Star and
Griesemer, 1989) based on a symbolic interactionist perspective. It describes how
interaction between multiple different, but intersecting social worlds is practically
achieved. Boundary objects facilitate collaboration across different social worlds so that
collective results can emerge from distributed work, even in the absence of consensus
(Star, 2010). The notion has become widely used, first and foremost to explain how
multi- or cross-disciplinary work can be productively organized. Along with its uptake
in other fields a more instrumental understanding has become evident, e.g. boundary
objects are often described as deliberately created ‘interventions’ to facilitate
collaboration (Zeiss and Groenewegen 2009; Trompette and Vinck 2009).
Corresponding author: University of Oslo, Department of Informatics, att/Nima Shidende, Postbox P.O.
Box 1080, N-0316 Oslo, Norway. E-mail: [email protected]
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 1
In this paper we seek to enrich the perception of how objects take part in the core
activities of organizing beyond their functioning as boundary objects. We seek to add
richness to the ‘object discourse’ in organizational studies, with a specific interest in
studying how workers change the objects of their work to achieve desired outcomes.
The research question underpinning this paper is therefore:
What are the ways and reasons for the change in objects’ roles across time
and space?
We have examined how the health workers have adapted the official tools for
information recording in order to conduct their job better, and the study documents the
health workers’ adaptations in three stages:
1. Prior to October 2012
2. After new official forms and books with a different format and layout was
introduced in October 2012
3. After two new vaccinations were introduced in January 2013.
In the next section we review literature on the roles of objects in organizing practices. In
section 3 we present the empirical context in the Tanzanian primary health care sector
and the qualitative research approach that we employed. Our empirical findings are
described in section 4. Here we describe the variation in usage of tools between the four
facilities before and after the changes. In section 5 we discuss how the analysis enable
us to describe objects differently, while the concluding discussion in section 6 compares
the insights from our analysis against related research presented in section 2.
2. RELATED RESEARCH: MULTIPLE ROLES OF OBJECTS IN
ORGANIZING PRACTICES
Our study is positioned within practice-based studies (e.g. Gherardi, 2009; Nicolini et
al. 2003; Nicolini, 2013) and engages in particular with the works that thematize the
roles of objects. The recognition of the role of materiality for organizing is on the one
hand fundamental and widespread in this tradition, for instance Schatzki defined
organizations as a “bundle of practices and material arrangements” (Schatzki 2006;
1863), where the notion of arrangement pointed to “assemblages of material objects –
persons, artifacts, organisms and things” (ibid.; 1864) that can causally support the
ongoing activities. We follow Suchman and conceptualize organizations as ongoing
performances which involve “heterogeneous modes of action and materialization, both
of which must be actively affiliated and aligned across a range of often unruly
contingencies” (Suchman, 2000, p.313). Suchman’s emphasis on alignment of actions
and artefact in relation to contingencies is of central relevance to our study, where the
contextual situation is significantly constrained on resources of different kinds. On the
other hand, the predominant analytic frame for investigating objects has been to see
them as boundary objects that facilitates the integration of distributed or crossdisciplinary work (see e.g. Henderson 1991; Carlile 2002, 2004; Bechky 2003; Levina
2005; Barrett and Oborn, 2010). However, several works points to a broader
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 2
conceptualization of the role of objects. For instance, Nicolini et al. (2012) employ
multiple theoretical lenses to study the role of objects: the object notion in activity
theory (Miettinen and Virkkunen, 2005), the epistemic object (Rheinberger, 1997,
Knorr-Cetina, 1999), boundary object (Star and Griesemer, 1989) and objects as
material infrastructure (Star, 1999; Orlikowski, 2007). Nicolini et al. (2012) argue that
objects in a setting of cross-disciplinary research collaboration perform three functions:
1) they provide motives and drive for collaboration to emerge (this role is indicated by
the object notion in activity theory and the notion of the epistemic object), 2) they allow
participants to work across different types of boundaries (the “classic” boundary object
notion) and, 3) they constitute the fundamental material infrastructure of the activity.
The authors also show how these object roles change along with the nature and
unfolding of the activity itself, and how they may change from being mere tools, ‘readyat-hand’, to become an object of inquiry (e.g. in breakdown situations), or fade into the
background as ‘infrastructure’. Also, Neyer and Maicher (2013) argue that the role
which an object plays in interactive innovation is influenced by the situational,
individual or process dimension apart from multiple perspective on objects, i.e. material
infrastructure, boundary objects, epistemic objects, and activity objects (p. 542).
Several of the studies of objects examine research settings and experimental practices
that are explicitly oriented towards creating new knowledge about specific objects
(Swan et al., 2007). These settings may warrant the distinction between an epistemic
object (an object of inquiry, the elusive and ever-unfolding knowledge object) and a
technical object (a well understood, stabilized object) (Rheinberger, 1997). Other
studies analyse non-research-oriented work settings, however, also here is described a
dynamic nature of objects’ roles even if it does not emerge from an explicit epistemic
activity such as research. For instance, in Lindberg and Walter’s (2013) study of a
malfunctioning infusion pump in a hospital, they describe how the breakdown situation
was a trigger for a re-conceptualization of the tool. From an unquestioned, black-boxed
tool, an “object-in-use”, it became a visible “object-of-inquiry”, an object that was
enacted along with the construction of action nets. Similarly, Miettinen and Virkkunen
(2005; 443) describes how gradually emerging contradictions in the practice of
occupational safety inspectors triggered a process of learning, re-design and “retooling” of the practice. To date there is, however, not a significant amount of attention
given to the nature and driver of such changes of the object. Consequently, we wish to
contribute to this gap in the literature through analysing the ways and reasons for the
change in objects’ roles across time and space.
In order to do this we draw on studies that emphasize the infrastructural role of objects.
As illuminated by e.g. Star (1999) and Bowker and Star (1999), an infrastructural lens
may help to see the fundamental supporting substructure that facilitates human activity,
which is often invisible or taken-for-granted. Foregrounding a relational perspective,
Star and Ruhleder (1996) describe how the information infrastructure relates to the
work, as it “links with conventions of practice” and is “learned as part of membership”
(ibid., p. 113). These information infrastructures emerge through a co-evolution of the
work practice and its tools, through the community seeking convergence between its
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 3
information needs and the information capabilities that are offered by the tools (Star et
al., 1997). From a learning perspective, Kaasbøll et al. (2010) argue similarly that for
users to develop information systems competence they need to master both aspects
related to information needs, technology and work practices.
In this paper we argue that objects can play important roles as information artefacts.
Information artefacts, called “cognitive artefacts” by Donald Norman (1993), refer to a
category of artefacts that are predicated on representations and abstractions. The way
the representation is matched to the task is crucial for their utility. This goes both for
finding the relevant information and computing the desired conclusion. As an example
Norman (ibid., p. 57) contrasts different ways to organize flight time information into
tables. Some layouts of the table require cognitive restructuring and additional
computations, while other representations give a more immediate, experiential access to
the information, where the desired conclusion (“which flight has the shortest
duration?”) is available simply through scanning the diagram. This kind of direct
interpretation, which Norman calls experiential cognition, is desirable as it involves less
mental expenditure. “The form of representation most appropriate for an artifact
depends upon the task to be performed. The same information may need to be
represented differently for different tasks and with the appropriate choice of
representations, hard tasks become easy” (p. 75). It therefore matters how information
elements gets ordered into horizontal and vertical layers, what sequence and format is
chosen, and whether elements like arrows, lines, color usage, emphasis is added in a
helpful way. All tools are not similar, e.g. while simple lists are organized after one
factor, matrixes provides the combination of two organizing principles. Also different
kinds of graphical illustrations, e.g. pie charts, line graphs, bar graphs etc. are suited for
specific and different informational purposes. The utility of different forms of
representations of course also links with the human capabilities of the readers, which
may vary. In general, there is a crucial difference between artefacts that offer surface
representations (i.e. through a visible and physically accessible mode of working) and
those who offer internal representations (where the mode of working is invisible and
inaccessible). Norman uses the notion of affordances, and point out that “different
technologies afford different operations. That is, they make some things easy to do,
others difficult or impossible” (ibid., p. 106). Specific demands accrue if these artefacts
shall constitute a shared work space, or facilitate cooperative work. Moreover, the
artifacts need to provide permanence of memory, offer a high enough quantity of
memory, as well as completeness, precision, and be updated, etc. in order to alleviate
the humans of a heavy memory and computational load.
Summing up, an underlying premise for many practice-based studies is that the relations
of a tool to the practice of use is particular (Suchman 2000) and situated (Gherardi,
2009; Nicolini, 2009; 2013). The actual playing out of activities in a location, with a
given texture to the relationships, is the “site” of the practice (Nicolini, 2011), and
change, organizing, meaning, coordination and learning are all present in practice
(Huizing and Canavagh, 2011). Therefore, three core dimensions of our analytic
approach to the study are: a) the practices in which we study the role of objects, are
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 4
situated, i.e. significantly shaped by the context, b) the relation between an object and a
practice has evolved historically as participants have been seeking convergence and c)
objects in practices may have become routine, taken for granted until a disturbance
occur. Moreover, they are embedded into ensembles of other tools and practices, and
therefore we claim that they have an infrastructural quality.
3. RESEARCH CONTEXT AND APPROACH
We will now describe the research context – the primary health sector of Tanzania with
a focus on preventive health services (immunization). We then describe the research
approach and how we conducted our data collection and analysis.
The Tanzanian health care sector, the vaccination situation and the
information systems in use
This study has been conducted in Tanzania under the Health Information System
Program (HISP) umbrella. This is a long-standing action research initiative which has as
its primary goal to enhance the information use by health managers, planners and
workers in different levels of the health sector of developing countries (Braa et al.,
2004; Braa et al., 2007). The aim is to strengthen national health management
information systems (HMIS) through capacity building and providing open source
software tools for data gathering, reporting and analysis. The District Health
Information Software (DHIS2) is a national standard in Tanzania and it is used to
process aggregate data at district, regional, and national levels.
The administrative structure of the Tanzanian health sector is organized into five levels:
national, zone, regional, district and facility levels. The national level is responsible for
formulation of health policy and guidelines. The zone and regional levels’ main roles
are to co-ordinate and support the regions and districts in planning and implementing
the services, as well as conducting monitoring and evaluation. The district is the focal
point for the implementation of the health policy and interventions. Its major role is
formulating, carrying out and monitoring health intervention packages and quality
assurance programs. Essentially the district team will supervise all health facilities. At
the health facility level health facility management team will implement the health
policy and interventions. The primary health facilities, dispensaries and health centers,
are responsible for providing health services to the population. The service provision
may also involve the community leaders in order to mobilize, educate and sensitize
community members to receive facility heath services.
In Tanzania, the mortality rate for children below five years of age is 54 per 1000 live
birth.3 This implies that each year 45,000 babies die before they are one month and in
total 157,000 children die before their first birthday. Child health depends heavily on
2
Software product from the HISP network
In comparison the similar number in Norway is 3, United Kingdom 5, and 7 in the US.
http://data.worldbank.org/indicator/SH.DYN.MORT
3
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 5
availability of and access to immunization, quality management of childhood illnesses
and proper nutrition. The immunization activities are organized under the Reproductive
and Child Health Section, as a donor-supported Expanded Programme on Immunization
(EPI) programme that since 1975 has offered vaccinations to children across the
country. However, not all children get immunized. For example, a survey from 2010
showed that only 75 percent of children between 12 and 23 months were fully
immunized, a modest increase from 71 percent in 2005 and 68 percent in 1999 (TDHS,
2010).Tanzania is committed to achieving the Millennium Development Goals, out of
which the goal no. 4 aims to reduce child mortality, goal no. 5 is to improve maternal
health and goal no. 6 is to combat HIV/AIDS, malaria and other diseases, by the year
2015. One of the operational targets to be achieved by 2015 is to increase immunization
coverage to above 90% in 90% of the districts. In line with that, Tanzania has
implemented the Reaching Every District (RED) strategy to improve immunization
coverage, and in addition periodic supplementation campaigns are conducted every
three years. The activities that we address in this paper, the health workers’ provision
and documentation of vaccination services, are therefore subjected to a relatively high
level of monitoring and supervision through reporting.
The primary health facilities are provided with official information artefacts, such as
tally sheets, patient cards, report forms and register books from the Ministry of Health.
The facility-based recording of information help them organizes the service provision
locally, as well as to generate regular reports to the district, regional and national
authorities. In October 2012 some of these information artefacts were replaced with a
redesigned version together with new guidelines for recording and reporting
information.
Construction of the research material
This study was conducted in the Dodoma region in two districts namely Bahari and
Bwawa4. Dodoma region is one of the poorest regions in Tanzania (TDHS, 2010).
Despite several challenges for adequate health care provision in the region, the
percentage of children under one year that were fully vaccinated was 93% for year
2012. We got the impression during our fieldwork that the relatively high percentage
were achieved by deliberate efforts, including active defaulter tracing systems5.
Defaulter tracing systems are recognized and realized by various actors from regional,
district, facility and community levels and aims at improving retention of patients in
health programs such as HIV care and vaccination services. Table 1 provides an
overview of the socio-economic context in the two districts.
4
The district and facilities are anonymized
5
Procedures and tools that can enhance monitoring of patients who have dropped out of health programs.
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 6
Table 1: Overview of health and social welfare indicators in the two districts
District
Area
Rural/urban
2
(km )
Bahari
Bwawa
8742
2769
Population
(2012
projection)
Rural
Mixed
319 044
552 135
Socioeconomic
indicators
Maternal
mortality
Infant
mortality
Literacy
rate 60 %
(declining)
33/100 000 live
births
51/1000
births
Among the
poorest and
least
developed
districts
(unreliable data
because lack of
information
from
community
level)
(unreliable,
lack
of
information
from
community
level)
Literacy
rate 84 %
97/100 000 live
births
51/1000
births
live
live
Health
staff
shortage
Other challenges
49 %
Nomad
and
pastoralist
communities, i.e.
migrating
population
35 %
Weak
coordination
among
donor/partners
driven activities.
This setting poses huge challenges for conducting adequate health work. We may thus
argue that with the strategic importance for the problem our case represents a ‘critical
case’ (Flyvbjerg, 2012); if the health-workers devise tools and practices that work in
this resource-constrained context, it can potentially also work elsewhere. The critical
and central role of the information artefacts and their relation to the context makes this
context a uniquely illustrative setting in which to see the practical significance of the
understanding that we seek around the pivotal role of objects in organizing.
Two public health facilities which offered child healthcare were selected in each of the
districts; one health centre and one dispensary. Data collection was conducted mainly
by the first author in multiple site visits between August 2011 and April 2013. We have
employed a qualitative research strategy utilizing ethno-methodological ethnographic
(Blomberg and Karasti, 2012) field methods. Ethno-methodologically informed
ethnography helps researchers to understand and describe practice by foregrounding
how practitioners accomplish practices through situated resources (Nicolini, 2013,
p.221). This includes open ended (formal) interviews, participant observations together
with concurrent in-situ interviews (Jordan and Henderson, 1995), as well as document
(artefact) analysis.
Ethno-methodological ethnographic inspired observation involved participant
observation in four facilities during health services provision focusing on how they use
information artifacts while offering health services. The total number of sessions of
observation in all facilities was 90 which ranged from three to seven hours. The
individuals that we observed consisted of 16 nurses, 6 doctors, 1 health officer and 5
village health workers, all of whom were working with child health. We also conducted
observations at community level by attending 4 events regarding vaccination services.
In both facility and community levels, our role altered between observer as participant,
and participant as observer (Bryman, 2008).
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 7
During the observation we performed in-situ interviews (Jordan and Henderson, 1995).
We also conducted 10 individual and 2 group interviews to community leaders. The
theme for the interviews was the reasons for information practices observed such as use
of improvised artefacts and recording and tracking tasks. The interviews were also
conducted with immunization, health officers and program coordinators at regional and
district levels. The duration of interviews ranged from 30 to 90 minutes. In total we
conducted 20 interviews with 8 Reproductive and Child Health coordinators (RCH) at
district level, 4 interviews with 2 health officers and 3 interviews with 2 information
systems officers. We also conducted 6 interviews with 4 RCH coordinators at regional
and zone levels. The theme of our interviews, both with zonal, regional and district
officials, was the essence of defaulter tracing practices, and its relations to information
work practices at the facility level.
A diary was kept during the entire fieldwork in order to document the data collection
process. In addition, notes were used to document the ethnographic study. Some
conversations occurring during participant observation were recorded and later
transcribed as extensive field notes. Both English and Kiswahili were used in the field
notes. The ethnographic study was conducted in Kiswahili, since it is the most-used
language in Tanzania. Later, the notes were translated into ethnographic description in
English.
Artifacts reviewed included children cards, antenatal cards, defaulter lists, children
registers and other tools which are used for interaction between facility and community
(such as village registration system and the village health workers’ notebooks).
The empirical study was approved by University of Dodoma’s Research and Publication
ethical committee. Ethical concerns were kept in mind during fieldwork, for example,
through gaining written permission prior to the fieldwork from both districts managers,
and obtaining oral consent from health workers prior to conducting interviews or
observations.
Data analysis strategies
Data have been analyzed by using qualitative techniques (Miles and Huberman, 1984;
Bryman, 2008; Walsham, 2005). Notes were taken during participant observations,
informal conversations, and document analysis and later transcribed as expanded field
notes. These transcribed field notes were read through several times to obtain a sense of
the whole and of the theme of changes to the information artefacts.
For the first level of analyses, we prepared an “ethnographic description” (Emerson et
al, 1995) for each facility in order to understand the specifics of its situation, the
contingencies and the reasons given for the health workers’ adaptation of the
information tools. In the second step, we identified the roles of information artefacts in
each facility according to the framework provided by Nicolini et al. (2012). We then
compared the roles across facilities and noted that the roles are situated and affected by
the information infrastructure of the place.
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 8
There are two limitation of the study. Firstly, our findings are drawn from four facilities
only out of 107 health facilities in the two districts; the reasons being the difficulties of
conducting multiple ethnographic studies at the same time. Secondly, we were not
present at all the times in the facilities; our findings are based on the participant
observations which were conducted only during the visits to the facilities. However, the
strengths of our study are that, the findings have been discussed with health workers of
different levels of the health sector; health facilities, district, regional and zone levels.
In addition, we have observed the fields from different types of health care provision
(health centre and dispensary and in rural and urban geographical settings) and the
fieldwork had been conducted for a long time. Also, comparing the findings to practice
based perspective discourse, makes it possible to transfer the contribution and
implication to other health information system researches.
4. IMMUNIZATION INFORMATION PRACTICES, ARTEFACTS,
CHANGES AND CHALLENGES
First we describe the general information flow and activities around the immunization
activities, and then the activities that specifically address the children which do not
receive the scheduled vaccinations (defaulters). Our analytic interest is in this practice
of defaulter tracing, which consist of several sub-tasks. The health workers need first to
identify defaulters and then trace them in their home. The outcome (whether the child
was found and vaccinated or not) must be documented and reported to the facility. We
describe first how defaulter tracing was done in the beginning of our study, and then
how it was done after changes to the information tools that the health workers used in
October 2012, and finally after new vaccines were introduced in January 2013.
4.1 Organization of Information Practices around Immunization services
before October 2012
Children would be enrolled into child health programs when they were six weeks old.
On the first visit, the nurses would create a child health card to be kept by the child’s
carer (the card is called RCH1) with e.g. vaccination information and the time for the
next appointment. In addition, the nurses would enter information for this particular
child into one row in the facility-based children register book. The nurse would enter
this on a page where the name of the street or hamlet of the child’s residence was
written on top.
The information in the register book was updated on every visit of the child (up to five
years of age), and was used both to generate the monthly reports to the district level and
for tracing defaulters. To identify defaulters, nurses would scrutinize the records to
determine if children were overdue in their vaccination schedule. They would make
notes about defaulters on a paper slip, where they wrote the name of the child and the
mother, the street/hamlet and the vaccination(s) not received. Then sometimes the
nurses would conduct home visits, or community health workers and community
leaders (whose name was written in the register) would be involved to search for the
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 9
defaulters. If they were found, the outcome could be that they did receive the service,
they could have received the service in another facility (documented in their RCH1
card), or they may have migrated or were deceased. Information about this outcome was
included as comments in the register.
Clients move between facilities for multiple reasons, and health workers recorded
information both about residents and non-residents within their residential population
area. Even though they were not required to report on non-residents, and coverage was
only calculated related to the residential population, they had to account for vaccination
usage and needed to document also the non-residents that received services. When we
started our study in 2011, they did this on a separate page in the register book or in a
different register (in the facilities that had many non-residents). However, the tracing of
defaulters was focused on residents. Here we see that the multiple practices intersect
and shape the usage of the information artifacts: recording, tracing defaulters, reporting,
resource management, and improving vaccination coverage. Thus information artifacts
such as the child register have multiple roles and a multitude of artifacts play together.
While the health services offered to children encompassed vaccination, vitamin A
provision and weight monitoring; the tracking of defaulters were done only for
immunization services. Two health workers had the following to say regarding tracing
defaulters activities:
“We trace defaulters only for immunization services because for other services we do
not have their information” (informant 1).
“In reality we are supposed also to track stunting and underweight children and those
who have not come for subsequent immunization services. But we cannot know who is
who, because after measles vaccination, information is registered in the card which
mother goes with and not in the register which remain in the facility” (informant 2).
The excerpts from health workers show that the design of the register afforded some
usages and not others. The next section presents our observations of how these practices
changed when there was a change in the official system.
4.2 Introduction of new register books and vaccines
During 2012 the Ministry of Health implemented a new version of the national health
management information system where both procedures (guidelines) and information
tools were changed. Also, the Ministry of Health introduced new vaccines (Rotavirus
and PCV13) in January 2013.In the Dodoma region new registers were introduced in
facilities from 1st October 2012. The new children register now contained more
information than before. The layout of the register indicated that client information
would no longer be organized street/hamlet wise, but be recorded sequentially. This
implied that there would be no separation between residents and non-residents, and that
the clients on a certain page would not necessarily have the same residential address.
For instance, the area of residence and the name of the street/hamlet leader, which
earlier were written on top of a page; was now supposed to go into a new column. The
rationale for mixing residents and non-residents was to improve reporting procedures as
explained in the following comment:
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 10
“In training of new system we were told that the new practice would improve unique
registration within the facility. Hence it will facilitate the process of determining total
number of children registered at any time by looking the last registration number
instead of aggregating” (informant 3 at district level).
Facility one is situated at the heart of town; near the commercial center and with good
access by public transports. This facility was one of the busiest facilities in town and
could serve 150 children per day. Most children were non-residents, and the staff had
previously collected information in two register books, one marked “residents” and one
marked (outsiders). When we visited the facility in the end of December 2012; there
were complaints from the health workers about the new registers. The format of the new
register stipulated that residents and non-residents would be collected together,
however, with a high number of non-residents this made the required calculations
difficult: “We have few clients since most buildings in our catchment area are for
commercial or official use”. Most of our clients are outsiders. Look at vaccination
report for December 2012, how can you extract information for residents in group of
more than 300 clients for BCG?”(Informant 4) “Now hamlet name and hamlet leaders’
information is supposed to be recorded into one cell. The space is too small” (informant
5). It seems that there was an intention to change the tool: “I went to store but there
were no spare children register. We are waiting when there is new spare register, we
would move back to ‘insiders and outsiders’ data recording which would facilitate
tracing defaulters” (Informant 6). In the next field visit in January 2013, the facility had
started to document in two registers again, one for residents and another for nonresidents.
When the new vaccines were introduced in January 2013, this facility introduced a
separate register book for recording these vaccines. This implied additional work of
entering the child’s name into another register book.
Facility two is located in the outskirts of town. Previously, the registration had
happened in one register book, with one page allocated to each street and one page
reserved for non-residents. This facility was challenged with problem of achieving
immunization targets:
“We have problems with achieving the immunization target. One of the reasons
is that, the targets set by district from census are higher than reality (Facility Incharge).
A similar comment was given during the interview by one district coordinator: “Among
the facilities who are not achieving their vaccination immunization targets are Maji and
Kisima. I think they share common problem that clients could move to other facilities”
Because of the awareness of the requested immunization coverage, tracing of defaulters
was a strongly institutionalized practice. When the new register came, the nurses (in
consultation with facility management) changed the register format immediately. In
order to support defaulter tracing practices, the use of the register followed the old
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 11
practice, i.e. the street name was put on the top of a page. The new column intended for
street name and leader was used to record street or hamlet leader’s name only.
When the new vaccines were introduced in January 2013, this facility changed the way
they used the existing register. They stopped recording hamlet name and leader in the
column intended for this, and used the space for recording new vaccines instead. One
health worker commented that “we know the names of our hamlet leaders”, and we
could also observe the list of hamlet leaders displayed on one wall in the facility.
Facility three is one of the biggest facilities in the Bahari district and it is serving
Mabwawani village. In most cases the children that attended were from the village. The
facility had developed its own information infrastructure, which consisted of local preprepared clinic schedules, facility daily meetings and RCH facility reports.
The facility had organized its clinic schedule streetwise for the purpose of managing
client workflow, predicting the workload, and as a way to trace defaulters on a daily
basis. After the new registers were introduced, the use of the old registers stopped. One
day the first author overheard the following conversation which took between two
health workers during vaccination services: “What about follow up? We are not telling
clients to inform defaulters”. “In the new registers addresses are mixed up. There is the
additional task of searching then announcing to the clients from a certain street”. Later,
around two months after the change, one nurse commented to us the following:
“Today it was internal supervision for data management. We noted in the
meetings while examining each page for the register that some children had
defaulted and follow up had not been done. It was discussed that it is because
clients are mixed across streets. We have decided to continue using the old
registers because it is easier to search and hence make follow up using clients”
(nurse).
The facility stopped using new register and moved to back to using some unused pages
in the old register book. However, health workers were worried what would happen if
district supervisors will come. In order to balance the requirement to use new register
and tracing defaulters, in January 2013 they moved back to new registers but changed
the format. They transcribed all data registered in the old books into the new register but
it was entered into the register in a one-page-per-street format.
When the new vaccines were introduced in January 2013, the facility introduced a
separate register for recording rotavirus and PCV13 vaccines. However, after a while
the health workers in facility three found that it was tedious to record in multiple places,
and in March 2013 they decided to use the “Remarks” column to record new vaccines in
order to record in only one place.
Facility four serves the clients of Mabwepande village and most clients are residents.
The facility has organized the children register into pages which are allocated to streets
and one page for non-residents. Also here defaulter tracing practices were very much
institutionalized, drawing on the assistance of community leaders. One time we
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 12
observed the village executive officer requesting names of defaulters, and when we
inquired about this he said:
“I usually ask names of defaulters and afterwards I send SMS to hamlet leaders
who will trace the clients. Or I can give names to community guards to pass to
hamlet leaders. I started this behavior two years ago because previous district
commissioner wanted me to give frequent reports about education, health and
other activities. Now it has become one of my duties”.
In contrast to the other facilities, this facility used the new registers as intended and had
not attempted to change it during our period of study (until April 2013). This might be
because the facility receives few clients. For example in the new register there were
only 23 clients from October to December 2012 and 40 registered clients from October
2012 to April 2013. With few clients it was easier to search for and identify defaulted
clients, even with the new format. We observed nurses twice when they were searching
for clients. However, though there were few clients on just one page of the register, one
nurse erroneously termed someone as a defaulter while the date was not yet due.
For the new vaccines, also this facility used the ‘remarks’ column to record the new
vaccines, which they had learnt from the nearby facility three’s practice.
5. ANALYSIS
In our study we have followed how practices of information recording and defaulter
tracing change across time and space together with the local information infrastructure.
In this section we will compare findings across the cases and discuss how they were a)
situated, i.e. related to the features of the context, b) how they showed a temporally
evolving trajectory as the participants sought to achieve convergence, and c) how they
had an infrastructural quality in that they were part on a heterogeneous, underlying,
supporting substructure for work.
5.1 Situated nature of objects and practices
The contextual differences between the facilities shaped their information needs. The
two health centers (one in each district) had a high number of clients compared to the
two dispensaries, which served a smaller population. With a low number of cases to
monitor, less sophisticated techniques can be used than when the numbers are higher.
Hence, one would expect that the practices differ between these sites. In addition, the
social cohesion and demographics of the residential population area had an impact both
on how the service provision was organized, on how big a problem defaulting was, and
on which means were available to address it.
In the rural areas, neighbors who lived in the same vicinity would know each other, and
when they were called on the same day and time, the health workers could therefore ask
the group about the whereabouts of the mother or child that did not show up. This is in
contrast to the situation in the city where the health workers could to a lesser degree
assume this knowledge to be present. Consequently, there were good reasons for not
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 13
attempting to do defaulter tracking the way they did in the rural areas. Meanwhile if we
acknowledge how practices and objects are culturally, socially and historically situated
(Nicolini et al., 2003; Nicolini, 2013) we should not take for granted that it is sufficient
to just draw a distinction between city and rural areas. When looking closer at the case
we actually find that the rural population in this district would regularly migrate to other
locations during the sowing and harvesting seasons, so defaulting was in itself a larger
problem than it was in the city. Needless to say clients were also moving in the cities,
but in such cases mainly on a permanent basis rather than for shorter periods of time.
They would therefore then belong to other facilities. Moreover, when the health worker
relied on the local authorities, such as a street leader or village head, to trace defaulters,
these officials would posses more power in the rural than in the city context. Hence
being part of fairly similar practices in different contexts had different effects in our
study. The location of a facility (the accessibility by public transport) impacted also the
mix of clients, whether they would be mainly residents or non-residents.
5.2 Evolving design of information artefacts to achieve convergence
The issues mentioned in the paragraph above all impacted the information needs of the
health workers. When we started our study we observed that they had modified the
official information artefacts according to their perceived information needs. This
illustrates that a process of seeking convergence (Star et al, 1997) between the
information needs of the practice and the informational affordance of the artefact, had
happened. When the health authorities introduced new information artefacts and
guidelines in October 2012, we could observe how the health workers reacted to this
change of their established practices. For some (the small facilities with few clients) the
changes to the format and layout of the register books were not perceived as a problem
– they could maintain their practice in either way. For other facilities (refer the case of
facility one and three), the change constituted a disturbance that triggered a process of
seeking new convergence. This implied a “re-tooling” of the practice, where new ways
of organizing information was employed. When the official tools were changed, some
health workers would comply with the new requirements, and then modify them as they
discovered that it undermined their practice, while other modified the tools
immediately. This illustrates that the process of emerging information tools have a
temporal nature, and that one needs to apply a historical sensitivity to the analysis of
such processes. The specific nature of this process depends on the contextually shaped
information needs as described above, as well as the resources available to the workers
(for instance whether they have access to register books, know the community well etc.)
The findings also show that each facility has learned how to organize its information
management using information artefacts which support situated concerns. It is a form of
experiential learning (Holmqvist, 2004) acquired through working. For instance, facility
1 had learned that information tasks are easier with a separation of insiders and
outsiders. Facility two has learned that the previous information organization helped to
track clients. Facility three had learned how to use informal information artefacts to
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 14
support the official information tools in order to organize work. The practical skills are
recognized by members of the facility and it aims at making work being accomplished.
5.3 An Infrastructural Lens on the Role of Objects
If we take an infrastructural lens on the information artifacts in use, we see that they
could work in the background until a disturbance occured, after which they were
foregrounded. The breakdown that occurred when the artefact changed pointed to the
taken for granted nature of it, which is typical for an infrastructural resource. Another
feature that this lens helps us see is that they are not standalone entities, rather they
work together with other surrounding artefacts. For instance, since a list of hamlet
leaders’ names hung on the wall, the health workers didn’t need to register it in the book
and could use the space for something else. The tools are serving multiple purposes,
both the health workers primary work of providing services in a routine way and
specifically the task of tracing defaulters, and the secondary purpose of reporting
information to the authorities. The information demands for the defaulter tracing were
different from, but not incompatible with the information demands from the other two
practices (documenting the service provision and reporting). Some layouts of the
information tool facilitated this, others not. The information artefacts that support a
multi-purpose information infrastructure must themselves support multiplicity.
6. DISCUSSION
In the previous section we described three dimensions about some of the findings across
the cases. By doing so we look at how objects as information artefacts play an important
role in organizing activities and practices. The earlier works on objects focused how
boundary objects succeed in maintaining coherence across multiple social worlds. The
focus was in other words on how such objects enabled them to overcome what Carlile
(2002) describe as semantic boundaries (Swan et al., 2007: 1829). By taking a practicebased approach to objects the contribution from Swan et al. (2007) was to show how
object also enable collaboration across other types of boundaries, even in cases where
the actors have different interests. In our study we have to a lesser degree looked into
the politics around these objects, even though we observed as mentioned in the previous
section, that in some settings a difference in status between the actors that handled the
objects.
Nicolini et al. (2012) provides a very useful overview of how different theoretical lenses
enable us to understand objects; activity theory, epistemic objects, boundary objects and
material infrastructure. In our case the objects that we have studied did not play the role
as epistemic objects motivating the different actors to collaborate to generate knowledge
and triggering innovation. We are instead more concerned with how tools worked in
order to support the practice of immunizing children in a resource constrained setting,
and in particular how they worked as infrastructures. In line with Nicolini et al.
(2012:13) we find it likely that objects take this role in cases “when boundaries are
uncontested and fall into the background”. In our study we found that:
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 15
i.
The local context varied among the different facilities in terms of challenges,
involvement of community leaders, district supervision, and clients’ behaviour. The
objects’ roles were shaped by the alignment with these contextual factors. Hence,
there is convergence (Star et al., 2003) between multiplicities of roles, the nature of
facilities or boundaries existing within particular facilities and its information
infrastructures. These infrastructures are heterogeneous networks where objects are
not only connected to other objects, but also to routines, resources, skills and so
forth, spanning beyond the local, both in time and space, exemplified in our case
with how schemes and books were produced by the government.
ii.
There is a relationship between the multiplicity of roles the object has to perform in
certain context and the change process of that object. The objects could be enacted
due to conflicts between different roles the artifact have to perform. Sometimes
there was tension between recording, reporting and tracking activities, and changes
were done in order to resolve the tensions. This is similar to Lindberg and Walter
(2013) who argue that objects are enacted in the organizing process, when actionnets are formed and that new practices could be established in the organizing
process.
iii.
The multiplicity of roles the object has to play can change over time, as there are
changing contextual concerns or other supporting tools in the information
infrastructure. If the ability of an artifact to perform its multiple roles is hindered,
this is a trigger to action (changing the usage of the object). This is in line with
Lindberg and Walter (2013) who argued that organizing as the construction and
maintenance of action-net.
We see that a core demand to the information artefacts were that they should present
actionable information. To analyze the way the artefacts succeeded or failed in this
required a more granular attention to the material qualities of the artefacts: how the
layout, the organizing principle of tabulation, in other words the ‘information
architecture’ of the tool itself was changing. This interrelated in different ways with the
practice in which the health workers were using the tools. As we recall from Norman
(1993), artefacts could have different types of representations ranging from offering
surface representations to those that offer internal representations. These artefacts
should contain enough quantity of memory and be precise enough to help those using
them. We observed that some of the changes of the tool had such consequences as to
shift the ‘computational load’ of identifying defaulters that the practice of tracing them
temporarily stopped. Moreover, any practice is also situated, and the actions with and
on an object, were shaped by the local contingencies of action. Some of the adaptations
were temporary and local, while others were shared across context (learned). In some of
the facilities, the health workers offered the tools extensive explicit attention, meaning
that was usually was secondary or even tertiary objects, changed into primary objects of
attention.
A central observation is that objects’ enabling and limiting characteristics are important
in the organizing process. The study was conducted within a larger initiative where
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 16
computerization of the health management information system was an aim (both from
donors and ministry of health), and the study has implications for design and
implementation of health information systems in Tanzania. The current information
tools in primary health facilities were all paper-based. Health workers were able to
change the tools to fit organizing of activities due to paper technology characteristics.
Paper offers certain possibilities; such as malleability, low cost and relatively high
availability (Sellen and Harper, 2003), while its constraints are related to the limitation
with respect to searching and storage. The changing of these tools to fit the work
practices does not require novel skills, but draw on the skills acquired during the work.
With computerized systems, the possibilities of artifacts would be different, related to
display, searching, shareability and storage. However, a limitation of computer based
technology is that it requires special skills to manipulate the technology; skills which
reside with information technologists and not with practitioners. The message from this
paper, to emphasize the material form of core information tool and the impact this has
on modifications and improvements, is crucial in order to temper such computerization
efforts.
7. CONCLUSION
Here we have studied information artefacts that play a significant role in enabling or
constraining what the first line health workers can do related to follow-up of defaulters.
Our case analysis has demonstrated the co-constitutive nature of artefacts, context,
practices and organizational forms, and it contributes through shedding light on how the
organization of work can be crucially dependent on information artefacts and the
multiple and temporally shifting roles they play in situated practices. We have also
argued that in order to understand the health workers’ actions and adaptations of the
tools, we needed to work at a level of empirical granularity that to date has been fairly
uncommon in organizational studies. More specifically we needed to understand the
changes of the layout of a page in the register book, and chose to utilize Norman’s
concept of experiential or explicit computation to explain why one layout worked better
than the other to support the practice. We also aim to contribute by indicating how
objects as part of infrastructures are not only something that becomes visible when there
no longer are challenges, but can be a visible part highly connected to heterogeneous
networks of objects, routines and practices. Changing the ways objects are part of
organizing activities can also have implications for roles and status of the actors
involved, and how such processes evolve is an area that could be further investigated.
8. REFERENCES
Ancona, D.G. and Caldwell, D.F. (1992) Bridging the boundary: External activity and
performance in organizational teams. Administrative Science Quarterly, 37(4):634–665.
Barrett, M. and Oborn, E. (2010) Boundary object use in cross-cultural software development
teams. Human Relations, 63:1199-1221.
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 17
Bechky, B. (2003) Creating Shared Meaning Across Occupational Communities: An
Ethnographic Study, Organization Science, Vol. 14(3):312-330.
Blomberg, J. and Karasti, H. (2012) Ethnography: Positioning Ethnography within Participatory
Design. Simonsen, J. & Robertson, T. (eds.) Routledge International Handbook of
Participatory design Routledge New York, 86–116.
Bowker, G.C. and Star, S.L. (1999) Sorting things out: Classification and its consequences.
MIT press.
Braa, J., Monteiro, E., and Sahay, S. (2004) Networks of action: sustainable health information
systems across developing countries. MIS Quarterly, 28(3):337-362.
Braa, J., Monteiro, E., Sahay, S., Staring, K., and Titlestad, O.H. (2007) Scaling up local
learning: Experiences from South-South-North Networks of Shared Software Development.
In IFIP WG9. 49th International Conference Taking Stock of E-development.
Bryman, A. (2008) Social research methods. 3rd ed., Oxford University Press Inc, New York,
Carlile, P.R. (2002) A Pragmatic View of Knowledge and Boundaries: Boundary Objects in
New Product Development. Organization Science, 13(4):442-455.
Carlile, P.R. (2004) Transferring, translating and transforming: An integrative framework for
managing knowledge across boundaries. Organization Science, 15(5):555-568.
Fennel, M.L. and Alexander, J.A. (1987) Organizational boundary spanning in institutionalised
environments. Academy of Management Journal, 30(3):456–476.
Flyvbjerg, B. (2012) Five Misunderstandings About Case-Study Research. Qualitative Inquiry,
12(2):219-245
Gherardi, S. (2009) Introduction: The Critical Power of the ‘Practice Lens’, Management
Learning Journal, 40(2):115-128.
Henderson, K. (1991) Flexible sketches and inflexible databases: Visual communication,
conscription devices, and boundary object in design engineering. Science, Technology, &
Human Values, 16(4):448-473.
Hernes, T. (2004) Studying composite boundaries: A framework for analysis. Human Relations,
57(1): 9–29.
Holmqvist, M. (2004) Experiential learning processes of exploitation and exploration within and
between organizations: An empirical study of product development. Organization
science, 15(1):70-81.
Huizing, A. and Cavanagh, M. (2011) Planting contemporary practice theory in the garden of
information science. Information Research Journal, 16(4):497 – 512.
Jordan, B. and Henderson, A. (1995) Interaction analysis: Foundations and practice. The
Journal of the Learning Sciences, 4:39-103.
Kaasbøll, J., Chawani, M.S., Hamre, G.A. and Sandvand, J. (2010) Competencies and Learning
for Management Information Systems. Journal of Information, Information technology and
Organizations, 5:85-100.
Knorr Cetina, K.D. (1997) Sociality with Objects. Social Relations in Postsocial Knowledge
Societies, Theory, Culture and Society, 14(4):1-30
Levina, N. and Vaast, E. (2005): The Emergence of Boundary Spanning Competence in
Practice: Implications for Implementation and Use of Information Systems, MIS Quarterly,
29(2):335-363.
Lindberg, K. and Walter, L. (2013) Objects-in-Use and Organizing in Action Nets: A Case of an
Infusion Pump, Journal of Management Inquiry , 22(2):212– 227.
Miettinen, R. and Virkkunen, J. (2005) Epistemic objects, artefacts and organizational
change. Organization, 12(3):437-456.
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 18
Miles M. and Huberman, M. (1984) Qualitative Data Analysis, Sage Publication, United States
of America.
National Bureau of Statistics (2010) Tanzania Demographic and Health Survey 2010. Dar es
Salaam, Tanzania.
Neyer, A.K. and Maicher, L. (2013). Understanding the Role of Objects in Interactive
Innovation. In Wirtschaftsinformatik, 537 -549).
Nicolini, D. (2009) Zooming In and out: Studying Practices by Switching Theoretical Lenses
and Trailing Connections, Organizational studies, 30:1391-1418.
Nicolini, D. (2011). Practice as the Site of Knowing: Insights from the Field of
Telemedicine. Organization Science, 22(3):602-620.
Nicolini, D. (2013) Practice Theory, Work, and Organization: An Introduction, Oxford
University Press, Oxford.
Nicolini, D., Gherardi S. and Yanow, D. (eds.) (2003) Introduction: Toward a Practice-Based
View of Knowing and Learning in Organizations. In Knowing in Organizations: A Practice
Based Approach M.E. Sharpe Inc. New York USA.
Nicolini, D., Mengis, J. and Swan, J. (2012). Understanding the role of objects in crossdisciplinary collaboration. Organization Science, 23(3):612-629.
Norman, D.A. (1993). Things that make us smart: Defending human attributes in the age of the
machine. Basic Books.
Orlikowski, W.J. (2007). Sociomaterial practices: Exploring technology at work. Organization
studies, 28(9):1435-1448.
Rheinberger, H.J. (1997). Toward a History of Epistemic Things: Synthesizing Proteins in the
Test Tube (Writing Science).
Schatzki, T.R. (2006). On organizations as they happen. Organization Studies, 27(12):18631873.
Sellen A. and Harper, R. (2003) The myth of paperless office. MIT press
Star, S.L. (1999) The ethnography of infrastructure. American behavioral scientist, 43(3), 377391.
Star, S.L. and Griesemer, J.R. (1989) Institutional Ecology, 'Translations' and Boundary
Objects: Amateursand Professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39,
Social Studies of Science, 19(3):387-420.
Star, S.L. and Ruhleder, K. (1996). Steps toward an ecology of infrastructure: Design and
access for large information spaces. Information systems research, 7(1):111-134.
Star, S.L. (2010) This is Not a Boundary Object: Reflections on the Origin of a Concept.
Science Technology Human Values, 35(5):601-617.
Star, S.L. Bowker, G.C. and Neumann, L.J. (2003) Transparency beyond the individual level of
scale: Convergence between information artifacts and communities of practice. Digital
library use: Social practice in design and evaluation, 241-269.
Suchman, L. (2000) Organizing alignment: A case of bridge-building. Organization, 7:311-327.
Swan, J., Bresnen, M., Newell, S. and Robertson, M. (2007) The object of knowledge: The role
of objects in biomedical innovation, Human Relations, 60(12):1809–1837.
Trompette, P., and Vinck, D. (2009): Revisiting the notion of boundary object. Revue
d’anthropologie des connaissance, 3:3-25.
Walsham, G. (2005) Interpretive Case studies in IS research: nature and method. European
Journal of Information systems, (4):74-81.
Zeiss, R. and Groenewegen, P. (2009) Engaging Boundary Objects in OMS and STS? Exploring
the Subtleties of Layered Engagement. Organization, 16(1):81-100.
Proceedings of Organizational Learning, Knowledge and Capabilities Conference. Oslo, Norway. 2014. Page 19
Appendix 4: Paper IV
Shidende, N.H. and Mörtberg C. (2014). Re-visiting Design-After-Design: Reflecting
Implementation Mediators Connectedness in Distributed Participatory Design Activities,
In Proceedings of the 13th Participatory Design Conference: Research Papers-Volume 1 (pp.
61-70). ACM.
IV
#!
$%&'&()*+,$
-.! !"
/3 < 8 $
6.
=8$6> = . ? $
99> ! - . ! $ -!.6%
[email protected] '>!A.!-
!1 A.. . .1 B
6.=6>+
-.
4!
.! / - ! '' 6* . !
!= . 8C.''> 8%[email protected] '>
* ! .! :
/ * !. . 0 . .
- ! . -!
!.!!
1 -!.-
. .. -! . " "
* / 1 . !! . . -! 2 . /3 ! 2
. ! . .-
! .- . .! ! ! *4 .!
. " ..
5 -!.
.. 5 .. .!
-!. ?. =99 @'> !
A..1 / !
.
-! ! /. - .
.! ?. 1 =->
!..
!
"#
-! 6 . 7 .2. 8 $
. - . !
D..1 ! . =?
$99> .-!
! 6 . 1 !! " =C. 8" ''> . - .! - =?
-. ''> .
! . =#. 2"
99> / . ! ? 6! =9'> / =9'> -!
. 1!-
" .- /
"#
9 : . 7. 8!
; 8 & :
6.
$ %& %$
/ 1 5 !" 6
"."
!.!
-! . -!. B.
" - 2B7 ! -
2-. / !- -! E! F GE! 6
. 6B0H-9H5"
-
B.9H2B7'[email protected]&99&FHFJ& :FF* .F H&F9H& 9HI % . #!=9&>!
2 ! -! 6
- ". #.1 2"
1=99>
*.1
5
- . ! #.
2"
=- '>.!
-! A
. A ;! .
! =$!
99> - !
! . 8 /3! 8$6 .- " 5 ! . /
- -! . =6! 9'4 / 9'4
? $ 99> 61 !
-! . ! . . 1-!6+
- ? $ =99 9HI> 3 . -
A E!
- ! 1 /!
-
.3 ! -! .
8!"
!
! .
< .- . =/ 9'4?
$ 99> / ! . . /! - 1 . -! .. 5-!!6
! . #. 2"
=99> ! . 1 ! /
! - A.. !
-!1=%[email protected]> -!
.-/
/ $!$! -
8$6$!2''H<&=?$
99> ?$=99 9>-8$6
" .- " ! L- $! / .- " ! $! 2 + + !- " / -8$6 !
6 . !. $ 6 / - ! . 8 $
=8$> ! -! .
" .
!-..
!- 8$6 =? 9H> #. 2"
=99 9> 3 .
!! -
! . . - " . !. =$, '@> /! * - ! #.1 2"
1
! . !
1 .! %1
[email protected]>A....1 /!! *
!- ! - . -. !! ! $
. / " 1 ! - ! # ! 8$6 - $! 2 - "-!. 61
: . / . $! 2 ! 9H /
-!--.-
!.3=/ 9'> /61
8$6-
. ". . =9<
9> . --
=9<> 6.
-8$61
-
E!L-$!=?$
99> / 6<8$6 3-
"
!
- -
.. - - ! -
! -. - .- 8$
" /! ! %'
()$
*
%$ $%&
$
(+%,& %$'
%(
(
/ .
1 . !
.
$
!'I=?K" 9H4C.
L-!
99> 6
=C.L-!
99 >!
6 $! !! . ! 6 =- > .
.3 6
! B
!- . =$ 99>-!6 . 8
$
6.
=8$6> *
6 . - =? $ 99> 6
61!!
.*=? 99> 9
2!. 8$6 .-" . . -"
! . !!."=$!
99> /!-
((-
(
!
-!
*
!
/ ! - ! .C
9/3
3- ! /8$...
/3!
... . / !! 8$6 /3 *
!8$ $)
$
%-(
8 $
6.
=8$6> . ! 8 $
=8$> /8$
8$/"
!!.
-!- / ! ! . !
!. !.
- . ! .
;! ! . 8( .
= . ; 9I47 99>
!
"..! -!
.-/
!
0
/ 8 $
=8$> .''I$!2=?
$ 99> 2... ! .
! 8$ 8$ .! - -! 9H 8$ 9 8 8$ 9 .
!
H!- .3!
8$ !
- ! ! 8$ ... 8$ - 8$/" ! / 8$
... ! 8$
.
... !!!
L$ - - / 8$
- ! . 8$ " . ! ! .
- -
8$9 8 "1 ! - 2!.! 9 - 9 2 . !"
. 8$ /" 7! I-
! / 8$/" -..
-99 /
! 1 4
. 7$ ! % $
+ - 99 5 ! -9929!.
!!.
! 8$ . 2 9 ! .! ! . " 8$ "
/ = !> !
8$ 2!.!99
-9 $ . .! !
" / - . !. !.. . 8$
" ! 8$ " -!'
-!.8$
"1 . / .! . 9 H
! / !
-
" ! !/" / . 8$ " . .! /
.8$1!
!
3-!1. / 8$ /" ! . - - !. * / ! -!. .. ! ! !3 . . ". -=L399> /!
E!...-
.
E! . ; 8$ "1 ! !..
".
! .
- !.
- /!.-
! ! * .
! !-! 8(F2$
. .! ! /
8$"
-!!
L!?!7/3. 2 " !. !
! !
. ! . ! ?%.C-!
/ . ! . ! .! ! ! C
!.!./3 #
. 9
2
"1 !. 11 .
-. /.!
!-! ".
2
! F
!. - " ! F " !.=.-!> %. / .- ! %. !
! "! (!
/ !-
-
1G6!-%
B
M! 9 % " !. " *
!. ..
"
-. " !. - " ! =
8$ "> .
! - - . !
! / : ; ! -
.
!
-!
-!
- /.!-
.
;
" ! -
! .
. / 1
! - !. - / - !
! !. .
-!1!!
.
/ !1 ! . <
. 5 ! ! * 6!
!
! ! !
$!*-
!!
A ! .! -! .
N N
! ! !. ! ! *A =B
-
%!-"=9H I9> 2
6 !. - .!
. " - !! =G-
5.99 &> 8!.!-
!.-
!
.
". ! ! - . .
=.> "
!=!>!
! => ! !! => -! ! ! - !. .
". ! .
". " ! $7$ -! ! -
"(1
()$
%$ (2 /8$"9 '-
. /3 =.
99> / ! / " . ! / !
3 - !
3.
-
. -! - . ! .
; *
!
3 ! 1 !-E!
- 7E!
!.-!1
;*
/3
!3
""H"'
/
8$ /" ! /3!.
.:FFH"-
. ! ? . .
8$ "
! - 1 !3 ! ;! !
. ! - !
.
- !
/ $7$ .
! ! -! !!
!
! /!-.8$"
!
3 .! "1!8$"
/ 1 8$ " ! -! ! .! .!. - .. .
"! /
.!- ()$3 (()$
/ 1
. ! .-.-.
8$6.-" /"
.!.-!!
8$ /" " .3 .. %[email protected]>
A....1!!
1 -<
< .- . / . / !. " ! ! -" 5 !
-.
H
!
,
(!!
3
! 5
* 9 - 99 " ! .. ! ! - * %* .
!-! /!!
1 : . - !
- ! =.
. !
> /.
!!-
.
!
5 .
!
! ! ! -! .
!!
-F
.
!
)
*
/ . ! ! -
2!.! 9 M! 99 7 !.8$/"
!.
*.
.
! 8$/" . / . !
. ! .
8$"! .!.
- " ! / -
! " ! ! -
.-!!
" : - . . ! . ! ! -!.
/ 8$ "1 - ! . . 8$
" ! - !3 .
- . !
2$7$!!
8$ " ! - .
!
- - !
/
!-
- ! ! - -
.- =B
- %!-" 9H> -
!..-
-.-*-!
.- . . 8$
" ; *
- .-
. .- * ! !!!.
.
/!!
. = . . ! >
.
! ! . . .
/ * . /.-*
!.
/!!
! !
- . !3
!
!. .
!
/.!
! . -!
.!.!.
8 *. 8$ ! .
/ 1 *. 8$
- .!
E! ! 8$ " ! . ! /
! ! - !. ! ! !
"
!.
4
4
*
3
!
2 !
- - .!
/
.!!
! -
1 F!
! !! =
- > ! 5 " - !.!
; !
. ; 9 ! - ! . -
! ! . !
.
/ !. E! . !
" .-
.8$/" /
8$".!..!
. / 8$ .
!
5
. .! !E!
.-
. -
!-.-.-=?
-.
''> 5..
!-.
/
!
!
.
.!--
! - !
5
.
"
,
)
-
0
2!.! 99 ! ! -! ! ". &
!-!! 2;9
!. -!
1 . 8$ " 5 ! - . !
. : . . .
! !. ! !.
- 5 8$ " ! .
!. ! - .
". 5 !
! - . 8$
" 5 8$ " !
! - . $7$ !.
!
-
;
" .
8$/"
-".
..8$"
/ -! * ! ! " !!
-. /
" !. /. - -!. ! ! !!. /.!-
!
=? 5 7$ * > / .
-! !.
! !. 8$
" ; *
7$ %* ". !1 - . !
-!8$
"--
!"#$
A %
&'
(
)
* +
& 1
+-
.8$"!
- .! - -
-! " ! .. -! ! ! !.!
.3 . *
!
/ ! ! . /6
- .. -
!! ! !- =G- 5. 99> / ! .- -
/! ! .-.-=#.
2"
99> !" $
A+ ' & &
,&-
1
%* "
!8$"
!
. !! ! . "1*!
* ! -! *.
. /! " . * ! !.-
!
! !
-! %* .
= > ! ! +!
". -
0
&
-!
3
. ;*
""
. .
* . /
-.!
-
" . ! .8$"!!
.! ! " . / ! .! !29 / . *. ! 8$ " . -! .
.-
;
! ! " . !
67/B/=67/B/
8(> !.
. 8 67/B/
!8$"
/
* - " ". 67/B/:
;!.!!
- ! . " . . / !.!!!
-
!8$
"1 !. ?! !
". E! !. . ! -
!" $
.+ / !")$
) % )
*
A)
.+ / 1
/ * ! ".
67/B/ /
" !.. . ! F67/B/.
/ ! . . !.
=!. * !.
.> ! * ! .! * !
- =%
[email protected]> /
!.
1!-!
* *. ! ! -
=$!
99> / ! ! . -.-.
! - ..
!
!
! ! ; 67/B/
.
! " /
!.!
!" /.
8$ " . ! ..
/ ! -! * - ". -! . ! - 8 * ! ! -
. ! . 5 - . ! !...:
! !.. . . .
- *. ! .
! ! -! 67/B/ ?. ! !. . .8$"
! % !. ! *
! .! .. - E!
. !
.- . ".
1 ! E!
.-
/
!*!- !")
$
A 0
'
%* ! " .!1-!.
1 - =
!> "4 . . 8$ " / 1 1 . 1! /!...!
!. ! E! !. . ! ! - !
! / . . ! !
!. 1 =% [email protected]> ?!
.. 1
1
=- > E! .!..
=?K" 9H> 2+
)
&
3 1
;
3 .!. 8$ " ! " 3. ! /.1
!".
. 8$ - ! - -! ! =5> /
A -! 4 1=5/! 9 9> / 1 ! *
!. ! - ! < ! =B
- %!-"
9H> .
.-!
-
!
-
- 8$". 8
-
!
-
.
.
. 67/B/.
. 1 - !
- /! -!!
- -! * ! :
/3.!-
I
-.-
! / . 1 ! BE! -!.
-!
-
.
;!
-! ! -
!
- : .-!
. - / -! .. ! .!
#. 2"
=99>.!!
. . !.*
8 -! .8$
.-!!
-"!.$7$
.
- - .- !...! 2 9 ! -! . 8$ ! ! $7$
.
! !
-! . .
/8$"!
! /3 BE! E!
3
<-.E!
! 3 . . - ! !! !$7$!
!
.
! 3
!
/ " .- . ! " !
. " .- / . !
!.-
. - . ! -
!:
.!
!
.!
! . ! - .3 . *
!
/ .-*-
!.
..
!1* 2 - .- !. 1 /
8$"
!
..!
.
E! - /
.- - / ! ! .. ! ! ! /
. - .- ". "
" &%$
/ ..
!. 1
. -!
!. /8$6"
8$
*.
!
/3 2 E!
! - ! !! 5.
!!
.- . / ! .! ! !
-.-
! - . =?K" 9H4C.L-!
99> ; *
*-!!
!
".-
.!-!!
- .. 1 .! "
! .
..- .- . "
! " ! !F
/
1 ! !
! * =B
- %!-" 9H>
" / . ! 1.
8$"
83-! . - . / - - .- /
-
! -.
- .*.!
.- /
< .- 1 3 . E!
2 $7$ ! .
! !3 *.
;! ! ! .- ! !. ! .. =6! 9'4/ 9'4?$
99> ?!.#.2"
1!
!. ! *
.! . ! A.. . .1 - 1=">.
!..=%[email protected]> 8.!A..
..1 @
.1 ". . !3 ! " . "
- ! / ! ! .. -
. ! . / .
E!!!
.-..
.! / ! - . . .-.!
- . !
-! . . . -
! -! . . .
! 1
!. 2!. " " -! !.
!.
0
=$!
99> % !. ! ! . -! *
. !. -
=$, '@> "O..!-!P
O.. ..P =% [email protected] '> ? $ =99> .! * .
E!
/ ! .! ": -
!
3 ". 8$ 5 ! .!
-! " A.!. .. 1 .! /!
< . ! .!
! 8$
" - !. - / - !
- . .!.
%$,&%$
. . 6 5 ! -! !
. ! . 5 ! . ! -! 6 . / - !"-!!.-!
!.-!
- -. " /! ! ! . -
" .- E!
#" !--
"
* . !
.- .2 / 8$9 8$" 7! - !
3 /
=> .
!
3 8$ " .".!
. !.. / . !. 8$ "
- . .- . 6. . ! ! * .! 1
*--!-+8$6"
. =% [email protected] ?. 994?$99> /"
. 1 .
!.. 8 1 !
. !. -"=#.2"
99> 6 . !. ! => -! . .
! - " . . -! . .- . -.
/"-
.- !. /"-!!
. - !. *
.!.. 2
. ! 6
-
! ! * ! !.
-!
!!.-
! .- . * -
:-
". ! .
. ; *
-
!..
-! !-!.
.-
1 . =#. 2"
99> /E!
!
- ! ! . 3 / . % =9> . ! . L- $! ! L- 4 !
.
/ .- 1 . 6 '
-
.
&24&/
56;5 "$%5,()($ 7-! 5 ": " ! =M /- C3
C"> 6+ % 8 !
! 82 C.7 /1"
:
!..! L-!
M QC.7 = > 4 (
'$ /
4 /
# %-!
2
M$2 ''9'H (3(($(
?
-. M # L
M 7 2 $
8 6 %. ; 7 /
G . .
4$
. . 8 M : #
%-!
2''9<& C.; 8"7 C L.(:
;!! 5" 7 . L-!
M Q C. 7 = > 4 (
'$
/
4 /
#
%-!
2M$2 ''&&I ? M $ $ 8 $
6.
: 6 . 8$6
" $
M Q G- / = >
G!. &
,
' .
4G!.R"999&& C. ; L-!
M 8.: . $
M Q G- / = > &
,
' .
4
G!.R"999 ? % ? / $ % / /E!: .. . ". . $
M Q G- / = >
&
,
' .
4G!.R"99H&@ #! ? $ 1
'1
6*9& ?. 7: .. . .! 6 . + $
M Q G- / = > &
,
' .
4
G!.R"99IHH 7$ $..
".
/3:.
32G/ &:./+ 99=9>:@9& #. 2 2"
R / !
!:
..
! .
.4/;5I ?K" C C. ; $
M .
& $ '
7/$2 9H .
B B. . B
!3 - /".
$
: 6 %* 7 8
B 6 & 1256;5/
99' B
- %!-" ( 7". $ -: 6! /. O!P
7. .
.4/5667&I 6! $ $ $ # M ?!. 6
8$ ": 2 B $! C &
#3
9''@ $ B B
2 6" ( B
! 6.-
! $
M Q G- / = > &
,
' .
4
G!.R"[email protected]' G- / 5. %: ..
$
M Q
G- / = > &
,
' .
4 G!. R"[email protected]& %6 6./. .
.4/5668
'9 $, 2 $ ,
' ? ?" R"'@
% 6 . ; / 7,-. B 5
S 5 ".S: *. 6 /3T3- $ .
.4/ 5669
&H $!
# # 2!- /.
6! :
&
99H=9>:'& ; 8 $; ! .
8( /!-!: 2 . /. !
. :+&9I'=H>:9'
/ $. C ? M -!
%-6:7!
.8$6":
&
9=>:9I& L3 2 2 7!"+ 2 # M $ $ . .! : 5/! ?. N 2
! .
.4/5667 I
Appendix 5: Paper V
Shidende, N.H., Aanestad, M. and Igira, F.T. The role of context in the co-evolution of work
and tools: a case from the primary health sector in Tanzania, under review for publication in
the international, peer-reviewed IS Journal.
V
Page 1 of 25
7KHUROHRIFRQWH[WLQWKHFRHYROXWLRQRIZRUNDQGWRROVD
FDVHIURPWKHSULPDU\KHDOWKVHFWRULQ7DQ]DQLD
$EVWUDFW
3XUSRVH±7KLVSDSHUSUHVHQWVDZRUNFHQWHUHGVWXG\RIKRZLQIRUPDWLRQV\VWHPVSUDFWLFHVDQGWRROV
EHFRPH VKDSHG E\ WKHLU FRQWH[W 7KH SXUSRVH LV WR LQFUHDVH WKH XQGHUVWDQGLQJ RI KRZ SUDFWLFHV DQG
WRROVFRHYROYHZLWKDVSHFLILFIRFXVRQWKHUROHRIFRQWH[WDQGEDVHGRQWKLVWRRIIHUUHOHYDQWGHVLJQ
LPSOLFDWLRQV 2XU HPSLULFDO PRWLYDWLRQ FRPHV IURP DWWHPSWV WR XWLOL]H ,&7 LQ UHVRXUFHFRQVWUDLQHG
VHWWLQJV
'HVLJQPHWKRGRORJ\DSSURDFK ±(PSLULFDO ZRUN ZDV FRQGXFWHG LQ SULPDU\ KHDOWK FDUH IDFLOLWLHV LQ
7DQ]DQLDWKDWRIIHU3UHYHQWLRQRI0RWKHU7R&KLOG7UDQVPLVVLRQVHUYLFHV)RXUKHDOWKIDFLOLWLHVZLWK
GLIIHUHQWRUJDQL]DWLRQDODQGVRFLRHFRQRPLFFKDUDFWHULVWLFVZHUHVWXGLHGXVLQJHWKQRJUDSKLFPHWKRGV
SDUWLFLSDQWREVHUYDWLRQLQWHUYLHZVDQGGRFXPHQWDQDO\VLV:HKDYHHPSOR\HGDFWLYLW\WKHRU\DVRXU
WKHRUHWLFDO IUDPHZRUN VLQFH LW LV H[SOLFLWO\ SODFHV KXPDQ DFWLYLW\ ZLWKLQ D FXOWXUDO VRFLDO DQG
WHPSRUDOGHYHORSPHQWDOFRQWH[W6SHFLILFDOO\WKHFRQFHSWRIPHGLDWLRQEUHDNGRZQZDVXVHGIRUGDWD
DQDO\VLVDWDFWLYLW\DFWLRQDQGRSHUDWLRQOHYHOV
)LQGLQJV ±%\IRFXVLQJDQDO\WLFDOO\RQVLWXDWLRQVRIPHGLDWLRQEUHDNGRZQLQWKHVLWXDWLRQRIXVHDW
ERWKDQDFWLYLW\DFWLRQDQGRSHUDWLRQOHYHOZHKDYHDFKLHYHG DQXQGHUVWDQGLQJRIKRZLQIRUPDWLRQ
WRROV DUH EHLQJ DGDSWHG WR ERWK WKHLU FRQWH[WXDO FRQGLWLRQV DQG WKH LQIRUPDWLRQ QHHGV RI WKH
FRPPXQLW\RIXVHUV
5HVHDUFKOLPLWDWLRQVLPSOLFDWLRQV±0RUHEURDGO\WKHVWXG\LOOXVWUDWHVWKHGHFLVLYHUROHWKDWFRQWH[W
PD\SOD\LQVKDSLQJWKHDFWXDOXVDJHRILQIRUPDWLRQWHFKQRORJ\$QDZDUHQHVVRIWKHUROHRIFRQWH[W
PD\OHDGWRPRUHUREXVWDSSURDFKHVWRWKHLQWURGXFWLRQRI,&7VROXWLRQV%\IRFXVLQJDQDO\WLFDOO\RQ
VLWXDWLRQV RI PHGLDWLRQ EUHDNGRZQ LQ WKH VLWXDWLRQ RI XVH DW ERWK DQ DFWLYLW\ DFWLRQ DQG RSHUDWLRQ
OHYHO ZH KDYH DFKLHYHG DQ XQGHUVWDQGLQJ RI KRZ LQIRUPDWLRQ WRROV DUH EHLQJ DGDSWHG WR ERWK WKHLU
FRQWH[WXDOFRQGLWLRQVDQGWKHLQIRUPDWLRQQHHGVRIWKHFRPPXQLW\RIXVHUV
2ULJLQDOLW\YDOXH±:KLOHDFWLYLW\WKHRU\OLWHUDWXUHRIIHUVLQVLJKWKRZWRDQDO\]HFRQWH[WKRZHYHUWKH
GLVFXVVLRQLVOLPLWHGWRWKHXQGHUVWDQGLQJRIKRZFRQWH[WFDQEHPRGHOOHGLQWRDUWHIDFWV7KH DUWLFOH
VXJJHVWVWKDWWKHFRQWUDGLFWLRQFRQFHSWLVXVHIXOLQLQIRUPLQJWKHUROHRIFRQWH[WDQGLWVLPSDFWLQFR
HYROXWLRQ RI ZRUN DQG LQIRUPDWLRQ WRROV 7KH VWXG\ KDV DOVR FRQWULEXWHV WR WKH GLVFRXUVH LQ KHDOWK
LQIRUPDWLRQ V\VWHPV +,6 LQ GHYHORSLQJ FRXQWULHV E\ HPSKDVL]LQJ WKH FUXFLDO UROH RI WKH IURQW OLQH
KHDOWKZRUNHUV¶RZQSUREOHPVROYLQJLQYHQWLRQDQGDGDSWDWLRQRILQIRUPDWLRQWRROV
.H\ZRUGV 3DWLHQW&DUH ,QIRUPDWLRQ 6\VWHPV 'HYHORSLQJ &RXQWULHV (WKQRJUDSKLF 6WXGLHV $FWLYLW\
7KHRU\5ROHRI&RQWH[W
3DSHU7\SH5HVHDUFKSDSHU
r
ee
rP
Fo
–”‘†—…–‹‘
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
7KH SDSHU WDNHV LWV SUDFWLFDO PRWLYDWLRQ IURP WKH FKDOOHQJHV RI HPSOR\LQJ LQIRUPDWLRQ DQG
FRPPXQLFDWLRQ WHFKQRORJLHV ,&7 WR IXUWKHU VRFLRHFRQRPLF GHYHORSPHQW LQ GHYHORSLQJ
FRXQWULHV+RZWRDYRLGIDLOXUHKDVEHHQDPDMRUWKHPHLQWKH,&7IRUGHYHORSPHQW,&7'
OLWHUDWXUH WKXV WKH GLVFRQQHFW EHWZHHQ LQIRUPDWLRQ V\VWHPV GHYHORSPHQW DQG LWV FRQWH[W RI
XVH RU ZKDW +HHNV FDOOV WKH ³GHVLJQUHDOLW\ JDS´ LV RIWHQ DGGUHVVHG $ FRUH
UHTXLUHPHQWIRURYHUFRPLQJWKLVJDSLVWRFRQVLGHU,6GHVLJQDVDVLWXDWHGDFWLRQWKDWLQFOXGHV
WKHHPEHGGLQJRIWKHVRFLDOFRQWH[W+HHNV&DOOVIRUFRQWH[WXDOO\DSSURSULDWHGHVLJQ
SURFHVVHV DQG SDUWLFLSDWRU\ GHYHORSPHQW KDV WKHUHIRUH EHHQ VWURQJ LQ WKLV OLWHUDWXUH HJ
7RQJLDDQG 6XEUDKPDQLDQ /XQG DQG 6XXWLQHQ 3LWXODDQG 5DGKDNULVKQDQ 7KH PDMRULW\ RI WKHVH ZRUNV DGGUHVVHV DVSHFWV RI FRQWH[W WKDW DUH UHOHYDQW VSHFLILFDOO\ WR
GHVLJQ DQG LPSOHPHQWDWLRQ RI ,&7V &KDUDFWHULVWLFV VXFK DV FRPSXWHU OLWHUDF\ DQG OLPLWHG
ϭ
Information Technology & People
PDWHULDO UHVRXUFHV KDYHD GLUHFW LPSDFW RQ WKHGHVLJQ DQG GHYHORSPHQWDFWLYLWLHVDQG PDQ\
,&7'VWXGLHVDGGUHVVWKHVHNLQGVRIFRQWH[WXDOFKDOOHQJHV)RULQVWDQFH6LQQLJHWDO
SURSRVHDGLJLWDOVWRU\WHOOLQJVWUDWHJ\IRUHOLFLWLQJXVHUV¶UHTXLUHPHQWVLQRUGHUWRDGGUHVVWKH
SUREOHPRIORZOLWHUDF\0DUDLVVXJJHVWWKDWDPDLQIRFXVVKRXOGEHRQLGHQWLI\LQJDQG
QXUWXULQJ WKH ORFDO KXPDQ DQG LQVWLWXWLRQDO FDSDFLW\ WKDW LV UHTXLUHG IRU VXVWDLQDELOLW\ RI WKH
LQWHUYHQWLRQ:KLOHZHZKROHKHDUWHGO\VXSSRUWWKHVHDGYLFHVLQWKLVSDSHUZHDUJXHWKDWLWLV
DOVR UHOHYDQW WR XQGHUVWDQG WKH UROH RI FRQWH[W EH\RQG D VSHFLILF GHVLJQ VLWXDWLRQ :H VWXG\
QRW RQO\ WKH DVSHFWV RI FRQWH[W GLUHFWO\ UHOHYDQW IRU ,&7 SURMHFWV EXW PRUH EURDGO\ KRZ
FRQWH[W VKDSHV WKH H[LVWLQJ ZRUN SUDFWLFHV WKDW XWLOL]H SDSHUEDVHG LQIRUPDWLRQ V\VWHPV DQG
WRROV :H DUJXH WKDW DFFRXQWLQJ IRU KRZ WKHVH WRROV KDYH HPHUJHG DQG FRHYROYHG WRJHWKHU
ZLWKWKHZRUNSUDFWLFHVZLOO\LHOGYDOXDEOHLQVLJKWVWKDWFDQLQIRUPGHVLJQ2XUVWXG\RIIHUV
ERWK LPPHGLDWH GHVLJQ LPSOLFDWLRQV UHJDUGLQJ HJ IXQFWLRQDOLW\ FRQWHQW DQG OD\RXW RI
GLJLWDO WRROVDQGDGHHSHU XQGHUVWDQGLQJ RIZKLFKFRQWH[WXDODVSHFWV LQWHUDFW ZLWK WKH ZRUN
SUDFWLFHVDQGWRROV6XFKXQGHUVWDQGLQJFDQFRQWULEXWHWRPRUHUREXVWDQGVXVWDLQDEOHFKDQJH
LQLWLDWLYHV,JLUD2XUFRQWULEXWLRQWRWKH,&7'OLWHUDWXUHLVWRHPSKDVL]HWKHUROHRI
FRQWH[W DW D PRUH JUDQXODU OHYHO EH\RQG VLPSO\ HTXDWLQJ LW ZLWK JHRJUDSKLF ORFDWLRQ WKDW
FRPHVZLWKFHUWDLQVRFLRHFRQRPLFYDULDEOHV
r
ee
rP
Fo
2XUHPSKDVLVRQZRUNSUDFWLFHVIRUHJURXQGVWKHRSHUDWLRQDOZRUNRIILUVWOLQHKHDOWKZRUNHUV
VLQFH WKH DFWLYLWLHV RI WKH ZRUNHUV DUH FHQWUDO IRU RUJDQL]DWLRQDO OHDUQLQJ DQG LQQRYDWLRQ
%URZQDQG'XJXLG7KHZRUNHUV¶QHHGIRUWRROVWKDWKHOSWKHPFRQGXFWWKHLUZRUNLQ
DJLYHQVLWXDWLRQGULYHVWKHFRHYROXWLRQRIWRROVDQGSUDFWLFHVWRZDUGVFRQYHUJHQFHLHZKHQ
LQIRUPDWLRQ DUWHIDFWV ILWV WKH QHHGV RI WKH XVHUV 6WDU HW DO 6WDU HW DO 7KH
VSHFLILF ZRUN VLWXDWLRQ *DVVHU LV WKH µVLWH¶ RI WKLV OHDUQLQJ DQG LQQRYDWLRQ 1LFROLQL
7KLV IRFXV RQ XQGHUVWDQGLQJ WKH ZRUN SUDFWLFH RQ LWV RZQ WHUPV VHWV RXU VWXG\ DSDUW
IURP PDQ\ RWKHU VWXGLHV RI ,&7V LQ GHYHORSLQJ FRXQWULHV 5HVHDUFK ZLWKLQ WKH +HDOWK
,QIRUPDWLRQ 6\VWHPV +,6 OLWHUDWXUH KDYH PDLQO\ EHHQ RULHQWHG WR WKH DVSHFWV RI KHDOWK
LQIRUPDWLRQ V\VWHPV WKDW DUH UHOHYDQW IRU KHDOWK PDQDJHUV¶ XVHRI LQIRUPDWLRQ VXFK DV GDWD
TXDOLW\ WLPHOLQHVV RI UHSRUWLQJ DQDO\WLF IXQFWLRQDOLW\ RI WRROV DQG VXVWDLQDELOLW\ DQG
VFDODELOLW\RI+,6LQLWLDWLYHVVHHIRUH[DPSOH6DXHUERUQDQG/LSSHYHOG0R\R
6LPED DQG 0ZDQJX 0SKDWVZH HW DO 6DHER HW DO .DUDUD HW DO ([FHSWLRQV LQ WKH +,6 OLWHUDWXUH DUH D QXPEHU RI VWXGLHV RI SULPDU\ KHDOWK ZRUNHUV DQG
FRPPXQLW\YROXQWHHUVZKLFKGHVFULEHVKRZFRQWH[WVKDSHVLQIRUPDWLRQV\VWHPV)RUH[DPSOH
UHVHDUFKHUVGHVFULEHVHYHUDOFRQWH[WXDOSUREOHPVIDFLQJGDWDPDQDJHPHQWDWWKHIDFLOLW\OHYHO
VXFKDVODFNRILQIRUPDWLRQWRROV,JLUD.DQMR1JRPDDQG,JLUDODFNRI
VWDQGDUGL]HG WRROV 'DPWHZ KLJKHU QXPEHU RI FOLHQWV DQG SRRU ZRUNLQJ FRQGLWLRQV
ZKLFK LPSDFW RYHUDOO GDWD PDQDJHPHQW DFWLYLWLHV 6KR]L HW DO +RZHYHU DOVR WKLV
UHVHDUFKRQ3+&ZRUNHUVVHHPWRSD\PRUHDWWHQWLRQWRLPSURYLQJDYDLODELOLW\RILQIRUPDWLRQ
IRUPDQDJHUVDQGOHVVRQKRZLQIRUPDWLRQV\VWHPFDQEHXVHGDQGLPSOHPHQWHGWRIDFLOLWDWH
KHDOWKFDUHGHOLYHU\2XUVWXG\DGGVWRWKLVOLWHUDWXUHE\SURYLGLQJDPRUHVWXG\WKDWH[DPLQHV
WKH UROH LQIRUPDWLRQ SOD\V IRU KHDOWKFDUH SURYLVLRQ DQG DOVR E\ RIIHULQJ D PRUH GHWDLOHG
XQGHUVWDQGLQJRIKRZWKHFRQWH[WRIWKHKHDOWKZRUNHUVLPSDFWVWKHLULQIRUPDWLRQSUDFWLFHV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 2 of 25
Ϯ
Page 3 of 25
2XUHPSLULFDOPDWHULDOFRQFHUQVLQIRUPDWLRQSUDFWLFHVDQGWRROVHPSOR\HGE\KHDOWKZRUNHUV
ZKR ZRUN ZLWK FKLOGUHQ ERUQ IURP +,9 SRVLWLYH SUHJQDQW ZRPHQ 7KH ZRPHQ DUH HQUROOHG
LQWRWKH3UHYHQWLRQRI0RWKHU7R&KLOG7UDQVPLVVLRQ307&7SURJUDPLQ7DQ]DQLDZKLFK
DLPV DW SUHYHQWLQJ WUDQVPLVVLRQ RI +,9$,'6 IURP +,9 SRVLWLYH PRWKHUV WR WKHLU FKLOGUHQ
%DVHG RQ JXLGHOLQHV IURP WKH :RUOG +HDOWK 2UJDQL]DWLRQ :+2 WKH SURJUDP RIIHUV
LQWHUYHQWLRQVERWKGXULQJSUHJQDQF\ELUWKDQGEUHDVWIHHGLQJSHULRG7KHRIILFLDOLQIRUPDWLRQ
V\VWHP WKDW VXSSRUWV WKH 307&7 SURJUDP FRPSULVHV SDSHU IRUPV DQG UHJLVWHUV ZKHUH
LQIRUPDWLRQ DERXW HDFK FOLHQW LV UHFRUGHG 7KLV LQIRUPDWLRQ LV DJJUHJDWHG DQG UHJXODUO\
UHSRUWHG WR SURJUDP PDQDJHUV IRU PRQLWRULQJ HYDOXDWLRQ DQG SODQQLQJ SXUSRVHV 7KH IRUPV
DQGUHJLVWHUVDOORZWKHKHDOWKZRUNHUVWRPDLQWDLQDFOLHQWFHQWHUHGDFFRXQWRIHQFRXQWHUVDQG
LQWHUYHQWLRQVD³UHFRUG´+RZHYHUWKHVHWRROVRIIHUOLPLWHGVXSSRUWZKHQWKHKHDOWKZRUNHUV
QHHG WR KDQGOH GHYLDWLRQV VXFK DV WUDFNLQJ FOLHQWV ZKR GR QRW VKRZ XS IRU WKH VFKHGXOHG
DSSRLQWPHQWV2XUVWXG\VSHFLILFDOO\DGGUHVVHGWKHFRPSOHPHQWDU\WRROVWKDWKHDOWKZRUNHUV
GHYLVHG LQ RUGHU WR EH DEOH WR GHDO ZLWK GHYLDWLRQV DQG ZH FRPSDUH WKH YDULDWLRQ RI WRROV¶
GHVLJQ DQG XVDJH DFURVV YDULRXV KHDOWK IDFLOLWLHV LQ RUGHU WR H[DPLQH WKH FRQWH[WUHODWHG
UHDVRQVIRUGLIIHUHQFHV:HKDYHFKRVHQDFWLYLW\WKHRU\(QJHVWU|PDVRXUWKHRUHWLFDO
IUDPHZRUN $FWLYLW\ WKHRU\ H[SOLFLWO\ SODFHV KXPDQ DFWLYLW\ ZLWKLQ D FXOWXUDO VRFLDO DQG
WHPSRUDOGHYHORSPHQWDOFRQWH[W,QDGGLWLRQZHEXLOGRQ/HRQWHY¶VQRWLRQRIKLHUDUFK\RI
DFWLYLW\DVDFWLYLW\DFWLRQVDQGRSHUDWLRQV/HRQWHYLQRUGHUWRXQGHUVWDQGWKHGHWDLOHG
ZRUNSUDFWLFHV
r
ee
rP
Fo
,Q WKH QH[W VHFWLRQ ZH EULHIO\ SUHVHQWV FRUH FRQFHSWV IURP DFWLYLW\ WKHRU\ DQG VSHFLILFDOO\
GLVFXVV KRZ FRQWH[W KDV EHHQ XQGHUVWRRG ,Q VHFWLRQ WKUHH ZH SUHVHQW WKH HPSLULFDO VHWWLQJ
DQG RXU UHVHDUFK DSSURDFK :H SUHVHQW WKH HPSLULFDO PDWHULDO LQ VHFWLRQ IRXU WR GLVSOD\ WKH
YDULDWLRQLQWRROV¶XVDJHRYHUWLPHDQGDFURVVWKHYDULRXVKHDOWKIDFLOLWLHV,QVHFWLRQILYHZH
WXUQ WR DQ DQDO\VLV RI WKH UHODWLRQ EHWZHHQ WKH FRQWH[W DQG WKH ZRUN SUDFWLFHV 6HFWLRQ VL[
GLVFXVVHVWKHPDLQILQGLQJVRIWKHVWXG\/DVWO\WKHLPSOLFDWLRQDQGFRQFOXVLRQDULVLQJIURP
WKHVWXG\DUHHODERUDWHG
Š‡‘”‡–‹…ƒŽˆ”ƒ‡™‘”
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
,Q WKLV VHFWLRQ ZH SUHVHQW FRUH FRQFHSWV IURP DFWLYLW\ WKHRU\ WKDW UHODWHV WR RXU DLP RI
XQGHUVWDQGLQJKRZFRQWH[WVKDSHVZRUNSUDFWLFHVDQGLQIRUPDWLRQWRROV
&RUHFRQFHSWVRI$FWLYLW\7KHRU\
7KH EDVLF XQLW RI DQDO\VLV LQ DFWLYLW\ WKHRU\ LV KXPDQ DFWLYLW\ ZKLFK LV GULYHQ E\ FHUWDLQ
QHHGVZKHUHSHRSOHZLVKWRDFKLHYHFHUWDLQSXUSRVHV%DQQRQDQG%¡GNHU$QDFWLYLW\
V\VWHP LV FRPSRVHG RI VXEMHFW REMHFW FRPPXQLW\ DV ZHOO DV WKH PHGLDWLQJ HOHPHQWV RI
DUWHIDFWV UXOHV DQG GLYLVLRQ RI ODERXU (QJHVWU|P 7KH VXEMHFW DLPV WR DFFRPSOLVK
VRPHWKLQJHJLQWKLVFDVHWRRIIHUKHDOWKFDUHVHUYLFHVWRFOLHQWVDQGLVPRWLYDWHGE\WKLV
REMHFWLYH 7KH DFWLYLWLHV DUH PHGLDWHG E\ LQVWUXPHQWV HJ LQIRUPDWLRQ DUWHIDFWV VXFK DV
SDWLHQW UHJLVWHUV RU PHGLFDO VXSSOLHV VXFK DV GUXJV RU VNLOOV HJ UHODWHG WR KHDOWK FDUH
SURYLVLRQ RU LQIRUPDWLRQ PDQDJHPHQW 7KH UHODWLRQVKLS EHWZHHQ VXEMHFW REMHFW DQG
FRPPXQLW\LVDOVRPHGLDWHGE\UXOHVVXFKDVPHGLFDOHWKLFVDQGVRFLDOUXOHVDQGGLYLVLRQRI
ϯ
Information Technology & People
ODERXU KRZ WDVNV DQG UHVSRQVLELOLWLHV DUH DOORFDWHG DPRQJ SDUWLFLSDQWV IRUPLQJ WKH
FRPPXQLW\HOHPHQW)LJXUHVKRZVWKHVWUXFWXUHRIDQDFWLYLW\V\VWHP
r
ee
rP
Fo
$FWLYLW\WKHRU\DFNQRZOHGJHVWKDWDQDFWLYLW\V\VWHPLQWHUDFWDQGRYHUODSZLWKRWKHUDFWLYLW\
V\VWHPV 7KLV UHTXLUHV WKDW WKH XQLW RI DQDO\VLV LV H[WHQGHG WR LQFOXGH DW OHDVW WZR
LQWHUFRQQHFWHGDFWLYLW\V\VWHPV)LJXUH,QWKLVLQWHUDFWLRQWKHREMHFWRIDQDFWLYLW\V\VWHP
H[SDQGVIURPDQLQLWLDOVWDWHREMHFWWRDQRWKHUVWDWHREMHFWDQGDSRWHQWLDOO\VKDUHGRU
MRLQW REMHFWV %\ SURYLGLQJ IHHGEDFN WR WKH UHVSHFWLYH DFWLYLW\ V\VWHPV WKH WKLUG REMHFW
REMHFWJLYHVULVHWRWUDQVIRUPDWLRQZLWKLQDQHWZRUNRIDFWLYLW\V\VWHPV(QJHVWU|P
(QJHVWURPHWDO1LFROLQL
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 4 of 25
)LJXUH7ZRLQWHUDFWLQJDFWLYLW\V\VWHPV(QJHVWU|P
$SUHPLVHIRUDFWLYLW\WKHRU\ZKLFKLVEXLOWXSRQDGLDOHFWLFDOSHUVSHFWLYHLVWKDWFKDQJHDQG
GHYHORSPHQW LV WULJJHUHG E\ FRQWUDGLFWLRQV (QJHVWU|P ,Q ZRUN SUDFWLFHV
FRQWUDGLFWLRQV FDQ PDQLIHVW WKHPVHOYHV DV SUREOHPV UXSWXUHV EUHDNGRZQV RU FODVKHV %OLQ
DQG0XQUR0LOQHHWDODOVRJHQHUDOO\FDOOHGGLVWXUEDQFHV'LVWXUEDQFHVFDQEH
FKDUDFWHUL]HG DV GHYLDWLRQV IURP WKH QRUPDO IORZ RI ZRUN DQG WKH\ FDQ SRLQW WR XQGHUO\LQJ
GHYHORSPHQWDO FRQWUDGLFWLRQV RI WKH DFWLYLW\ V\VWHPV LQYROYHG +HOOH WKH VWXG\ RI
WKHVH GLVWXUEDQFHV LV WKHUHIRUH D YDOXDEOH DQDO\WLF VWUDWHJ\ 0LOQH HW DO &KDQJH RI
SUDFWLFHV FDQ RFFXU LI HQGXVHUV ZLOO QRW LJQRUH GLVWXUEDQFHV DQG LQVWHDG GHFLGH WR DGGUHVV
WKHP IRU H[DPSOH E\ GHYHORSLQJ QHZ VROXWLRQV &OXWV ZKLFK FDQ EH GHYLDWLRQ RI
ϰ
Page 5 of 25
IRUPDO SUDFWLFHV VXFK DV DGDSWDWLRQV RU LPSURYLVDWLRQ *DVVHU )ROFKHU $OWHU
5HOHYDQFHRIDFWLYLW\WKHRU\IRUVWXG\LQJWKHLPSDFWRIFRQWH[W
6RPHUHVHDUFKHUVDSSURDFKFRQWH[WLQDSUDJPDWLFVWUDLJKWIRUZDUGPDQQHU'H\HWDO
GHILQHV FRQWH[WDV ³DQ\ LQIRUPDWLRQ WKDW FDQ EH XVHG WRFKDUDFWHUL]H WKH VLWXDWLRQ RI HQWLWLHV
LHZKHWKHUDSHUVRQSODFHRUREMHFWWKDWDUHFRQVLGHUHGUHOHYDQWWRWKHLQWHUDFWLRQEHWZHHQ
D XVHU DQG DQ DSSOLFDWLRQ LQFOXGLQJ WKH XVHU DQG WKH DSSOLFDWLRQ WKHPVHOYHV´ LELG S
([DPSOHV RI FRQWH[W LGHQWLILHG DV UHOHYDQW DUH HJ WLPH ORFDWLRQ PDWHULDO DQG SHRSOH
LQYROYHG%RHUHWDO%DUGUDPDQG'RU\DE'RU\DEDQG%DUGUDP$FWLYLW\
WKHRU\ LV IXQGDPHQWDOO\ RULHQWHG WR DFFRXQWLQJ IRU FRQWH[W EHFDXVH WKH DFWLYLW\ LV VHHQ DV
FXOWXUDOO\VRFLDOO\DQGKLVWRULFDOO\VLWXDWHG0RUHRYHUDFWLYLW\WKHRU\GRQRWRQO\VHHFRQWH[W
DVDQH[WHUQDO³FRQWDLQHU´IRUWKHDFWLYLW\V\VWHPFRQWH[WLVDOVRXQGHUVWRRGWREHLQWHJUDOWR
WKHDFWLYLW\V\VWHP$FFRUGLQJWR1DUGL
r
ee
rP
Fo
³$FWLYLW\WKHRU\WKHQSURSRVHVDYHU\VSHFLILFQRWLRQRIFRQWH[WWKHDFWLYLW\LWVHOILVWKHFRQWH[W:KDW
WDNHVSODFHLQDQDFWLYLW\V\VWHPFRPSRVHGRIREMHFWDFWLRQVDQGRSHUDWLRQLVWKHFRQWH[W&RQWH[WLV
FRQVWLWXWHGWKURXJKWKHHQDFWPHQWRIDQDFWLYLW\LQYROYLQJSHRSOHDQGDUWLIDFWV&RQWH[WLVQRWDQRXWHU
FRQWDLQHURUVKHOOLQVLGHRIZKLFKSHRSOHEHKDYHLQFHUWDLQZD\V3HRSOHFRQVFLRXVO\DQGGHOLEHUDWHO\
JHQHUDWH FRQWH[WV DFWLYLWLHV LQSDUWWKURXJKWKHLURZQ REMHFWVKHQFHFRQWH[WLVQRWMXVWCCRXWWKHUH
&RQWH[WLVERWKLQWHUQDOWRSHRSOH²LQYROYLQJVSHFLILFREMHFWVDQGJRDOV²DQGDWWKHVDPHWLPHH[WHUQDO
WRSHRSOHLQYROYLQJDUWLIDFWVRWKHUSHRSOHVSHFLILFVHWWLQJV7KHFUXFLDOSRLQWLVWKDWLQDFWLYLW\WKHRU\
H[WHUQDODQGLQWHUQDODUHIXVHGXQLILHG´1DUGLS
7KH WKHRU\ WKHUHIRUH GRHV QRW RQO\ SRLQW WR H[WHUQDO FRQGLWLRQV DV FRQWH[W EXW DOVR RIIHUVD
ULFKHUDQGPRUHFRPSOH[XQGHUVWDQGLQJRIFRQWH[WDVERWKLQWHUQDODQGH[WHUQDOWRWKHDFWLYLW\
7KH HPSKDVLV RQ ³LQWHUQDO FRQWH[W´ QHFHVVLWDWHV D VROLG DQDO\VLV RI WKH DFWLYLW\ LWVHOI 7KH
UHFRJQLWLRQ RI ³H[WHUQDO FRQWH[W´ IDFLOLWDWHV WKH LQFRUSRUDWLRQ RI FRQWH[W EH\RQG WKH DFWLYLW\
LWVHOI 7KH XQGHUVWDQGLQJ WKDW PXOWLSOH DFWLYLW\ V\VWHPV FDQ EH FRQQHFWHG SRLQWV WR WKH
UHOHYDQFH RI VWXG\LQJ WKH UHODWLRQVKLS EHWZHHQ RQH DFWLYLW\ V\VWHP DQG RWKHU LQWHUDFWLQJ
DFWLYLW\V\VWHPVVXFKDVHJHFRQRPLFV\VWHPLQGXVWU\VXSSO\FKDLQRUJDQL]DWLRQ%RHUVHW
DO WKDW HQFRPSDVV WKH ODUJHU ODQGVFDSH .RUSHOD HW DO 2WKHU UHVHDUFKHUV KDYH
VXJJHVWHGWRDQDO\]HFRQWH[WVSHFLILFDOO\UHODWHGWRHDFKHOHPHQWLQWKHDFWLYLW\V\VWHPVXFK
DV VXEMHFW FRQWH[W DUWHIDFW FRQWH[W REMHFW FRQWH[W DQG VRFLDO FRQWH[W &DVVHQ DQG .RIIRG
3HWHUVHQ .RIIRG3HWHUVHQ DQG &DVVHQ +XDQJ DQG *DUWLQHU 0RUHRYHU
UHVHDUFKHUV FDQ XVH WKH HOHPHQWV RI DFWLYLW\ DV IUDPHZRUN WR JXLGH UHVHDUFK DFWLYLWLHV ZKHQ
WKH\ZDQWWRLGHQWLI\FRQWH[WXDOIHDWXUHV.DSWHOLQLQHWDO:HEXLOGRQWKLVGLVFRXUVH
WR GHPRQVWUDWHWKDWFRQWUDGLFWLRQFRQFHSW LV XVHIXO LQ LGHQWLI\LQJ WKH LQIOXHQFH RI FRQWH[WLQ
ZRUNSUDFWLFHV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
,WLVDOVRUHOHYDQWWRQRWHWKDWDFWLYLW\WKHRU\DVVLJQVDFHQWUDOUROHWRDUWHIDFWV$UWHIDFWVDUH
PHGLDWLQJ WKH VXEMHFWV¶ ZRUN WRZDUGV WKH REMHFW 7KLV FRQFHSWXDOL]DWLRQ RI WKH DUWHIDFWV
FHQWHUVWKHDQDO\VLVRQWKHLUUHODWLRQWRDFWXDOXVHUDWKHUWKDQDVVWDQGDORQHHQWLWLHV7KHPRVW
VLJQLILFDQW FRQWH[W IRU DQ DUWLIDFW LV WKHUHIRUH WKH VLWXDWLRQ RI XVH ZKHUH D VXEMHFW ZRUNV WR
DFKLHYH DQ REMHFW )XUWKHUPRUH WKH DQDO\VLV RI WKLV FRQWH[W RI XVH FDQ EH DLGHG E\ VHHLQJ
ϱ
Information Technology & People
DFWLYLWLHVDVKLHUDUFKLFDOO\RUJDQL]HGLQWRDFWLYLW\DFWLRQVDQGRSHUDWLRQV/HRQWHYDV
ZHOODVFRPSRVHGRIVHTXHQFHVRIVWHSVZKLFKFDQKDSSHQVHTXHQWLDOO\RULQSDUDOOHOPDQQHU
.XXWWL .DSWHOLQLQ DQG 1DUGL $FWLYLWLHV DUH GLUHFWHG DW PRWLYHV ZKLOH DFWLRQV
DUHFRQVFLRXVSURFHVVHVGLUHFWHGDWJRDOVZKLFKPXVWEHSHUIRUPHGWRDFFRPSOLVKWKHREMHFW
$FWLRQV DUH DFKLHYHG WKURXJK WKH RSHUDWLRQV ZKLFK DUH URXWLQH SURFHVVHV SURYLGLQJ
DGMXVWPHQWV WR WKH DFWLRQ LQ WKH VLWXDWLRQ .DSWHOLQLQ DQG 1DUGL S )LJXUH UHSUHVHQWWKHKLHUDUFKLFDOVWUXFWXUHRIWKHDFWLYLW\
$FWLYLW
0RWLYH
$FWLRQ
$FWLRQ
2SHUDWLRQ
*RDOV
$FWLRQ
r
ee
rP
Fo
2SHUDWLRQ
2SHUDWLRQ
&RQGLWLRQV
)LJXUH7KHKLHUDUFKLFDOVWUXFWXUHRIDQDFWLYLW\DGDSWHGIURP.DSWHOLQLQDQG1DUGL
.DSWHOLQLQDQG1DUGLDUJXHWKDWWKHKLHUDUFKLFDOVWUXFWXUHRIWKHDFWLYLW\V\VWHPRSHQV
XS D SRVVLELOLW\ IRU DQDO\VLV RI PRWLYDWLRQDO DFWLYLW\ OHYHO JRDO GLUHFWHG DFWLRQ OHYHODQG
RSHUDWLRQDO DVSHFW RSHUDWLRQ OHYHO RI KXPDQ DFWLQJ LQ WKH ZRUOG E\ DGGUHVVLQJ LVVXHV OLNH
ZK\ ZKDW DQG KRZ S 7KLV KLHUDUFKLFDO VWUXFWXUH FDQ DOVR DVVLVW WKH DQDO\VLV RI KRZ
FRQWH[WLPSDFWVDFWLYLWLHV6RPHUHVHDUFKHUVKDYHDWWHPSWHGWRVWXG\WKHLQIOXHQFHRIFRQWH[W
DW WKH DFWLRQV DQG RSHUDWLRQV OHYHOV HJ .RIIRG3HWHUVHQ DQG &DVVHQ +XDQJ DQG
*DUWLQHU %DUGUDP DQG 'RU\DE VXJJHVW WKDW KXPDQ DFWLYLWLHV DUH SHUIRUPHG WKURXJK
RSHUDWLRQVZKLFKDUHDGMXVWHGDFFRUGLQJWRWKHVSHFLILFFRQGLWLRQVRIWKHUHDOZRUOG%DUGUDP
DQG 'RU\DE )RU H[DPSOH WKH RSHUDWLRQ SURFHVVHV DUH FORVHO\ OLQNHG WR WKH FRQFUHWH
SK\VLFDO RU VRFLDO FRQGLWLRQV IRU FRQGXFWLQJ WKH DFWLRQ %DQQRQ DQG %¡GNHU 7KHVH
LQVLJKWV IURP $FWLYLW\ 7KHRU\ VWXGLHV KDYH KHOSHG XV WR IRUPXODWH RXU DQDO\WLF IUDPHZRUN
ZKHUH ZH VHHN WR GHOLQHDWH KRZ LQIRUPDWLRQ WRROV UHODWHV WR ZRUN DW DQ DFWLYLW\ DFWLRQ DQG
RSHUDWLRQOHYHODQGKRZFRQWH[WXDOIHDWXUHVLQIOXHQFHWKLVUHODWLRQDWHDFKRIWKHWKUHHOHYHOV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 6 of 25
:H EXLOG RQ WKH SUHPLVH WKDW WKH DUWHIDFWV¶ UROH LQ WKH ZRUN VLWXDWLRQ LV WKH FRUH HPSLULFDO
IRFXVRIRXUVWXG\)XUWKHUZHSD\DWWHQWLRQWRLPSURYLVDWLRQVFRQGXFWHGE\XVHUVWRLGHQWLI\
EUHDNGRZQV LQ ZRUN SUDFWLFHV +DVX DQG ZH VHDUFK IRU GLVWXUEDQFHV ZKLFK VLJQDOV
PHGLDWLRQEUHDNGRZQRIWKHDUWHIDFWVDQGVWXG\WKHKHDOWKZRUNHUV¶LPSURYLVDWLRQVWRFRXQWHU
WKHVH GLVWXUEDQFHVDVWKLV SRLQWVWR XQGHUO\LQJ WHQVLRQV DQGFRQWUDGLFWLRQV:H ZLOODQDO\]H
GLVWXUEDQFHV DW ERWK DFWLYLW\ DFWLRQ DQG RSHUDWLRQ OHYHO LQ RUGHU WR SLQSRLQW KRZ FRQWH[W
LPSDFWVWKHZRUNVLWXDWLRQ
ϲ
Page 7 of 25
’‹”‹…ƒŽ‡––‹‰ƒ†‡•‡ƒ”…Š’’”‘ƒ…Š
7KLV VWXG\ LV SDUW RI WKH +HDOWK ,QIRUPDWLRQ 6\VWHP 3URJUDP +,63 ZKLFK LV UXQ E\ WKH
'HSDUWPHQW RI ,QIRUPDWLFV 8QLYHUVLW\RI 2VOR7KHSULPDU\ JRDO RI +,63 LV WRHQKDQFH WKH
LQIRUPDWLRQXVHE\KHDOWKPDQDJHUVSODQQHUVDQGZRUNHUVLQWKHKHDOWKVHFWRURIGHYHORSLQJ
FRXQWULHV E\ VWUHQJWKHQLQJ ORFDO SURIHVVLRQDO FDSDFLW\ IRU WKH GHYHORSPHQW RIDVXVWDLQDEOH
+,6 %UDD HW DO +,63 KDV GHYHORSHG VRIWZDUH NQRZQ DV '+,6 'LVWULFW +HDOWK
,QIRUPDWLRQ6RIWZDUHZKLFKLVXVHGWRPDQDJHGDWDLQYDULRXVOHYHOVRIWKHKHDOWKVHFWRULQ
PRUH WKDQ GHYHORSLQJ FRXQWULHV :KLOH WKH '+,6 DGGUHVV WKH URXWLQH UHSRUWLQJ RI
DQRQ\PL]HGGDWDUHFHQWO\+,63KDVDOVRGHYHORSHGDSDWLHQWRULHQWHGPRGXOHNQRZQDVWKH
'+,6 7UDFNHU 7KLV FDQ EH XVHG IRU PDQDJLQJ QDPHEDVHG SDWLHQW GDWD DQG DOVR IRU
SURYLGLQJUHPLQGHUVDQGDOHUWVWRSUDFWLWLRQHUVDQGFOLHQWV*L]DZHWDO6DXJHQH
2XU VWXG\ IHHGV LQWR WKH GHVLJQ SURFHVV RI WKH '+,6 7UDFNHU E\ SURYLGLQJ D SUDFWLFHEDVHG
XQGHUVWDQGLQJRIWKHH[LVWLQJLQIRUPDWLRQSUDFWLFHVLQKHDOWKIDFLOLWLHV
r
ee
rP
Fo
,Q WKLV VHFWLRQ ZH RIIHU EDFNJURXQG LQIRUPDWLRQ RQ WKH VLWXDWLRQ LQ 7DQ]DQLD UHODWHG WR FDUH
IRU +,9SRVLWLYH SUHJQDQW PRWKHUV DQG WKHLU FKLOGUHQ :H GHVFULEH WKH RUJDQL]DWLRQ RI WKH
307&7SURJUDPDQGWKHRIILFLDOLQIRUPDWLRQV\VWHPSURYLGHGWRWKHKHDOWKZRUNHUV7KHQZH
GHVFULEHRXUUHVHDUFKDSSURDFKLQWHUPVRIVHOHFWLRQRIVLWHVGDWDFROOHFWLRQPHWKRGVDQGGDWD
DQDO\VLVSURFHVV
+,9$,'6DQG307&7VHUYLFHVLQ7DQ]DQLD
,Q 7DQ]DQLD WKH SUHYDOHQFH RI +,9 DPRQJ ZRPHQ ZKR DWWHQG DQWHQDWDO FDUH LV DQG
FXUUHQWO\ WKH HVWLPDWHG SHUFHQWDJH RI LQIDQWV ERUQWR +,9 LQIHFWHG PRWKHUV ZKR WKHPVHOYHV
EHFRPHV+,9LQIHFWHGLVDURXQG02+6:$SSURSULDWHLQWHUYHQWLRQVFDQUHGXFH
WKH WUDQVPLVVLRQ UDWH WR OHVV WKDQ DQG WKH 7DQ]DQLDQ 0LQLVWU\ RI +HDOWK DQG 6RFLDO
:HOIDUH FROODERUDWHV ZLWK LQWHUQDWLRQDO SDUWQHUV WR LPSOHPHQW WKH 3UHYHQWLRQ RI 0RWKHU7R
&KLOG7UDQVPLVVLRQ307&7SURJUDPDFURVVWKHFRXQWU\7KH307&7SURJUDPRIIHUV+,9
WHVWLQJ DQG FRXQVHOOLQJ IRU SUHJQDQW ZRPHQ DQG WKHLU SDUWQHUV LQ DQWHQDWDO FDUH $1&
GHOLYHU\ RI DQWLUHWURYLUDO $59 SURSK\OD[LV RU WUHDWPHQW WR SUHYHQW WUDQVPLVVLRQ VDIHU
GHOLYHU\ SUDFWLFHVFRXQVHOOLQJDQG VXSSRUW UHODWHG WREUHDVWIHHGLQJ RI WKH LQIDQWDV ZHOODV
WUHDWPHQWIRUPRWKHUVDQGFKLOGUHQOLYLQJZLWK+,9
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
:KHQ D SUHJQDQW PRWKHU JHWV GLDJQRVHG ZLWK +,9$,'6 VKH ZLOO EH HQUROOHG LQWR WKH
307&7SURJUDPDQGWKH+,9&DUHDQG7UHDWPHQW&HQWHUV&7&SURJUDPWRUHFHLYHHLWKHU
$59SURSK\OD[LVRU$59WUHDWPHQW1RWDOOKHDOWKIDFLOLWLHVZRXOGRIIHU$1&307&7DQG
&7&VHUYLFHVDQGWKHSDWLHQWZLOOWKHQEHUHIHUUHGWRRWKHUIDFLOLWLHVIRURQHRUPRUHRIWKHVH
VHUYLFHV ,Q SUHSDUDWLRQ IRU WKH ELUWK RI WKH FKLOG WKH 307&7 LQIRUPDWLRQ PXVW EH
FRPPXQLFDWHGWRWKHGHOLYHU\ZDUGVRWKDW$59PHGLFDWLRQFDQEHJLYHQGXULQJODERXUDQG
WR WKH FKLOG DIWHU LWV ELUWK 7KH LQIRUPDWLRQ UHJDUGLQJ $59 PHGLFDWLRQ KDYH WR EH
FRPPXQLFDWHGWRSRVWQDWDOVHUYLFHVDIWHUWKHFKLOGLVERUQVLQFHERWKWKHPRWKHUDQGWKHFKLOG
ZLOOUHFHLYHSRVWQDWDOVHUYLFHV7KHPRWKHUZLOOFRQWLQXHWRUHFHLYHVHUYLFHV$59WUHDWPHQW
IURPD&7&FOLQLF7KHFKLOGLQDGGLWLRQWRIROORZLQJWKHUHJXODUVHUYLFHVLQWKHFKLOGFOLQLF
LPPXQL]DWLRQ JURZWK PRQLWRULQJ HWF ZLOO FRQWLQXH LQ WKH 307&7 SURJUDP XQWLO PRQWKVROG7KHFKLOGVKRXOGUHFHLYHSURSK\ODFWLFWUHDWPHQWIURPELUWKDQGXQWLOWZRPRQWKV
ϳ
Information Technology & People
DIWHU EUHDVWIHHGLQJ VWRSV 7KLV LV JLYHQ LQ WKH IRUP RI D GDLO\ GRVDJH RI 1HYLUDSLQH V\UXS
ZKLFK LVDGPLQLVWHUHG E\ WKHFKLOG¶V FDUHU 7KHFKLOGLV VXSSRVHG WREHEURXJKW IRU PRQWKO\
YLVLWVWRWKH307&7FOLQLF$QLQLWLDO+,9WHVWLVFRQGXFWHGZKHQWKHFKLOGLVEHWZHHQDQG
ZHHNV RIWHQ FRLQFLGLQJ ZLWK WKH FRPPHQFHPHQW RI WKH RUGLQDU\ YDFFLQDWLRQ VFKHGXOH $
VHFRQG WHVW LV FRQGXFWHG ZKHQ WKH FKLOG LV PRQWKV DQG D ILQDO FRQILUPDWRU\ WHVW LV
FRQGXFWHG ZKHQ WKH FKLOG LV PRQWKV 'HSHQGLQJ RQ WKH +,9 VWDWXV RI WKH FKLOG LW ZLOO
HLWKHUILQLVK307&7DVKHDOWK\RUEHWUDQVIHUUHGWR&7&WRFRQWLQXHWRUHFHLYHDQWLUHWURYLUDO
WKHUDS\ WKURXJKRXW LWV OLIH +RZHYHU VWDUWLQJ &7& ZLOO KDSSHQ HYHQ EHIRUH WKH DJH RI PRQWKVLIWHVWVDUHSRVLWLYHGXULQJILUVWDQGRUVHFRQGWHVWV,QWKHIROORZLQJVHFWLRQZH]RRP
LQ RQ 307&7 VHUYLFHV RQ WKH IROORZXS RI D FKLOG DIWHU LW LV ERUQ DQG ZLOO GHVFULEH WKH
LQIRUPDWLRQDUWLIDFWVDQGSUDFWLFHVWKDWDUHHPSOR\HG
7KURXJK WKH 307&7 0RQLWRULQJ DQG (YDOXDWLRQ 0( IUDPHZRUN WKH KHDOWK DXWKRULWLHV
SURYLGH JXLGHOLQHV DQG LQIRUPDWLRQ WRROV WR HQVXUH WKDW UHOLDEOH FRUH GDWD LV FROOHFWHG DQG
UHSRUWHGLQRUGHUWRPRQLWRUDQGHYDOXDWHWKH307&7SURJUDP7KHLQIRUPDWLRQWRROVLQFOXGH
$1& 307&7 5HJLVWHU /DERXU DQG 'HOLYHU\ 307&7 5HJLVWHU 307&7 &DUH UHJLVWHU IRU
SUHJQDQWZRPHQDQG307&70RWKHU,QIDQW)ROORZXS5HJLVWHUDQGWKHLUPRQWKO\VXPPDU\
IRUPV 2WKHU LQIRUPDWLRQ WRROV DUH 307&7 'UXJ GLVSHQVLQJ 5HJLVWHU DQG 307&7
WUDQVIHUUHIHUUDO IRUP 'XULQJ WKH ILHOGZRUN SHULRG D QHZ JHQHUDO GHOLYHU\ UHJLVWHU ZDV
LQWURGXFHG LQ WKH 'HOLYHU\ :DUG 7KLV 'HOLYHU\ 5HJLVWHU KDV D VHFWLRQ IRU 307&7
LQIRUPDWLRQ DQG UHSODFHG WKH SUHYLRXV VHSDUDWH UHJLVWHU ERRNVIRU 'HOLYHU\ DQG /DERXUDQG
'HOLYHU\ 307&7 7KH RIILFLDO 307&7 UHJLVWHUV DUH DQRQ\PL]HG WR SURWHFW WKH FOLHQWV¶
SULYDF\(DFKIDFLOLW\LVVXHVDXQLTXHDOSKDQXPHULFDOFRGHHJ;20..WKDWLVXVHGLQ
WKH UHJLVWHUV DQG RQ WKH FOLHQWV¶ FDUGV FDOOHG 5&+ FDUG VLQFH LW LV XVHG ZLWKLQ WKH
5HSURGXFWLYHDQG&KLOG+HDOWKSURJUDP
r
ee
rP
Fo
7KHSUREOHPRI³GHIDXOWHUV´
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 8 of 25
7KH RIILFLDO LQIRUPDWLRQ V\VWHP LV EXLOW RQ WKH DVVXPSWLRQ WKDW FOLHQWV ZLOO FRPSO\ ZLWK WKH
VHUYLFHV LH WKDW WKH\ ZLOO VHHN WUHDWPHQW ZKHQ LW LV GXH LQ WKH IDFLOLW\ ZKHUH WKH\ EHORQJ
+RZHYHULQSUDFWLFHWKLVLVQRWWKHFDVH&OLHQWVPD\FKRRVHWRJRWRDQRWKHUIDFLOLW\ZKLFKLV
HDVLHUWRUHDFK2WKHUVVHHNWRDYRLGWKHVRFLDOVWLJPDDQGSUHIHUWRVHHNWUHDWPHQWLQIDFLOLWLHV
WKDW DUH ORFDWHG IDU DZD\ IURP WKHLU KRPH DUHD 6RPH FOLHQWV DUH GLUHFWHG WR RWKHU IDFLOLWLHV
ZKHQ WKHUH LV D ODFN RI WHVW UHDJHQWV RU PHGLFDWLRQ LQ WKHLU ORFDO IDFLOLW\ 6RPH FOLHQWV DUH
UHOXFWDQWWRDFFHSWWKHLUGLDJQRVLVDQGFKRRVHWRVWD\DZD\IURPWUHDWPHQW,QDGGLWLRQHYHQ
WKRXJKDQWLUHWURYLUDOWUHDWPHQWLVPRVWO\DYDLODEOHLWKDVVHULRXVVLGHHIIHFWVWKDWVFDUHPDQ\
FOLHQWVDZD\IURPXWLOL]LQJLW7KHUHLVDOVRDZLGHVSUHDGVHQWLPHQWWKDWLWMXVWSRVWSRQHVWKH
XQDYRLGDEOHDQGLVUHODWLYHO\LQHIIHFWLYH&OLHQWVPD\JHWWRRLOOWRWUDYHOWRWKHFOLQLFRUPD\
GLHZLWKRXWWKHFOLQLFEHLQJQRWLILHGRIWKLV7KHFOLQLFVWKHUHIRUHH[SHULHQFHZKDWWKH\WHUP
³GHIDXOWHUV´ QRQFRPSO\LQJ FOLHQWV 7KLV KDV EHHQ UHFRJQL]HG DV D ODUJH DQG JHQHUDO
SUREOHPWKDWLVSUHYDOHQWLQPDQ\VHWWLQJVGHDOLQJZLWKERWKWXEHUFXORVLVDQG+,9WUHDWPHQW
-RQHV HW DO )RUVWHU HW DO 7KRPVRQ HW DO 1JOD]LHW DO .DOHPER
DQG =JDPER 7KHVH ORQJLWXGLQDO IROORZXS SURJUDPV UHTXLUH FRPSOLDQFH DFURVV
ϭ
ZZZWXQDMDOLRUW]FDUHHUIDFWRUVDVVRVFLDWHGSOKLYORVWWUHDWPHQW
ϴ
Page 9 of 25
H[WHQGHGWLPHSHULRGVOLIHORQJIRU+,9$,'6WUHDWPHQWZKLFKLVLQPDQ\ZD\VGHPDQGLQJ
IRU WKH SDWLHQWV )RU LQVWDQFH WKH\ PD\ KDYH WR UHSODFH LQFRPHJHQHUDWLQJ DFWLYLW\ ZLWK
H[SHQVLYH DQGRU WLPHFRQVXPLQJ WUDYHOV WR WKH IDFLOLW\ ,Q PDQ\ UXUDO DUHDV VHDVRQDO ZRUN
PLJUDWLRQLVDOVRFRPPRQ)RULQVWDQFHWKHUXUDOSRSXODWLRQRIRQHRIWKHVWXG\GLVWULFWRIWHQ
PLJUDWHV WR DQRWKHU UHJLRQ EHWZHHQ 2FWREHU DQG )HEUXDU\ DQG DJDLQ EHWZHHQ 0D\ DQG -XO\
IRU IDUPLQJ DQG KDUYHVWLQJ 'XULQJ WKHVH SHULRGV WKHUH LV D SHDN LQ WKH GURSRXW UDWHV IRU
KHDOWKVHUYLFHV,QGLIIHUHQWPDQQHUEXWZLWKVLPLODUHIIHFWWRWKHFDUHWUDMHFWRU\FOLHQWVIURP
WKH RWKHU XUEDQ GLVWULFW PD\ PRYH EHWZHHQ IDFLOLWLHV GXH WR WKHLU FRQYHQLHQFH VXFK DV
FRPELQLQJKHDOWKFDUHDQGGDLO\OLIHUHVSRQVLELOLWLHV
7KXV GHIDXOWLQJ KDSSHQV IRU D YDULHW\ RI UHDVRQV EXW PRVW KHDOWK IDFLOLWLHV KDYHHVWDEOLVKHG
URXWLQHV WR IROORZ XS WKH GHIDXOWHUV 8VXDOO\ WKH ILUVW VWHS LV WR LGHQWLI\ FOLHQWV ZKR DUH
RYHUGXH LQ WKHLU VFKHGXOHV EDVHG RQ ZKDW LV UHFRUGHG LQ WKH UHJLVWHU ERRNV 7KHQ WKH QH[W
VWHSLVWRORFDWHRUFRQWDFWWKHPE\UHWULHYLQJLQIRUPDWLRQDERXWWKHLUDGGUHVVRUPRELOHSKRQH
QXPEHU/DVWO\WKHRXWFRPHRIWKHGHIDXOWHUWUDFLQJDFWLYLW\LVGRFXPHQWHGDQGWKHUHJLVWHUV
XSGDWHG ZLWK LQIRUPDWLRQ HJ DERXW VHUYLFHV RIIHUHG FOLHQW PRYHG RU GHFHDVHG +RZHYHU
WKHRIILFLDOLQIRUPDWLRQV\VWHPVDUHQRWGHVLJQHGWRVXSSRUWWKHLQIRUPDWLRQQHHGVDVVRFLDWHG
ZLWKWKHVHDFWLYLWLHV7KH\KDYHEHHQGHVLJQHGDFFRUGLQJWRDUHSRUWLQJORJLFFRQFHUQHGZLWK
GDWDJDWKHULQJDQGDJJUHJDWLRQIRUWKHLQIRUPDWLRQQHHGVRIKLJKHUOHYHOVLQWKHKHDOWKV\VWHP
KLHUDUFK\ 7KH KHDOWK ZRUNHUV QHHG LQIRUPDWLRQ WR VXSSRUW D SDWLHQW RU WUDMHFWRU\FHQWHUHG
PRGH RI ZRUNLQJ DQG PRVW LPSRUWDQWO\ WR EH DEOH WR KDQGOH GHYLDWLRQV IURP WKH SODQQHG
SURFHVVHV 7KLV DFWLYLW\ RI GHIDXOWHU WUDFLQJ UHTXLUHV DGGLWLRQDO LQIRUPDWLRQ HOHPHQWV DQG
WKHUHIRUH XQRIILFLDOLPSURYLVHG DGGLWLRQV WRWKHRIILFLDOLQIRUPDWLRQV\VWHPDUH ZLGHVSUHDG
DFURVV307&7IDFLOLWLHV
r
ee
rP
Fo
'DWDFROOHFWLRQPHWKRGV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
:H FRQGXFWHG DQ LQWHUSUHWLYH DQG TXDOLWDWLYH VWXG\ LQ RUGHU WR H[SORUH WKH SKHQRPHQRQ ZH
ZHUH LQWHUHVWHG LQ :DOVKDP 0\HUV &UHVVZHOO :H KDYH HPSOR\HG
HWKQRJUDSKLF PHWKRGV VXFK DV SDUWLFLSDQW REVHUYDWLRQ LQWHUYLHZV DQG GRFXPHQW DQDO\VLV
DFURVVIRXUVWXG\VLWHVIXUWKHUGHVFULEHGEHORZ+RZHYHUWKHPDLQGDWDFROOHFWLRQWHFKQLTXH
HPSOR\HG ZDV SDUWLFLSDQW REVHUYDWLRQ E\ WKH ILUVW DXWKRU %U\PDQ FDWHJRUL]HV D
SDUWLFLSDQW REVHUYHU¶V UROHV LQWR FRPSOHWH SDUWLFLSDQW SDUWLFLSDQWDVREVHUYHU REVHUYHUDV
SDUWLFLSDQWDQGFRPSOHWHREVHUYHU$FRPSOHWHSDUWLFLSDQWLVDIXOO\IXQFWLRQLQJPHPEHURID
VRFLDOVHWWLQJDQGRWKHUPHPEHUVDUHQRWDZDUHRIWKHUHVHDUFKHU¶VLGHQWLW\$SDUWLFLSDQWDV
REVHUYHULVVLPLODUWRFRPSOHWHSDUWLFLSDQWEXWRWKHUPHPEHUVDUHDZDUHRIUHVHDUFKHU¶VVWDWXV
$QREVHUYHUDVSDUWLFLSDQWUHVHDUFKHUSHUIRUPVVRPHREVHUYDWLRQZLWKYHU\OLWWOHSDUWLFLSDWLRQ
ZKLOHLQFRPSOHWHREVHUYDWLRQWKHUHVHDUFKHUGRHVQRWLQWHUDFWZLWKSHRSOH,QDOOIDFLOLWLHVWKH
UROHVRIWKHILUVWDXWKRUDOWHUHGEHWZHHQREVHUYHUDVSDUWLFLSDQWDQGSDUWLFLSDQWDVREVHUYHU,Q
VRPH FDVHV VKH SDUWLFLSDWHG LQ DFWXDO GDWD UHJLVWUDWLRQ GXULQJ VHUYLFH GHOLYHU\ DQG WKLV
LQYROYHPHQW LQ WKH DFWLYLWLHV LPSURYHG KHU XQGHUVWDQGLQJ RI 307&7 KHDOWK ZRUN SUDFWLFHV
DQG LQIRUPDWLRQ SUDFWLFHV 7KH RWKHU DXWKRUV SDUWLFLSDWHG LQ D IHZ REVHUYDWLRQV ,Q WRWDO ZH
KDG VHVVLRQV RI REVHUYDWLRQ :KLOH GRLQJ REVHUYDWLRQV ZH VRPHWLPHV FRQGXFWHG LQVLWX
LQWHUYLHZV ZLWK QXUVHVGRFWRUV ZRUNLQJ LQ WKH 307&7 VHFWLRQV 7KHUH ZHUH QR VSHFLILF
FULWHULDIRUVHOHFWLQJSDUWLFLSDQWVWREHLQWHUYLHZHGRUREVHUYHGRXUPDLQIRFXVZDVZLWKZRUN
ϵ
Information Technology & People
SUDFWLFHV LQ WKH VHFWLRQ ZLWK DQ\ VWDII RQ GXW\ 6RPH FRQYHUVDWLRQV RFFXUULQJ GXULQJ
SDUWLFLSDQW REVHUYDWLRQ ZHUH UHFRUGHG DQG ODWHU WUDQVFULEHG DV H[WHQVLYH ILHOG QRWHV 7KH
FRQYHUVDWLRQV DQG LQWHUYLHZV ZHUH FRQGXFWHG LQ .LVZDKLOL :H DOVR FRPSOHPHQWHG RXU
REVHUYDWLRQVHVVLRQVZLWKSKRQHFRQYHUVDWLRQV,QWRWDOZHKDGFRQWDFWZLWKQXUVHVLQWKH
IDFLOLWLHV DQG FRPPXQLW\EDVHG FDUH FRRUGLQDWRUV :H DOVR FRQGXFWHG VHPLVWUXFWXUHG
LQWHUYLHZVZLWKKHDOWKDGPLQLVWUDWRUVDWWKHGLVWULFWDQGUHJLRQDOOHYHOV,QWRWDOZHFRQGXFWHG
LQWHUYLHZVZLWK5&+FRRUGLQDWRUVDWGLVWULFWOHYHOLQWHUYLHZVZLWK5&+FRRUGLQDWRUV
DWUHJLRQDODQG]RQHOHYHOVLQWHUYLHZVZLWK+,9$,'6FRRUGLQDWRUVDWGLVWULFWOHYHODQG
LQWHUYLHZZLWK+,9$,'6FRRUGLQDWRUVDWUHJLRQDOOHYHO
7KHVWXG\ZDVFRQGXFWHGGXULQJWKHSHULRGIURP$XJXVW±-DQXDU\-XO\±$XJXVW
DQG1RYHPEHU±ġ 'HFHPEHUDQGIUHTXHQWYLVLWVZHUHPDGHWRIRXUIDFLOLWLHV
GXULQJ WKLV SHULRG :H DOVR FRQGXFWHG D GHWDLOHG VWXG\ RI WKH LQIRUPDWLRQUHODWHG DUWHIDFWV
XVHGWRRUJDQL]HWKHZRUNRIVHUYLFHSURYLVLRQPRVWQRWDEO\WKHYDULRXVUHJLVWHUERRNV7KLV
W\SHRIGRFXPHQWUHYLHZVHUYHGWRJLYHXVFRUHNQRZOHGJHDERXWWKHGRFXPHQWDWLRQSUDFWLFHV
RIWKHKHDOWKZRUNHUV,QDGGLWLRQWRILHOGQRWHVDGLDU\ZDVNHSWGXULQJWKHHQWLUHILHOGZRUN
LQRUGHUWRGRFXPHQWWKHGDWDFROOHFWLRQSURFHVV%RWK(QJOLVKDQG.LVZDKLOLZHUHXVHGLQWKH
ILHOGQRWHV/DWHUWKHQRWHVZHUHWUDQVODWHGLQWRHWKQRJUDSKLFGHVFULSWLRQLQ(QJOLVK
r
ee
rP
Fo
6HOHFWLRQRIVLWHV
7KHVWXG\UHSRUWHGLQWKLVSDSHUZDVFRQGXFWHGLQ,JRPEHEZDZDDQG.D]LPDEZDZDGLVWULFWV
LQWKH'RGRPD5HJLRQRI7DQ]DQLD)RXUGLIIHUHQWKHDOWKIDFLOLWLHVZHUHLQFOXGHGLQWKLVVWXG\
7KH\ ZHUH FKRVHQ WR VSDQ RXW D YDULHW\ RI VHWWLQJV ERWK JHRJUDSKLFDO UXUDOXUEDQ DQG
RUJDQL]DWLRQDOGLVSHQVDU\KHDOWKFHQWUH2IRXUREVHUYDWLRQVZHUHLQ0WRIDFLOLW\LQ
%ZDZDIDFLOLW\LQ&KHPFKHPIDFLOLW\DQGLQ.RURQJRIDFLOLW\QDPHVRIGLVWULFWVDQG
IDFLOLWLHVDUHQRW WKHLU UHDO QDPHV ,Q WDEOH VRPH UHOHYDQWIDFWVDERXW WKHIRXU FHQWUHVDUH
SURYLGHG
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 10 of 25
7DEOH6RFLR(FRQRPLF3URILOHRIIRXUIDFLOLWLHVWKDWZHUHVWXGLHG
'LVWULFW
5XUDO8UEDQ
'HPRJUD
SK\
.D]LPDEZDZD
8UEDQ
$UHD
NP
3RSXODWLRQ
/LWHUDF\
UDWHLV
+,9
SUHYDOHQFH
+HDOWK
)DFLOLWLHV
9LVLWHG
072
1XPEHURI
2EVHUYDWLRQV
6HUYLFHV
%:$:$
2XWSDWLHQWFOLQLF
5HSURGXFWLYHDQG&KLOG+HDOWK
FOLQLFLQFOGHOLYHU\ZDUGDQG
307&7+,9$,'6FOLQLF
GHQWDOFOLQLFWXEHUFXORVLV7%
DQGOHSURV\FOLQLFUHIHUUDOIRU
LQSDWLHQWFDUHWRZRPHQZKR
JLYHELUWKWKURXJKFDHVDUHDQ
VHFWLRQUHIHUUDOXQLWIRUVHYHUDO
SXEOLFDQGSULYDWHIDFLOLWLHVLQ
WRZQ
2XWSDWLHQWFOLQLF
5HSURGXFWLYHDQG&KLOG+HDOWK
FOLQLFLQFOGHOLYHU\ZDUGDQG
307&75HIHUSDWLHQWVWR0WR
KHDOWKFHQWUHRURWKHU+,9
FOLQLFVLQWRZQ
$YDLODEOH
5HVRXUFHVIRU
307&7
6HSDUDWHURRPIRU
307&7VHUYLFHV
7ZRQXUVHVIRU
307&7SUHJQDQW
ZRPHQDQG
307&7FKLOGUHQ
IROORZXS
UHVSHFWLYHO\
6DPHURRPXVHG
IRU307&7DQG
$1&VHUYLFHV2QH
QXUVHDOORFDWHGWR
ZRUNZLWK307&7
EXWZLOOZRUNZLWK
RWKHUWDVNVLIIUHH
ϭϬ
Page 11 of 25
,JRPEHEZDZD
5XUDO
$UHD
NP
3RSXODWLRQ
/LWHUDF\
UDWH
+,9
SUHYDOHQFH
&+(0&+(0
.2521*2
5HSURGXFWLYHDQG&KLOG+HDOWK
FOLQLFLQFOGHOLYHU\ZDUGDQG
307&7+,9$,'6FOLQLF
GHQWDOFOLQLFDVZHOODVRXW
SDWLHQWFOLQLFDQGLQSDWLHQW
GHSDUWPHQW5HIHUUDOXQLWIRU
IRXUWHHQGLVSHQVDULHVZLWKLQD
JHRJUDSKLFDOGLYLVLRQ,VWKH
RQO\&7&FOLQLFLQWKHDUHD
2XWSDWLHQWFOLQLFODERUDWRU\
XQLWDQG5&+VHFWLRQLQFO
GHOLYHU\ZDUGDQGOLPLWHG
307&7VHUYLFHV:LOOUHIHUWR
&KHPFKHP+&NPDZD\
$URRPDOORFDWHG
307&7DQG$1&
VHUYLFHV2QHQXUVH
DOORFDWHGZLOO
ZRUNRQRWKHUWDVNV
LIIUHHDOVRRWKHU
QXUVHVWUDLQHGWRGR
307&7
1RURRPDOORFDWHG
2QHQXUVHLV
DOORFDWHGEXWQRW
WUDLQHGZLOODOVR
GRRWKHUWDVNV
UHODWHGWRPDWHUQDO
DQGFKLOGKHDOWK
FDUH
'DWDDQDO\VLV
r
ee
rP
Fo
:H KDYH DQDO\VHG RXU GDWD E\ XVLQJ TXDOLWDWLYH WHFKQLTXHV 0LOHV DQG +XEHUPDQ %U\PDQ:DOVKDP%DVHGRQH[SDQGHGILHOGQRWHVFUHDWHGIURPSULPDU\QRWHV
ZHSUHSDUHGDQ³HWKQRJUDSKLFGHVFULSWLRQ´(PHUVRQHWDOIRUHDFKIDFLOLW\2XUIRFXV
ZDV RQ XQGHUVWDQGLQJ WKH UHODWLRQ EHWZHHQ LQIRUPDWLRQ QHHGV WRROV DQG ZRUN SUDFWLFHV $
SDUWLFXODU IRFXV ZDV RQ LGHQWLI\LQJ SUREOHPV ZLWK WKH WRROV XVHG IRU GHIDXOWHU WUDFLQJ 2XU
LQLWLDORULHQWDWLRQRIDQDO\VLVZDVWRZDUGVWKHWHQVLRQEHWZHHQWKHRIILFLDOKHDOWKLQIRUPDWLRQ
V\VWHP DQG WKH ZRUNUHODWHG GHPDQGV WKDW QHFHVVLWDWHG WKH XQRIILFLDO DGGLWLRQV DQG
PRGLILFDWLRQVRIWRROV:HFRQFHSWXDOL]HGWKHFKDOOHQJHVDVVRFLDWHGZLWKWUDFLQJRIGHIDXOWHUV
DVHPHUJLQJIURPWKHPHHWLQJRIWKHDFWLYLW\V\VWHPRI307&7FDUHGHOLYHU\DQGWKHDFWLYLW\
V\VWHPRIFOLHQWVOHDGLQJWRDQHZFKDQJHGREMHFWWKDWGURYHWKHHPHUJHQFHRIDQHZDFWLYLW\
GHIDXOWHUWUDFLQJ
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
+RZHYHUZKHQZRUNLQJZLWKWKHIRXUHWKQRJUDSKLFGHVFULSWLRQVZHUHDOL]HGWKHVLJQLILFDQW
GHJUHHRIYDULDWLRQVDPRQJWKHVLWHV7KLVOHDGWRDVHFRQGURXQGRIDQDO\VLVPRUHIRFXVHGRQ
WKH VLWXDWLRQDO LVVXHV :H VRXJKW WR XQGHUVWDQG ZKDW FKDUDFWHUL]HG WKH VLWXDWLRQ LQ ZKLFK
KHDOWKSURYLVLRQRFFXUUHGDQGHVSHFLDOO\VLWXDWLRQVZKHUHWKHDVVLJQHGWRROVZHUHQRWDEOHWR
PHGLDWH WKH DFWLYLW\ VDWLVIDFWRULO\ DQG RWKHU VROXWLRQV ZHUH IRXQG )RU WKLV DQDO\VLV ZH
VHDUFKHGIRUGLVWXUEDQFHVRUEUHDNGRZQVRIWKHDUWHIDFWVPHGLDWLQJUROHDFURVVDFWLYLW\DFWLRQ
DQGRSHUDWLRQOHYHOVDQGOLQNHGWKHVHZLWKFRQWH[WXDOFRQGLWLRQV
0HWKRGRORJLFDOUHIOHFWLRQV
7KH HPSLULFDO VWXG\ ZDV DSSURYHG E\ 8QLYHUVLW\ RI 'RGRPD¶V 5HVHDUFK DQG 3XEOLFDWLRQ
HWKLFDOFRPPLWWHH(WKLFDOFRQFHUQVZHUHNHSWLQPLQGGXULQJILHOGZRUNIRUH[DPSOHWKURXJK
JDLQLQJZULWWHQSHUPLVVLRQSULRUWRWKHILHOGZRUNIURPERWKGLVWULFWVPDQDJHUVDQGREWDLQLQJ
RUDO FRQVHQW IURP KHDOWK ZRUNHUV SULRU WR FRQGXFWLQJ LQWHUYLHZV RU REVHUYDWLRQV :H KDYH
SKRWRJUDSKHG UHJLVWHU ERRN SDJHV ZLWK SHUVRQDO LQIRUPDWLRQ EXW FRQFHDOHG WKH LGHQWLILDEOH
LQIRUPDWLRQZKHQSUHVHQWLQJRUSXEOLVKLQJWKHVHSKRWRV:HKDYHDOVRDQRQ\PL]HGWKHKHDOWK
ZRUNHUV DQG XVH SVHXGRQ\PV IRU WKH IDFLOLWLHV DQG GLVWULFWV ZKHUH ZH FRQGXFWHG RXU VWXG\
:KLOH WKH VWXG\ LV EDVHG RQ D VPDOO VHW RI FDVHV RXW RI KHDOWK IDFLOLWLHV LQ WKH WZR
ϭϭ
Information Technology & People
GLVWULFWV WKHVH FDVHV DUH VXIILFLHQWO\ GLIIHUHQWWR LQGLFDWHWKH UROH RI FRQWH[W 0RUHRYHU WKH
VWXG\LVEDVHGRQH[WHQGHGH[SRVXUHE\WKHILUVWDXWKRUWRWKHVHWWLQJDQGZRUNFRQGLWLRQVRI
WKHKHDOWKZRUNHUV$OVRWKHLQWHUPHGLDWHILQGLQJVIURPWKHVWXG\KDYHEHHQGLVFXVVHGZLWK
KHDOWK ZRUNHUV RI GLIIHUHQW OHYHOV RI WKH KHDOWK VHFWRU KHDOWK IDFLOLWLHV FRPPXQLW\ GLVWULFW
UHJLRQDODQG]RQHOHYHOV
ˆ‘”ƒ–‹‘’”ƒ…–‹…‡•‹†‹ˆˆ‡”‡–ˆƒ…‹Ž‹–‹‡•
3UDFWLFDO FKDOOHQJHV RQ WKH JURXQG LPSDFW ERWK VHUYLFH GHOLYHU\ DQG LQIRUPDWLRQ FROOHFWLRQ
DQG UHQGHU WKH RIILFLDO WRROV LQVXIILFLHQW IRU UHVROYLQJ WKH LQIRUPDWLRQ QHHGV RI WKH KHDOWK
ZRUNHUV 7KLV VHFWLRQ GHVFULEHV WKH REVHUYHG DGDSWDWLRQ RI LQIRUPDWLRQ DUWHIDFWV :H ILUVW
GHVFULEHWKHPRGLILFDWLRQVWKDWHPHUJHGLQRQHIDFLOLW\DQGWKHQKRZWKH\ZHUHWDNHQXSRU
QRWLQWKHRWKHUWKUHHIDFLOLWLHV)LQDOO\ZHUHWXUQWRWKHLQLWLDOIDFLOLW\DQGGHVFULEHKRZWKH
SURFHVVRIPRGLILFDWLRQVFRQWLQXHG
r
ee
rP
Fo
‹–‹ƒŽ‘†‹ˆ‹…ƒ–‹‘•‘ˆ–Š‡‘ˆˆ‹…‹ƒŽ”‡‰‹•–‡”•‹–‘
7KH0WRIDFLOLW\LVDEXV\XUEDQKHDOWKFHQWHUORFDWHGFORVHWRDSXEOLFWUDQVSRUWKXE+HUHWKH
DGDSWDWLRQV RI WKH RIILFLDO LQIRUPDWLRQ WRROV VWDUWHG 7KH RIILFLDO 307&7 UHJLVWHUV DUH
DQRQ\PRXVDQGZKHQDKHDOWKZRUNHUQHHGVWRFRQVXOWDUHJLVWHUERRNIRUDFHUWDLQFOLHQWVKH
GHSHQGVRQWKHFOLHQWWRSURYLGHWKH,'FDUGFRQWDLQLQJWKHLGHQWLILFDWLRQQXPEHU,QSUDFWLFH
WKHFOLHQWVGRQRWDOZD\VEULQJWKHLUFDUGV$KHDOWKZRUNHULQWKH0WRIDFLOLW\H[SODLQHG
³:HVWDUWHGVHUYLFHVLQ6HSWHPEHUZLWKD307&7FKLOGIROORZXSUHJLVWHU$WWKDWWLPH
SULYDF\ IRU 307&7 FOLHQWV ZHUH D PDMRU FRQFHUQ DQG KHQFH 307&7 UHJLVWHU KDG EHHQ
GHVLJQHGZLWKRXWQDPHV$IWHURQHZHHNZHUHDOL]HGWKDWZHQHHGHGH[WUDLQIRUPDWLRQEHFDXVH
VH IRXQG WKDW LW ZDV GLIILFXOW IRU RXU VWDII WR LGHQWLI\ FOLHQWV ZLWK QXPEHUV LQ WKH UHJLVWHU
EHFDXVHRIWKHQDWXUHRIWKHDFWLYLWLHV6RPHRIWKHFOLHQWVZRXOGFRPHZLWKRXWWKHLUFDUGVDQG
DVNIRUWKHLU UHVXOWV7HVWLQJNLWVZHUH VHQWWR'DU HV 6DODDPDQGLWZDVQRWFOHDUZKHQ WKH
UHVXOWV ZRXOG EH GXH ,W ZDV FRPPRQ IRU FOLHQWV WR SDVV E\ KHUH DQ\ WLPH WR DVN IRU WKHLU
UHVXOW´
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 12 of 25
:RUNHUVLQWKLVIDFLOLW\WKHUHIRUHFUHDWHGDORFDOUHJLVWHULQDQRUGLQDU\H[HUFLVHERRNERXJKW
LQWKHORFDOPDUNHW,QWKLVUHJLVWHUFOLHQWV¶IXOOQDPHVERWKWKHPRWKHUV¶DQGFKLOGUHQ¶VQDPHV
ZRXOGEHPDWFKHGZLWKWKHLUXQLTXHLGHQWLILFDWLRQQXPEHU7KLVZRXOGDOORZWKHZRUNHUVWR
ORRN XS WKHWHVW UHVXOWVLQWKHRIILFLDO UHJLVWHU ERRNHYHQ LIWKHFOLHQWGLG QRW EULQJ KHUFDUG
5DWKHUWKDQGRLQJWKLVWZRVWHSSURFHGXUHWKH\VWDUWHGWRUHJLVWHUDOVRWKHWHVWUHVXOWVLQWKH
ORFDO UHJLVWHU LQ RUGHU WR VDYHWLPH ZKHQ WKHFOLHQW DVNHGWKHP WR ORRN LWXS $FWXDOO\ WKHLU
RZQ QDPH IRU WKH UHJLVWHU ZDV ³'%6 UHJLVWHU´ VWDQGLQJ IRU 'U\ %ORRG 6SRW UHJLVWHU
LQGLFDWLQJWKHW\SHRI+,9WHVWXVHG
7KLV ORFDO '%6 UHJLVWHU DOVR FRQWDLQHG LQIRUPDWLRQ WKDW ZRXOG DOORZ WKH KHDOWK ZRUNHUV WR
WUDFN WKH FOLHQWV LQ FDVH WKH\ GLG QRW VKRZ XS IRU WKH VHUYLFH $ KHDOWK ZRUNHU H[SODLQHG
³$QRWKHUUHDVRQZHIRXQGRXWWKDWWRWUDFNFOLHQWVZHQHHGHGH[WUDLQIRUPDWLRQVXFKDVVWUHHWQDPH
WHQFHOOOHDGHUDQGSKRQHQXPEHU7KHUHZDVQRZKHUHVXFKLQIRUPDWLRQFRXOGEHIRXQG´7KHQDPH
RI WKH VWUHHW ZKHUH WKH FOLHQW OLYHG ZDV UHFRUGHG DQG DOVR WKH QDPH RI WKH DGPLQLVWUDWLYH
FRQWDFWSHUVRQLQWKHDUHDFDOOHGWHQFHOOOHDGHUKDPOHWOHDGHURUVWUHHWOHDGHUDVZHOODVWKH
SKRQHQXPEHURIWKHPRWKHUKHUVHOI
ϭϮ
Page 13 of 25
$V WKLVIDFLOLW\ ZDV WKHLQLWLDOVLWHIRU VWDUWLQJ307&7 VHUYLFHVLQWKHGLVWULFWLW VHUYHG DVD
WUDLQLQJFHQWUHIRUVWDIIIURPRWKHUIDFLOLWLHVLQWKHYLFLQLW\ZKHQ307&7VHUYLFHVZHUHUROOHG
RXW LQ $OVR WKH ORFDO '%6 UHJLVWHU ZDV SUHVHQWHG GXULQJ WKH WUDLQLQJ 7KURXJK WKLV
WUDLQLQJ WKH '%6 UHJLVWHU WKDW HPHUJHG LQ RQH IDFLOLW\ VSUHDG LQ WKH RWKHU WKUHH IDFLOLWLHV
)LJXUH,QWKHQH[WVHFWLRQVZHWXUQWRH[DPLQLQJWKHZD\WKH'%6UHJLVWHUZDVWDNHQXS
RUQRWLQRWKHUIDFLOLWLHV
r
ee
rP
Fo
&ŝŐƵƌĞϰ͗^ŝŵŝůĂƌůŽĐĂůƌĞŐŝƐƚĞƌƐŵĂĚĞĨƌŽŵůŽĐĂůůLJĂǀĂŝůĂďůĞĞdžĞƌĐŝƐĞŬƐ͕ĨŽƌƐƵƉƉŽƌƚŝŶŐ
ĚĞĨĂƵůƚĞƌƚƌĂĐŝŶŐŝŶƚŚƌĞĞĨĂĐŝůŝƚŝĞƐƌĞƐƉĞĐƚŝǀĞůLJ
ˆ‘”ƒ–‹‘’”ƒ…–‹…‡•‹–Š‡•‡”˜‹…‡•ƒ––Š‡‘–Š‡”ˆƒ…‹Ž‹–‹‡•
%ZDZDGLVSHQVDU\XUEDQ
$IWHUWKHLUWUDLQLQJDW0WRWKHVWDIILQ%ZDZDGLVSHQVDU\VWDUWHGWRXVHORFDOUHJLVWHUVVLPLODU
WRWKH RQHIRXQGDW0WR 7KH\ UHFRUGHGWKH UHVXOWVRIWHVWV LQWKH VDPH ZD\ DV 0WRIDFLOLW\
DQGWKHORFDOUHJLVWHUZDVXVHGWRVXSSRUW307&7VHUYLFHVILUVWIRUGHWHUPLQLQJFOLHQWV¶VWDWXV
DQGWKHQIRUVXSSRUWLQJGHIDXOWHUWUDFLQJDFWLYLWLHV+RZHYHUWKHUHZHUHIHZFOLHQWVUHFRUGHG
LQ WKH UHJLVWHU IURP WKH VWDUW LQ WR $SULO WKHUH ZHUH RQO\ FOLHQWV 7KH PRELOH
SKRQH QXPEHU ZDV UHFRUGHGIRU RQO\WZR RIWKHP 7RFRQWDFWWKH GHIDXOWHUV WKH VWDII ZRXOG
XVHKRPHYLVLWVRUHQFRGHGPHVVDJHWRSURWHFWSULYDF\WUDQVPLWWHGHLWKHULQDOHWWHURUDVRUDO
PHVVDJHFRQYH\HG WRFOLHQWV XVLQJ WKHLU QHLJKERUFOLHQWV 7KH UHDVRQV JLYHQ IRUIHZ FOLHQWV
UHJLVWHUHG ZHUH FRQQHFWHG WR WKH IDFW WKDW WKH GLVSHQVDU\ KDG QR &7& FOLQLF RI LWV RZQ
7KHUHIRUH+,9WUHDWPHQWFOLHQWVZHUHUHIHUUHGWRRWKHUIDFLOLWLHV7KHVWDIIDOORZHGFOLHQWVWR
YLVLWDQ\+,9$,'6FOLQLFRIWKHLUSUHIHUHQFHDQGWKHUHZHUHILYHIDFLOLWLHVZLWK&7&VHUYLFHV
LQWKHWRZQ,QIRUPDWLRQDERXWWKHFOLHQWVWKDWKDGEHHQUHIHUUHGWR0WRRURWKHUSODFHVZDV
QRW UHFRUGHG DQ\ZKHUH DQG WKH VWDII KDG QR SUDFWLFH RI IROORZLQJ XS WKHVH FOLHQWV WR WKH
IDFLOLW\ZKLFKFOLHQWVKDGEHHQUHIHUUHGWR+RZHYHUGXULQJRXUREVHUYDWLRQDW0WRZHQRWHG
WKDWWKHUHZHUHPDQ\!FOLHQWVWKDWKDGEHHQUHIHUUHGIURP%ZDZDWR0WRDQGUHJLVWHUHG
LQSDWLHQWUHJLVWULHV:HDOVRIRXQGWKDWFOLHQWVZHUHDGYLVHGE\0WRVWDIIWRFRPELQH307&7
DQG +,9$,'6 FOLQLF GD\ LQ RUGHU WR UHGXFH WKH EXUGHQ RI YLVLWLQJ WKH PDQ\ FOLQLFV 6LQFH
PRVWRIWKHUHIHUUHGFOLHQWVGLGQRWFRPHEDFNWR%ZDZDIRUWKHVHVHUYLFHVWKHVWDIIGLGQRW
DWWHPSW WR WUDFH WKHP 0RUHRYHU DV FOLHQWV KDG PXOWLSOH RSWLRQV DV WR ZKHUH WKH\ VHHN
+,9$,'6FOLQLFVHUYLFHVWU\LQJWRWUDFHWKHPZRXOGEHFKDOOHQJLQJ7KHUHZDVQRDWWHPSWWR
XVHWKHUHJLVWHUIRUWUDFLQJDQGWKHUHKDYHEHHQQRODWHUFKDQJHVWRLWVOD\RXWRUFRQWHQW
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
ϭϯ
Information Technology & People
&KHPFKHPKHDOWKFHQWUHUXUDO
7KH VWDIILQ&KHPFKHPKHDOWKFHQWUH XVHGWKH'%6 UHJLVWHUVWKDW WKH\ KHDUG RI GXULQJ WKHLU
WUDLQLQJDW0WR7KHOD\RXWDQGFRQWHQWVZHUHXQFKDQJHG)URPWKHVWDUWLQXQWLOHQGRI
$SULO D UHODWLYHO\ ORZ QXPEHU RI FOLHQWV KDG EHHQ UHFRUGHG KRZHYHU WKLV
FRPSULVHG WKH DFWXDO QXPEHU RI 307&7 FOLHQWV LQ WKH IDFLOLW\ 7KHUH ZHUH D ZHOOZRUNLQJ
LQIRUPDWLRQH[FKDQJHEHWZHHQWKH+,9$,'6FOLQLFDQG5&+FOLQLFVVLQFHWKHEXLOGLQJVZHUH
ORFDWHGQH[W WRHDFK RWKHUDQGWKH QXUVHV URWDWHGEHWZHHQWKHFOLQLFV 6LQFHWKHIDFLOLW\ ZDV
ORFDWHG LQ D UXUDO FRQWH[W ZLWK D FRQVLGHUDEOH GLVWDQFH WR WRZQ LW ZDV XQOLNHO\ WKDW FOLHQWV
ZRXOGJRWRRWKHU+,9$,'6FOLQLFV:KHQWKHKHDOWKZRUNHUVZRXOGWU\WRWUDFHDGHIDXOWHU
WKH\ZRXOGXVHERWKWKHORFDOUHJLVWHUDQGWKHRIILFLDOUHJLVWHU7KHWUDFLQJDQGIROORZXSRI
GHIDXOWHUV LQ &KHPFKHP GHSHQGHG VWURQJO\ RQ FRPPXQLW\ KHDOWK ZRUNHUV :KHQ WKH
GHIDXOWHUVKDGEHHQLGHQWLILHGWKHKHDOWKZRUNHUVZRXOGZULWHWKHQDPHVRQDSDSHUVOLSDQG
JLYHLWWRWKHFRPPXQLW\ZRUNHUWKDWZHUHUHVSRQVLEOHIRUIROORZLQJXS+,9$,'6SDWLHQWVLQ
JHQHUDO 7KH ZRUNHU ZRXOG VHHN RXW WKH PRWKHU XVXDOO\ LQ KHU KRPH RU LQ FRPPXQLW\ DQG
DVN KHU WR FRPH WR WKH FOLQLF ZLWK WKH FKLOG +RZHYHU WKHUH ZDV QR UHFRUGLQJ RU
GRFXPHQWDWLRQRIWKHRXWFRPHRIIROORZXSDFWLYLWLHV:HREVHUYHGDOVRRQVHYHUDORFFDVLRQV
WKDW WKH 307&7 QXUVH PDGH SKRQH FDOOV WR QXUVHV LQ RWKHU KHDOWK IDFLOLWLHV RU FRPPXQLW\
KHDOWK ZRUNHUV LQ WKH YLFLQLW\ 6KH DVNHG WKHP WR JLYH LQIRUPDWLRQ LI D FHUWDLQ ZRPDQ KDG
GHOLYHUHGDQGGHOLYHU\GHWDLOVDQGLQVRPHFDVHVVKHDVNHGWKHPWRWUDFHWKHFOLHQW7RVXSSRUW
WKLVWKHVWDIIKDGDOLVWRIWHOHSKRQHQXPEHUVRIWKHRWKHUIDFLOLWLHV
r
ee
rP
Fo
.RURQJRGLVSHQVDU\UXUDO
.RURQJR GLVSHQVDU\ LV ORFDWHG DERXW NLORPHWHUV IURP &KHPFKHP KHDOWK IDFLOLW\ DQG WKH
KHDOWKZRUNHULQWKH.RURQJRGLVSHQVDU\SUHIHUVWRUHIHUFOLHQWVWKDWQHHG307&7VHUYLFHWR
&KHPFKHP EHFDXVH VKH KDV QRW UHFHLYHG 307&7 WUDLQLQJ KHUVHOI :H IRXQG WKH RIILFLDO
307&7WRROVLQWKHIDFLOLW\KRZHYHUWKH\ZHUHQRWXVHG7KHKHDOWKZRUNHUH[SODLQHGWKDW
³,KDYHQHYHUUHFHLYHGDQ\WUDLQLQJLQ307&7VHUYLFHVVRKRZZLOO,RIIHU307&7FDUHDQG
ILOOWKHUHJLVWHU")RULQVWDQFH,PHWQXUVH%DQGVKHWROGPHWKDWQRZWKHVFKHGXOHIRULQIDQW
$59 SURSK\OD[LV KDYH FKDQJHG EXW , DP QRW DZDUH :KDW , GR LV WR PDNH VXUH WKDW FOLHQW
DWWHQG307&7DQG+,9$,'6FOLQLFLQ&KHPFKHP´6KHVDLG ³,NQRZDOOWKH307&7FOLHQWV
RIWKLVYLOODJH´DQGH[SODLQHGDERXW KHU DSSURDFK WRWUDFLQJUHIHUUHGFOLHQWVDQGLGHQWLI\LQJ
GHIDXOWHUV ³,XVXDOO\FRPPXQLFDWHZLWKWKHSURYLGHUVLQ&KHPFKHPWRHQVXUHWKDWWKHFOLHQW
KDVDWWHQGHGWKHFOLQLFLQFDVH,GRQ¶WVHHKHUKHUHLQWKHFOLQLF)RUWKHFKLOGUHQZKRDUHRQ
EUHDVWIHHGLQJ,XVHPRQLWRULQJTXHVWLRQV)RULQVWDQFH,DVNWKHPRWKHUDERXWWKHFRORXURI
WKHGUXJDERXWWKHGRVDJHVFKHGXOHDQGWKHDPRXQW,I,GRQRWUHFHLYHDVDWLVIDFWRU\DQVZHU
, PLJKW UHTXHVW WKH PRWKHU WR EULQJ WKH FDUG DQG ERWWOHV IRU FRWULPR[D]ROH V\UXS RU $59
SURSK\OD[LVWRWKHIDFLOLW\DVDSURRI´
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 14 of 25
Šƒ‰‹‰‹ˆ‘”ƒ–‹‘’”ƒ…–‹…‡•‹–Š‡–‘Š‡ƒŽ–Š…‡–”‡
:HREVHUYHGWKDWWKHVWDIIDW0WRHQFRXQWHUHGRWKHUFKDOOHQJHVWKDWOHGWRIXUWKHUFKDQJHVRI
WKHLQIRUPDWLRQWRROV0WRLVWKHEXVLHVWKHDOWKIDFLOLW\LQWKHGLVWULFWZLWKDGDLO\DWWHQGDQFH
RIDURXQGFOLHQWVLQWKHRXWSDWLHQWDQG5&+FOLQLFV7KHKHDOWKFHQWHULVORFDWHGQHDUWR
WKHPDUNHWDUHDDQGWKHUHIRUHLVHDV\WRUHDFKYLDSXEOLFWUDQVSRUW0DQ\QRQUHVLGHQWFOLHQWV
WKHUHIRUHSUHIHUWRXVHWKLVFHQWHU%HWZHHQDQG$SULOPRUHWKDQFKLOGUHQKDG
ϭϰ
Page 15 of 25
EHHQ HQUROOHG LQWR WKH 307&7 SURJUDP:H IRXQG WKUHH GLIIHUHQW UHJLVWHU ERRNV RQH ERRN
IRU6HSWHPEHU2FWREHUDVHFRQGERRNIRU1RYHPEHU'HFHPEHUDQGD
WKLUGERRNZDVVWDUWHGLQWKHEHJLQQLQJRI7KHVHUHJLVWHUVGLIIHUHGLQWKHLUOD\RXWDQG
FRQWHQW ,Q WKH VHFRQG UHJLVWHU ERRN VRPH QHZ FROXPQV KDG EHHQ LQWURGXFHG WKH GDWH WKDW
UHVXOWV ZHUH UHFHLYHGDWWKHIDFLOLW\ DQGWKHGDWHWKDW UHVXOWV ZHUH JLYHQWRFOLHQWV $ KHDOWK
ZRUNHUH[SODLQHGWKDW³$WRQHWLPHSHRSOHIURPWKHPLQLVWU\ZDQWHGXVWRUHSRUWQDPHVGDWH
IRUZKLFKUHVXOWVZHUHGXHDWIDFLOLW\DQGGDWHUHVXOWVZHUHJLYHQWRFOLHQWV7KHQZHGHFLGHG
WRDGGFROXPQVIRUWKDW´6LPLODUO\DKHDOWKZRUNHUH[SODLQHGWKDW ³,QPLG'HFHPEHU
WKHIDFLOLW\LQFKDUJHDVNHGXVDERXWQXPEHURIFOLHQWVZKRKDGEHHQWHVWHG+,9SRVLWLYHVLQFH
ZH VWDUWHG 307&7 VHUYLFHV LQ ,W ZDV GLIILFXOW WR ILQG RXW E\ JRLQJ WKURXJK GLIIHUHQW
SDJHV´ 7KH QHZ UHJLVWHU ERRN IURP -DQXDU\ ZDV GHVLJQHG VR WKDW LW RIIHUHG D EHWWHU
RYHUYLHZDOVRRYHUVXFKLQIRUPDWLRQ7KXVWKHVHFKDQJHVRFFXUUHGDVDUHVSRQVHWRFKDQJLQJ
GHPDQGVWRWKHUHSRUWLQJRIDFWLYLWLHV
r
ee
rP
Fo
:HDOVRREVHUYHGFKDQJHVLQWKHOD\RXWRIWKHUHJLVWHUERRNVZKLFKZHUHLQWHQGHGWRVXSSRUW
WKH VWDII¶V RZQ VHDUFKLQJ WDVN 0WR LV D ODUJH IDFLOLW\ ZKHUHPRVW YLVLWRUV DUH QRQUHVLGHQWV
DQGWKHVWDIIVWUXJJOHVWRDFKLHYHDGHTXDWHIROORZXSEHFDXVHRIWKHKLJKZRUNORDGDQGPDQ\
QRQUHVLGHQWV :KHQ D KHDOWK ZRUNHU VHHNV WR JR WKURXJK WKH ERRNV LQ RUGHU WR LGHQWLI\
GHIDXOWHUVRQHRIWKHLQIRUPDWLRQQHHGVLVWREHDEOHWRIROORZDSDUWLFXODUFOLHQW¶VWUDMHFWRU\
RYHUWLPH,QWKHILUVWERRNQHZFOLHQWVZKRFDPHIRUWHVWLQJDQGUHSHDWFOLHQWVZHUHPL[HG
LQWKHVDPHSDJHDQGVKHZRXOGKDYHWRVHDUFKPXOWLSOHSDJHVIRUWKHVDPH307&7LGHQWLILHU
+RZHYHUE\WKHWLPHWKH\VWDUWHGWRXVHWKHVHFRQGUHJLVWHUERRNWKH\FKDQJHGWKHOD\RXWRI
WKHERRNVRWKDWLWQRZKDGWKUHHW\SHVRISDJHVIRUQHZFOLHQWVIRUUHSHDWLQJFOLHQWVDQGIRU
RWKHU LQIRUPDWLRQ VXFK DV GRFXPHQWLQJ IROORZ XS RXWFRPH HWF $OVR WKLV ZD\ RI RUJDQL]LQJ
WKHLQIRUPDWLRQZDVQRWZLWKRXWSUREOHPV:HREVHUYHGDVLWXDWLRQZKHUHWKLVVWUXFWXUHKDG
OHGWRDSUREOHPGXHWRGXSOLFDWLRQRILQIRUPDWLRQRQHPRWKHUWRDFKLOGRQ307&7WKDWKDG
EHHQWHVWHGIRUWKHWKLUGWLPHLQ2FWREHUFDPHE\WKHIDFLOLW\IRXUPRQWKVODWHUWRDVNIRU
WKH WHVW 7KH UHVXOWV KDG DUULYHG ORQJ WLPH DJR EXW WKH KHDOWK ZRUNHUV KDG IRXQG QR SKRQH
QXPEHUUHFRUGHGIRUWKLVFOLHQWDQGZKHQVKHFDPHVKHZDVEODPHGIRUQRWKDYLQJUHJLVWHUHG
KHUSKRQHQXPEHU7KHPRWKHUDVVXUHGWKHKHDOWKZRUNHUVWKDWKHUSKRQHQXPEHUKDGLQGHHG
EHHQUHFRUGHG:KHQZHH[DPLQHGWKHORFDOUHJLVWHUZHIRXQGWKDWKHULQIRUPDWLRQKDGEHHQ
UHJLVWHUHGLQWKUHHVHSDUDWHSDJHV,QILUVWUHFRUGWKHSKRQHQXPEHUKDGEHHQUHFRUGHGZKLOH
LQWKHVHFRQGUHFRUGWKHSKRQHQXPEHUZDVQRWUHFRUGHG
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
7KH QHZ ORFDO UHJLVWHU VWDUWLQJ IURP -DQXDU\ KDG EHHQ GHVLJQHG ZLWK D QHZ OD\RXWLQ
RUGHUWRIDFLOLWDWHWKHKHDOWKZRUNHUV¶RYHUYLHZRIFOLHQWWHVWLQJDFWLYLWLHV,QWKHQHZUHJLVWHU
DOOLQIRUPDWLRQIRUDFHUWDLQFOLHQWZRXOGEHLQRQHSDJH7KHFKDQJHGOD\RXWZRXOGHQKDQFH
VHDUFKLQJRIFOLHQWVLQIRUPDWLRQSUHVHQWDWLRQDERXWUHVXOWVDQGIROORZXSLQIRUPDWLRQLQRQH
SDJH $ KHDOWK ZRUNHU H[SODLQHG LW OLNH WKLV ³,W LV DOVR GLIILFXOW WR JHW FOHDU SLFWXUH RI FOLHQWV¶
VHULHVRIWHVWLQJVFKHGXOHVDQGUHVXOWV7KLVSDJHJLYHVRYHUYLHZRISHUVRQDODWWULEXWHVWRJHWKHUZLWK
WHVWLQJVFKHGXOHVDQGUHVXOWV1RZDOOLQIRUPDWLRQLVLQRQHSDJH´7KHSUHFLVLRQRIWKHLQIRUPDWLRQ
ZDV DOVR LPSURYHG ³6RPH FOLHQWV GR QRW RZQ SKRQHV EXW ZH HQFRXUDJH WKHP WR UHJLVWHU SKRQH
QXPEHUVRIWKHLUUHODWLYHV+RZHYHUZHKDYHWRQRWHZKRWKHRZQHULVVRWKDWZHZLOONQRZKRZWR
FRPPXQLFDWHWKHPHVVDJH´)LJXUHVKRZVDVDPSOHSDJHLQWKHFKDQJHGUHJLVWU\
ϭϱ
Information Technology & People
0RELOH SKRQH QXPEHU
r
ee
rP
Fo
+XVEDQG¶V
SKRQHQXPEHU
+HUPRWKHU¶V
SKRQHQXPEHU
&ŝŐƵƌĞϱ͗ƉĂŐĞŝŶƚŚĞĐŚĂŶŐĞĚůĂLJŽƵƚŽĨƚŚĞƌĞŐŝƐƚƌLJ
7KH VWDII DW 0WR DOVR GHFLGHG WR FUHDWH D VHSDUDWH UHJLVWHU WR GRFXPHQW WKH DFWLRQV DQG
RXWFRPHV RI WUDFLQJ FOLHQWV 5HJXODUO\ QXUVHV ZRXOG VHDUFK WKURXJK WKH '%6 UHJLVWHU WR
GHWHUPLQH FOLHQWV ZKR ZHUH RYHUGXH LQ WKHLU VFKHGXOHV $V WKH ZRUNORDG LV WRR KLJK IRU WKH
0WRVWDIIWRSK\VLFDOO\WUDYHOLQWKHFLW\DQGVHHNRXWWKHFOLHQWVWKH\HPSOR\PRELOHSKRQH
EDVHG IROORZ XS LI D PRELOH SKRQH QXPEHU KDV EHHQ UHFRUGHG LQ WKH '%6 UHJLVWHU 7KH
RXWFRPHRIWKHIROORZXSDFWLYLW\ZDVGRFXPHQWHGLQWKHVHSDUDWHUHJLVWU\VLQFHVRPHFOLHQWV
FRXOG QRW EH UHDFKHG RU GLG QRW SLFN XS WKH SKRQH $QRWKHU DGGLWLRQDO UHJLVWHU KDG EHHQ
FUHDWHGWRVXSSRUWWUDFNLQJLQFDVHRIXQDYDLODELOLW\RIPHGLFDOUHVRXUFHVVXFKDVWHVWNLWVRU
GUXJV6RPHFOLHQWVKDGEHHQUHIHUUHGIURPRWKHUIDFLOLWLHVWR0WRLQRUGHUWRJHWQHYLUDSLQH
V\UXS $59 SURSK\OD[LV 7KH KHDOWKFDUH ZRUNHU H[SODLQHG WKDW ³6RPH FOLHQWV DUH UHIHUUHG
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 16 of 25
IURPRWKHUIDFLOLWLHVEXWWKHUHLVQRZKHUHWRUHFRUGDQG,FDQQRWUHFRUGLQ'%6UHJLVWHU7KH\KDYH
MXVW FRPH WR WDNH QHYLUDSLQH V\UXS , KDYH LQYHQWHG QHZ UHJLVWHU WR UHFRUG WKDW DFWLYLW\´ 7KLV
QHYLUDSLQHV\UXSUHJLVWHUZDVH[SDQGHGWRLQFOXGHDQRWKHUJURXSRIYLVLWRUVFOLHQWVZKRKDG
FRPH IRU WHVWLQJ EXW WKLV FRXOG QRW EH FRQGXFWHG EHFDXVH WKHUH ZHUH QR WHVWLQJ NLWV 7KH
KHDOWKFDUHZRUNHUVDLG ³,KDYHWRUHJLVWHUWKHVH,KDYHUHIHUUHGWKHPWRWKH0DMLKRVSLWDORUWRD
IDLWK EDVHG KRVSLWDO , FRXOG QRW UHJLVWHU WKHP LQ '%6 UHJLVWHU EHFDXVH WKH\ KDYH QRW WHVWHG , ZLOO
PDNHIROORZXSLIWKH\KDYHWHVWHGDQGDOVRLIQRWZKHQZHKDYHWHVWLQJNLWV,ZLOOLQIRUPWKHP´
ƒŽ›•‹•ǣˆ‘”ƒ–‹‘’”ƒ…–‹…‡•ƒ†…‘–‡š–
,Q WKH SUHYLRXV VHFWLRQ ZH KDYH GHVFULEHG WKH GHIDXOWHU WUDFLQJ SUDFWLFHV LQ IRXU GLIIHUHQW
KHDOWK IDFLOLWLHV IURP GLIIHUHQW JHRJUDSKLFDO VRFLRHFRQRPLF DQG LQVWLWXWLRQDO ORFDWLRQV ,Q
WKLVVHFWLRQZHZLOOIRFXVRQH[DPLQLQJWKHUROHRIFRQWH[WIRUWKHHPHUJHQFHRILQIRUPDWLRQ
ϭϲ
Page 17 of 25
UHODWHGSUDFWLFHVDQGWRROV:HZLOOGLIIHUHQWLDWHEHWZHHQWKUHHKLHUDUFKLFDOOHYHOVRI DFWLYLW\
DFWLRQV DQG RSHUDWLRQV /HRQWHY .DSWHOLQLQ DQG 1DUGL :KLOH WKH DFWLYLW\ DV
VXFK LV RULHQWHG WRZDUGV PRWLYHV DQGRU REMHFWLYHV WKH DFWLRQV DUH JHDUHG WRZDUGV UHDOL]LQJ
FRQFUHWHUHVXOWVWKDWFRUUHVSRQGVWRWKHVXEMHFW¶VJRDOV$OVRDFWLRQVFDQEHSRO\PRWLYDWHG
LH RQHDFWLRQ PD\ UHDOL]H VHYHUDO PRWLYHV %DUGUDPDQG 'RU\DE HPSKDVL]H WKDW WKH
SURFHVV RI UHDOL]LQJ WKH DFWLRQ WKURXJK FRQFUHWH RSHUDWLRQV LV VKDSHG E\ DQG DGMXVWHG WR WKH
PDWHULDOFRQGLWLRQVRIWKHVLWXDWLRQ:HZLOOKHUHWDNHDEURDGHUYLHZDQGDVNZKHWKHUFRQWH[W
LQIOXHQFH DOVR DFWLYLWLHV DQG DFWLRQV ,Q WDEOH ZH VXP XS WKH ILQGLQJV SUHVHQWHG LQ WKH
SUHYLRXVVHFWLRQ
7DEOH6XPPDU\RIWKH)LQGLQJV
+HDOWK)DFLOLW\
8VHRI7RROV
072
&HQWHU
0DLQ,VVXHV
+HDOWK 5HJXODU XVH RI WRROV IRU GHIDXOWHU
WUDFLQJ EXW FRQWLQXLQJ WHQVLRQV DQG
IXUWKHU PRGLILFDWLRQVDGGLWLRQV WR WRROV
OD\RXW RI '%6 UHJLVWHU DGGLWLRQDO
UHJLVWHUV
%:$:$
'LVSHQVDU\
.2521*2
'LVSHQVDU\
ew
vi
Re
&+(0&+(0
+HDOWK&HQWUH
+LJKQXPEHURIFOLHQWVQHFHVVLWDWHV
DWWHQWLRQ WR UHJLVWHU ERRNV¶
LQIRUPDWLRQOD\RXW+LJKQXPEHURI
QRQUHVLGHQW FOLHQWV DQG KLJK
PRELOLW\ RI FOLHQWV QHFHVVLWDWHV
GHYHORSPHQW RI DGGLWLRQDO WRROV WR
WUDFHWKHP
6WDII UHOLHV RQ PRELOH SKRQHV WR
FRQWDFWFOLHQWVGXHWRODUJHDUHDDQG
KLJKQXPEHURIFOLHQWV/DUJHUDUHD
DOVR PHDQV ZHDNHU OLQNV EHWZHHQ
FOLQLFDQGFRPPXQLW\
9HU\OLPLWHGXVHRIWRROVGXHWROLPLWHG &OLHQWVKDYHWREHUHIHUUHGWRRWKHU
GHIDXOWHUWUDFLQJ
IDFLOLWLHV WR DFFHVV &7& VHUYLFHV
&OLHQWV DOVR VHHN RWKHU VHUYLFHV
WKHUH DQG PRVW GR QRW UHWXUQ 7KH
DYDLODELOLW\RIVHYHUDOFOLQLFVPDNHV
WUDFLQJPRUHGLIILFXOW
r
ee
rP
Fo
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
5HJXODU XVH RI WRROV IRU GHIDXOWHU )RU FRQGXFWLQJ GHIDXOWHU WUDFLQJ
WUDFLQJ DQG UHODWLYHO\ ZHOO ZRUNLQJ SUDFWLFH WKH VWDII UHOLHV RQ
IROORZXS
FRPPXQLW\ KHDOWK ZRUNHUV WR
FRQWDFW FOLHQWV IHDVLEOH EHFDXVH
VWURQJOLQNVZLWKFRPPXQLW\
6SHFLILF LVVXH RI DUHD VHDVRQDO
ZRUNPLJUDWLRQ
1R XVH RI WRROV IRU GHIDXOWHU WUDFLQJ &OLHQWV DUH UHIHUUHG WR &KHPFKHP
VLQFHQR307&7VHUYLFHLVRIIHUHG
1XUVH PRQLWRUV DGKHUHQFH WR
307&7 EDVHG RQ SHUVRQDO
NQRZOHGJH RI FOLHQWV YLD FOLHQW
HQFRXQWHUV WKURXJK PRQLWRULQJ
TXHVWLRQV DQG FRQWDFW ZLWK
&KHPFKHPVWDII
&RQWH[WXDOLQIOXHQFHDWDFWLYLW\V\VWHPOHYHO
&OLHQWVWKDWGRQRWIROORZWKH307&7VFKHGXOHVUHVXOWVLQGLVWXUEDQFHVWRWKHQRUPDOIORZRI
307&7 FDUH WUDMHFWRU\ 7KH GHIDXOWHU WUDFLQJ SUDFWLFHV DLPV DW FRXQWHULQJ WKLV GLVWXUEDQFH
ZLWK WKH REMHFW RI HQVXULQJ WKDW FOLHQWV FRPSOHWH WKHLU FDUH WUDMHFWRULHV 7KH SUDFWLFH RI
GHIDXOWHU WUDFLQJ HPHUJHG DV D UHVXOW RI FRQWUDGLFWLRQV EHWZHHQ WKH 307&7 DFWLYLW\ V\VWHP
DQGWKH&OLHQWDFWLYLW\V\VWHP(QJHVWURPHWDO1LFROLQL+RZHYHULQWKHIRXU
ϭϳ
Information Technology & People
IDFLOLWLHVWKHVWDIIHQJDJHGLQGHIDXOWHUWUDFLQJLQGLIIHUHQWZD\VDQGWRGLIIHUHQWGHJUHHV)RU
LQVWDQFHLQ%ZDZDGLVSHQVDU\WKHFOLHQWVZHUHUHIHUUHGWRRWKHUIDFLOLWLHVQHDUE\DQGPRVWGLG
QRWUHWXUQWR%ZDZD'XHWR%ZDZD¶VJHRJUDSKLFDOORFDWLRQFORVHWRPDQ\RWKHUIDFLOLWLHVDQ
DWWHPSWWRIROORZXSWKHFOLHQWVZRXOGLQYROYHFRRUGLQDWLRQZLWKPDQ\IDFLOLWLHV7KHUHIRUHDV
D FRQVHTXHQFH RI WKH JHRJUDSKLFDO DQG RUJDQL]DWLRQDO ORFDWLRQ RI WKLV IDFLOLW\ XUEDQ
SHULSKHUDO XQLW ZLWK IHZ VHUYLFHV LQ WKLV IDFLOLW\ RQO\ D SDUWLDO GHIDXOWHU WUDFLQJ DFWLYLW\
V\VWHP KDG HPHUJHG ZKHUH WKH\ PRQLWRUHG RQO\ WKRVH FOLHQWV ZKR FRQWLQXHG WR XVH WKHLU
VHUYLFHV DQG QRW WKRVH WKDW KDG EHHQ UHIHUUHG 7KH GHIDXOWHU WUDFLQJ DFWLYLWLHV KDYH D VLPLODU
PRWLYHDFURVVIDFLOLWLHVWRGHWHFWGHIDXOWLQJFOLHQWVDQGFRQYLQFHWKHPWRUHVXPHVHUYLFH
)XUWKHUPRUH WKH VSHFLILF FRQILJXUDWLRQV RI WKH DFWLYLW\ V\VWHPV ZHUH KRZHYHU GLIIHUHQW
EHWZHHQ WKHVH IDFLOLWLHV 7R DQDO\]H WKLV GLIIHUHQFH IXUWKHU ZH ZLOO QHHG WR PRYH WR WKH
DQDO\WLFOHYHORIDFWLRQVDQGRSHUDWLRQV
r
ee
rP
Fo
&RQWH[WXDOLQIOXHQFHDWDFWLRQDQGRSHUDWLRQOHYHOV
7R IROORZ XS WKRVH FOLHQWV WKDW GHIDXOW WKH 307&7 VHUYLFH ZH GLVWLQJXLVK EHWZHHQ WKUHH
VHSDUDWH DFWLRQV HDFK VHHNLQJ WR DFKLHYH VSHFLILF FRQFUHWH UHVXOWV )LUVW WKHUH LV D QHHG WR
LGHQWLI\ ZKLFK FOLHQWV DUH RYHUGXH DFFRUGLQJ WR WKHLU VHUYLFH VFKHGXOH WKHQ WR WUDFH WKHPLQ
WKH VHQVH RI ORFDWLQJ WKHP SK\VLFDOO\ RU FRQWDFWLQJ WKHP E\ SKRQH RU RWKHU FRPPXQLFDWLRQ
PHDQVDQGILQDOO\WRRIIHUWKHUHTXLUHGVHUYLFHVDQGXSGDWHWKHUHJLVWHUV7KHWKUHHIDFLOLWLHV
ZKLFK WUDFH GHIDXOWHUV GLIIHU LQ KRZ HDFK RI WKHVH DFWLRQV DUH FDUULHG RXW ,Q 7DEOH ZH
SUHVHQW GLIIHUHQFHV IRU WKH RSHUDWLRQV IRU HDFK RI WKH WKUHH DFWLRQV 1RWH WKDW WKH FROXPQV¶
KHDGLQJ LQ WKH WDEOH SUHVHQW WKH WKUHH DFWLRQV IRUPLQJ WKH GHIDXOWHU WUDFLQJ DFWLYLW\ V\VWHP
ZKLOHWKHWH[WZLWKLQHDFKFROXPQUHSUHVHQWVRSHUDWLRQV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 18 of 25
dĂďůĞϯ͗&ĂĐŝůŝƚŝĞƐ͛ĚŝĨĨĞƌĞŶĐĞƐŝŶĐƚŝŽŶĂŶĚKƉĞƌĂƚŝŽŶƐǁŝƚŚŝŶĞĨĂƵůƚĞƌdƌĂĐŝŶŐĐƚŝǀŝƚŝĞƐ
,GHQWLI\LQJGHIDXOWHUV
0WR +HDOWK ,QWHUPLWWHQW VXUYH\V RI
UHJLVWHU ERRNV '%6
&HQWHU
UHJLVWHU FRQGXFWHG ZKHQ
WLPH DOORZV 6NLPPLQJ
HQWULHV WR GHWHFW ZKR LV
EHKLQG
LQ
VFKHGXOH
,PPHGLDWH FRQWDFWLQJ YLD
SKRQH &KDOOHQJH PDQ\
FOLHQWV
5HJXODU VXUYH\ RI UHJLVWHU
%ZDZD
'LVSHQVDU\ ERRN IRU WKH OLPLWHG
QXPEHU
RI
FOLHQWV
UHFRUGHG
&KHPFKHP
+HDOWK
&HQWHU
.RURQJR
7UDFLQJGHIDXOWHUV
8VLQJ PRELOH SKRQHV
QRW KRPH YLVLWV EHFDXVH
KLJK QXPEHU RI FOLHQWV
DQG ODUJH DUHD 8UEDQ
DUHD ZLWK PDQ\ RWKHU
DFWRUV
SXEOLF
DQG
1*2VWKXVZHDNHUOLQNV
EHWZHHQ FOLQLF DQG
FRPPXQLW\
6WDII XVH HQFRGHG OHWWHU
RU RUDO PHVVDJH YLD
FOLHQWV
'RFXPHQWLQJWKH2XWFRPH
7KH RXWFRPH RI IROORZXS
UHJLVWHUHG LQ D VHSDUDWH
UHJLVWHU FUHDWHG WR GRFXPHQW
IROORZXS
DFWLYLWLHV:KHQ
FOLHQWV UHWXUQ WR WKH FOLQLF
GRFXPHQWDWLRQ SURFHHGV DV
XVXDO
(PSOR\LQJ
WKH
FRPPXQLW\
KHDOWK
ZRUNHUV VXSSRUWHG E\
1*2
SURJUDP
IRU
+,9$,'6 FDUH WR
FRQWDFW SDWLHQWV DW KRPH
6RPH XVH RI PRELOH
SKRQHV
EXW
QRW
ZLGHVSUHDG
1XUVH NQRZV DOO 307&7 1XUVH
FRPPXQLFDWHV
FOLHQWV DQG FRQWLQXRXVO\ ZLWK SDWLHQWV ZKHQ
6WDII FRQGXFWV IROORZ XS EXW
GR QRW UHFRUG WKH RXWFRPH RI
IROORZ XS DFWLYLWLHV ,I FOLHQW
UHWXUQV WR VHUYLFH WKH VHUYLFH
SURYLVLRQLVGRFXPHQWHGLQWKH
XVXDOZD\
5HJXODUVXUYH\VRIUHJLVWHU
ERRNV ERWK 307&7 DQG
&7& ERRNV WR LGHQWLI\
GHIDXOWHUV 1DPHV DQG
FRQWDFW GHWDLOV ZULWWHQ RQ
SDSHUVOLS
7KH RXWFRPH LV QRW UHJLVWHUHG
DQ\ZKHUH
1XUVH PRQLWRUV RXWFRPH RI
GHIDXOWHU WUDFLQJ WKURXJK
ϭϴ
Page 19 of 25
'LVSHQVDU\
PRQLWRU WKHLU
DGKHUHQFH
VHUYLFH PHHWLQJWKHPLQPDWHUQDO FRQYHUVDWLRQ ZLWKRXW VSHFLILF
DQGFKLOGFDUHFOLQLFRULQ WRROV
YLOODJH
:H VHH WKDW WKH IDFLOLWLHV GLIIHU FRQVLGHUDEO\ ZKHQ ZH FRPSDUH DW WKH OHYHO RI DFWLRQV DQG
RSHUDWLRQVZLWKLQWKHGHIDXOWHUWUDFLQJDFWLYLW\V\VWHP7ZRUHODWLRQVVHHPWRLPSDFWWKLVPRVW
WKH UHODWLRQ EHWZHHQ WKH FOLQLF DQG WKH FOLHQWV DQG WKH UHODWLRQ EHWZHHQ WKH FOLQLF DQG WKH
FRPPXQLW\ KHDOWK V\VWHP ,Q WKH WZR UXUDO IDFLOLWLHV &KHPFKHP DQG .RURQJR WKH OLQNV
EHWZHHQ WKHFOLQLFDQGWKHFRPPXQLW\LVVWURQJDQG WKHDFWLRQ RI GHIDXOWHU WUDFLQJFDQ WDNH
DGYDQWDJH RI WKLV HLWKHU WKURXJK GLUHFW LQWHUDFWLRQV ZLWK WKH FOLHQWV DV LQ .RURQJR RU
WKURXJKPRELOL]DWLRQRIWKHFDGUHRIFRPPXQLW\KHDOWKZRUNHUVRUFRPPXQLW\OHDGHUVZKRLQ
&KHPFKHPDUHZHOONQRZQWRWKHVWDII,Q0WRZKLFKFDWHUVIRUDODUJHUSRSXODWLRQWKHUHLV
QRW VXFK D FORVH UHODWLRQ EHWZHHQ WKH FOLQLF DQG WKH FRPPXQLW\ KHDOWK ZRUNHUV ZKR PD\
EHORQJ WR GLIIHUHQW 1*2V DQGRU GLVWULFWV RU FRPPXQLW\ OHDGHUV ZKR DOVR PD\ KDYH OHVV
NQRZOHGJHRIWKHSRSXODWLRQWKDQDUXUDOFRPPXQLW\OHDGHU,QDGGLWLRQWKHKLJKQXPEHURI
FOLHQWVDQGWKHPL[WXUHRIUHVLGHQWDQGQRQUHVLGHQWFOLHQWVDUHSDUWLFXODUWR0WRWKHUHIRUHD
PDMRUUHDVRQIRUWKHGLIIHUHQFHEHWZHHQ0WRDQGWKHWZRRWKHUIDFLOLWLHV:HZLOOQRZWXUQWR
DPRUHGHWDLOHGDQDO\VLVDWWKHOHYHORIRSHUDWLRQVRIKRZWKHWRROVSHUIRUPDQGDUHXVHG
r
ee
rP
Fo
7KHVPDOOQXPEHURIFOLHQWVDW.RURQJRPDGHDOOWKHDFWLRQVRIGHIDXOWHUWUDFLQJLGHQWLI\LQJ
GHIDXOWHUV WUDFLQJ GHIDXOWHUV DQG PRQLWRULQJ RXWFRPH GRDEOH ZLWKRXW DQ\ LQIRUPDWLRQ WRRO
WR VXSSRUW WKH DFWLYLW\ ,Q &KHPFKHP WKH '%6 UHJLVWHU LQ LWV RULJLQDO OD\RXW ZRUNHG ILQH
KRZHYHU LQ 0WR WKH LGHQWLILFDWLRQ RI GHIDXOWLQJ FOLHQWV ZDV FKDOOHQJLQJ 2QH RI WKH FRUH
RSHUDWLRQV LQ TXHVWLRQ ZDV WKH VHDUFKLQJ WKURXJK UHJLVWHU ERRNV 7KH RULJLQDO '%6 UHJLVWHU
KDG WKH IROORZLQJ ILHOGV SHUVRQDO LQIRUPDWLRQ GHPRJUDSKLF LQIRUPDWLRQ PRELOH SKRQH
QXPEHUDQGWHVWUHVXOWZKLFKFDQEHIRUVWQGRUUGWHVWV7KHFOLHQWVZHUHUHFRUGHGLQ
PXOWLSOH URZV GHSHQGLQJ RQ WKH QXPEHU RI WHVWV ZKLFK KDG EHHQ FRQGXFWHG VR WR LGHQWLI\
GHIDXOWHUVKHDOWKZRUNHUVZRXOGVHDUFKWKHURZVWRGHWHUPLQHRYHUGXHFOLHQWVE\FRPSDULQJ
DJHRIWKHFKLOGDQGWKHQXPEHURIWHVWVZKLFKKDGEHHQSHUIRUPHG,Q0WRWKHVDPHWRROWKDW
ZRUNHGZHOOLQ&KHPFKHPGLGQRWZRUNZHOOHQRXJK7KLVVKRZVD³PHGLDWLRQEUHDNGRZQ´
RIWKHWRROGXHWRLWVFRQWH[WRIXVHKLJKQXPEHURIFOLHQWV7KHKHDOWKZRUNHUVLPSURYLVHGD
FKDQJHWRWKHOD\RXWRIWKHUHJLVWHUERRNVRWKDWLWZDVUHLQVWDOOHGDVDVXFFHVVIXOPHGLDWRURI
WKHLU ZRUN 7KH QHZ UHJLVWHU KDG DOO LQIRUPDWLRQ LQ RQH SDJH SHUVRQDO LQIRUPDWLRQ PRELOH
SKRQHDQGGHPRJUDSKLFLQIRUPDWLRQRQWKHXSSHUSDUWZLWKWHVWLQJVFKHGXOHVDQGUHVXOWVLQWKH
ORZHUSDUW7KLVZRXOGFRQILQHWKHVHDUFKRSHUDWLRQWRRQHSDJHUDWKHUWKDQDFURVVPXOWLSOH
SDJHVDVWKHSUHYLRXVIRUPDW:HDOVRVDZYDULDWLRQLQWHUPVRIKRZKHDOWKZRUNHUVFKRVHWR
FRQWDFW WKH FOLHQW :KHWKHU PRELOH SKRQHV ZHUH ZLGHVSUHDG GHWHUPLQHG ZKHWKHU WKH\ FRXOG
XVHLWWRFRQWDFWWKHFOLHQWVRUWKHLUIDPLO\RUQHLJKERXUV$OWHUQDWLYHO\WKH\ZRXOGFRQWDFW
WKHFOLHQWVLQSHUVRQRUHPSOR\WKHKRPHEDVHGFDUHZRUNHUVLQ+,9$,'6SURJUDPV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
‹•…—••‹‘
2XUFDVHDQDO\VLVGHPRQVWUDWHVWKDWWKHLQIOXHQFHRIFRQWH[WKDSSHQVDWGLIIHUHQWOHYHOVZLWKLQ
WKH KLHUDUFK\ RI DQ DFWLYLW\ V\VWHP )LUVWO\ VRPH DVSHFWV RI FRQWH[W VKRZHG LPSDFW DW WKH
OHYHORIWKHDFWLYLW\V\VWHPLWVHOI)RULQVWDQFHLQWKHSULPDU\KHDOWKFDUHVHWWLQJRI7DQ]DQLD
ϭϵ
Information Technology & People
WKH FOLHQWV¶ DFWLYLW\ V\VWHP ZDV VKDSHG E\ WKH VRFLRHFRQRPLF FRQWH[W VXFK DV SDWWHUQV RI
ZRUN PLJUDWLRQ SRSXODWLRQ PRELOLW\ SXEOLF WUDQVSRUW DQG WKH VRFLDO VWLJPD RI +,9$,'6
7KLV DIIHFWHG WKH KHDOWKFDUH DFWLYLW\ V\VWHP WKDW VRXJKW WR RIIHU 307&7 VHUYLFHV DQG WKH
FRQWUDGLFWLRQV EHWZHHQ WKHVH WZR FRXOG PDQLIHVW LQ WKH SUREOHP RI GHIDXOWLQJ FOLHQWV $
GHIDXOWHU WUDFNLQJ DFWLYLW\ V\VWHP HPHUJHG WR FRXQWHU WKLV 6SHFLILF WRROV DQG SUDFWLFHV WKDW
ZHUH QRW SDUW RI WKH RIILFLDO WRROV ZHUH GHYHORSHG E\ WKH KHDOWK ZRUNHUV WKHPVHOYHV 7KH
EURDGHUFRQWH[WDOVRFRXOGKDYHGLUHFWLPSDFWRQWKH307&7DFWLYLW\V\VWHPLWVHOIHJLQWKH
H[DPSOHVZKHUHWKHKHDOWKDXWKRULWLHVKDGQRWVXSSOLHGHQRXJKWHVWNLWVRUGUXJVWRRSHUDWHWKH
VHUYLFHVDVWKH\ZHUHVXSSRVHGWR6HFRQGO\FRQWH[WLPSDFWHGWKHZRUNRI307&7DOVRDWDQ
DFWLRQOHYHO7KHKHDOWKZRUNHUV¶FKRLFHRIDFWLRQVGHSHQGHGRQDVSHFWVRIWKHFRQWH[WVXFKDV
DYDLODELOLW\RIPRELOHSKRQHVDPRQJFOLHQWVDYDLODELOLW\RIFKHDSH[HUFLVHERRNVWKHUHODWLRQV
WR DQG UHVRXUFHV LQ WKH FRPPXQLW\ 7KHVH DVSHFWV VKDSHG WKH ZD\ WKH GHIDXOWHU WUDFLQJ
DFWLYLW\ V\VWHP ZDV GHVLJQHG DW WKH DFWLRQ OHYHO ZKLFK VKRZHG JUHDW YDULDWLRQ EHWZHHQ
IDFLOLWLHV 7KLUGO\ FRQWH[W ZDV LQIOXHQWLDO DOVR DW WKH RSHUDWLRQ OHYHO ZKHUH WKH VDPH WRRO
PHGLDWHG WKH DFWLYLW\ VXIILFLHQWO\ ZHOO LQ RQH IDFLOLW\ DQG QRW LQ DQRWKHU ,Q RXU HPSLULFDO
PDWHULDO PHGLDWLRQ EUHDNGRZQ UHODWHG WR LQIRUPDWLRQ WRROV FDQ RFFXU ERWK DW DFWLRQ DQG
RSHUDWLRQ OHYHO :KHQ LW RFFXUV DW WKH DFWLRQ OHYHO LW PD\ VWLPXODWH WKH VHDUFK IRU RU
GHYHORSPHQWRIDQHZWRROVXFKDVZKHQWKH'%6UHJLVWHUZDVFUHDWHGZKHQFOLHQWVGLGQRW
DOZD\V EULQJ WKHLU ,' FDUG WKDW ZDV D SUHUHTXLVLWH IRU WKH RIILFLDO DQRQ\PRXV UHJLVWHU WR
PHGLDWH WKH DFWLRQ :KHQ PHGLDWLRQ EUHDNGRZQ KDSSHQV DW DQ RSHUDWLRQ OHYHO LW PD\
VWLPXODWHWKH DGDSWDWLRQRI WKH WRRO :H VDZ WKDWWKHIRUPDW OD\RXWDQGRUJDQL]DWLRQ RIWKH
LQIRUPDWLRQZHUHPRGLILHG
r
ee
rP
Fo
/HRQWHY VWDWHG WKDW WKH VLWXDWLRQ DQG FRQGLWLRQV GHILQH KRZ DQ DFWLYLW\ LV DFFRPSOLVKHG
/HRQWµHYDQGDOVR(QJHVWU|PHPSKDVL]HGWKDWDFWLYLW\V\VWHPVKDYHDKLVWRU\DQGDUH
VLWXDWHG/LNHZLVH1DUGLVXJJHVWVWKDWFRQWH[WLVWKHDFWLYLW\V\VWHPLWVHOIDQGH[WHUQDO
FRQGLWLRQVDURXQGLW+RZHYHU7LHGHPDQQFODLPVWKDWWKLVLVQRWVXIILFLHQWO\VSHFLILF
:H DUJXH WKDW WR XQGHUVWDQG WKH LQIOXHQFH RI FRQWH[W RQO\ DW WKH OHYHO RI WKH VWUXFWXUH RI
DFWLYLW\LVWRREURDGDQGGRHVQRWSURYLGHDVXIILFLHQWO\FRQFUHWHDQGGHWDLOHGXQGHUVWDQGLQJ
RIWKHLQIOXHQFHRIFRQWH[W2XUDQDO\VLVFRQILUPVWKHQHHGWRLQFUHDVHWKHJUDQXODULW\DQGWR
VHDUFKIRULQIOXHQFHVRIFRQWH[WDOVRDWWKHDFWLRQDQGRSHUDWLRQOHYHOLQRUGHUWRXQGHUVWDQG
KRZ DUWHIDFWV DUH XVHG WR PHGLDWH ZRUN ,QYHVWLJDWLQJ LQVWDQFHV RI PHGLDWLRQ EUHDNGRZQV
KHOSHGXVSRLQWWRWKHVSHFLILFFRQFUHWHFRQGLWLRQVZKLFKFRQVWLWXWHGFKDOOHQJHVWRVXFFHVVIXO
ZRUNLQJRILQIRUPDWLRQWRROVDQGSUDFWLFHVLQDFHUWDLQFRQWH[W7KXVRXUZRUNFRQILUPVWKH
FODLPVE\UHVHDUFKHUVVXFKDV.RIIRG3HWHUVHQDQG&DVVHQ+XDQJDQG*DUWLQHU
DQG%DUGUDPDQG'RU\DEWKDWWKHLQIOXHQFHRIFRQWH[WVKRXOGEHVWXGLHGDWWKHDFWLRQV
DQG RSHUDWLRQV OHYHOV 6SHFLILFDOO\ ZH GHPRQVWUDWHG WKDW WKH LQIOXHQFH RI FRQWH[W FDQ EH
LQYHVWLJDWHG DW D YHU\ FRQFUHWH OHYHO E\ ORRNLQJ DW PHGLDWLRQ EUHDNGRZQV LQ ZRUNLQJ
VLWXDWLRQV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 20 of 25
:HFRQWULEXWHWRWKHOLWHUDWXUHRIDFWLYLW\WKHRU\E\H[WHQGLQJWKHGLVFXVVLRQVHHIRUH[DPSOH
.RIIRG3HWHUVHQDQG&DVVHQ&DVVHQDQG.RIIRG3HWHUVHQ+XDQJDQG*DUWLQHU
DURXQG FRQWH[W :KLOH WKLV GLVFXVVLRQ LV OLPLWHG WR WKH XQGHUVWDQGLQJ RI KRZ FRQWH[W
FDQ EH PRGHOOHG LQWR DUWHIDFWV &KHQ HW DO ZH GHVFULEH WKH UROH RI FRQWH[W DQG LWV
LPSDFWLQWKHXVHVLWXDWLRQ$OVRRXUUHVHDUFKVXJJHVWWKDWLGHQWLILFDWLRQRIFRQWUDGLFWLRQVDQG
ϮϬ
Page 21 of 25
PHGLDWLRQEUHDNGRZQVEHWZHHQFHQWUDODQGRWKHULQWHUDFWLQJDFWLYLW\V\VWHPVFDQ EHXVHGWR
LGHQWLI\LQIOXHQFHRIFRQWH[WLQZRUNSUDFWLFHV
,Q ,&7' XQGHUVWDQGLQJ RI ZRUN FRQWH[W LQIOXHQFLQJ WKH XVH RI FRPSXWHUL]HG V\VWHP LV
LPSRUWDQWLQRUGHUWRUHGXFHIDLOXUHRILQIRUPDWLRQV\VWHPV7LLKRQHQHWDO7KXVRXU
ZRUNFRQWULEXWHVWRLQIRUPDWLRQV\VWHPV¶VWXGLHVIRFXVLQJ RQDQDO\]LQJ LQIOXHQFH RIFRQWH[W
LQLQIRUPDWLRQSUDFWLFHVHJ7LLKRQHQHWDO.RUSHODHWDO7LLKRQHQHWDO
7LLKRQHQ 2QH H[DPSOH RI WKH SURSRVHG IUDPHZRUNV LV WKH /$&$6$ /HYHOV RI
$QDO\VLV &DWHJRULHV RI $QDO\VLV DQG 6FRSHV RI $QDO\VLV PRGHO WKDW FDQ EH XVHG WR DLG
LGHQWLILFDWLRQRIFRQWH[WXDOIHDWXUHVGXULQJLQIRUPDWLRQV\VWHPVGHYHORSPHQW7LLKRQHQHWDO
7LLKRQHQ HW DO 7LLKRQHQ +RZHYHU 7LLKRQHQHWDO LQGLFDWHG WKDW
WKHPRGHOZDVQRWXVHIXOWRDQDO\]HFRQWH[WXDOIDFWRUVRIWKHGDWDFROOHFWLRQWRROVUHJLVWHUV
ERRNVIRUPV7LLKRQHQHWDOS:HGRQRWZDQWWRGLVSXWHWKHFUHDWHGDQDO\WLFDO
IUDPHZRUNV IRU XQGHUVWDQGLQJ FRQWH[W LQ LQIRUPDWLRQ V\VWHPV VWXGLHV 5DWKHU RXU DQDO\VLV
KDYH VKRZQ WKDW DFWLYLW\ WKHRU\ FDQ EH XVHG WR DQDO\]H GDWD FROOHFWLRQ WRROV IRU LQVWDQFH WR
LGHQWLI\ FRQWH[WXDO IHDWXUHV ZKLFK LQIOXHQFH LWV XVDJH :H KDYH IXUWKHU HODERUDWHG KRZ
LQIRUPDWLRQWRROV¶XWLOL]DWLRQLVLQIOXHQFHGE\FRQWH[WXDOIHDWXUHV7KXVZHVXJJHVWWKDWLQ
,&7'ILHOGDFWLYLW\WKHRU\FDQDOVREHXVHGWRLGHQWLI\FRQWH[WXDOIHDWXUHVDQGLWVLQIOXHQFH
RQ ZRUN SUDFWLFHV ERWK DW PDFUR HJ VRFLRSROLWLFDO JHRJUDSKLFDO RUJDQL]DWLRQDO FXOWXUH
KXPDQUHVRXUFHVHFRQRP\WRWKHGHWDLOHGOHYHOPLFURZRUNSUDFWLFHVUHODWLRQVKLSEHWZHHQ
FOLHQWVDQGKHDOWKVHUYLFHV
r
ee
rP
Fo
’Ž‹…ƒ–‹‘•ƒ†‘…Ž—•‹‘•
7KH VWXG\ LOOXVWUDWHV WKH GHFLVLYH UROH WKDW FRQWH[W PD\ SOD\ LQ VKDSLQJ WKH DFWXDO XVDJH
LQIRUPDWLRQ WHFKQRORJ\ $Q DZDUHQHVV RI WKH UROH RI FRQWH[W PD\ OHDG WR PRUH UREXVW
DSSURDFKHV WR WKH LQWURGXFWLRQ RI ,&7 VROXWLRQV %\ IRFXVLQJ DQDO\WLFDOO\ RQ VLWXDWLRQV RI
PHGLDWLRQEUHDNGRZQLQWKHVLWXDWLRQRIXVHDWERWKDQDFWLYLW\DFWLRQDQGRSHUDWLRQOHYHOZH
KDYH DFKLHYHG DQ XQGHUVWDQGLQJ RI KRZ LQIRUPDWLRQ WRROV DUH EHLQJ DGDSWHG WR ERWK WKHLU
FRQWH[WXDOFRQGLWLRQVDQGWKHLQIRUPDWLRQQHHGVRIWKHFRPPXQLW\RIXVHUV7KLVXQGHUVFRUHV
WKHQHHGIRUOLQNLQJGHVLJQDQGLPSOHPHQWDWLRQDFWLYLWLHVWRDQXQGHUVWDQGLQJRIWKHVLWXDWLRQ
RQWKHJURXQGEHIRUHDQGDIWHUWKHLPSOHPHQWDWLRQHVSHFLDOO\LQORZUHVRXUFHVHWWLQJV
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
7KH VWXG\ KDV JHQHUDWHG D QXPEHU RI LPPHGLDWH GHVLJQ LQVLJKWV LQ WHUPV RI IXQFWLRQDOLW\
QHHGHG WR FDWHU IRU WKHVH SUDFWLFHV WKDW DUH QRW DGGUHVVHG LQ WKH RIILFLDO WRROV RU SROLF\
SURYLGHG E\ WKH 0LQLVWU\ RI +HDOWK )RU H[DPSOH GHVLJQ LQVLJKWV IRU WKH QHHG WR GHYHORS
JXLGHOLQHV DQG WRROV WR IDFLOLWDWH WKH GHIDXOWHU WUDFLQJ SUDFWLFHV ZLWKLQ PDWHUQDO DQG FKLOG
KHDOWKLQ7DQ]DQLDZHUHIRUPXODWHGUHI WREHDGGHG7KHGHVLJQLPSOLFDWLRQVIURPWKHVWXG\
KDV DOVR EHHQ IHG LQWR WKH RQJRLQJ GHVLJQ SURFHVV RI WKH '+,6 7UDFNHU PRGXOH UHI WR EH
DGGHGVRWKDWWKHVRIWZDUHFDQQRZRIIHUDVVLVWDQFHZLWKGHWHFWLRQRIRYHUGXHFOLHQWVJLYH
IOH[LELOLW\ WR UHFRUG WKH DSSURSULDWH FRQWDFW LQIRUPDWLRQ DQG FDQ JHQHUDWH GHIDXOWHU OLVWV IRU
SULQWRXW 7DEOH VXPPDUL]H WKH GHVLJQ DQG LPSOHPHQWDWLRQ LQVLJKWV JDLQHG IURP WKLV
XQGHUVWDQGLQJRIWKHLQIOXHQFHRIFRQWH[WLQZRUNSUDFWLFHV
Ϯϭ
Information Technology & People
dĂďůĞϰ͗ĞƐŝŐŶĂŶĚ/ŵƉůĞŵĞŶƚĂƚŝŽŶŝŶƐŝŐŚƚƐŐĂŝŶĞĚĨƌŽŵƚŚĞƐƚƵĚLJ
/HYHO
$FWLYLW\
&RQWH[WXDO$VSHFWV
,OOXVWUDWLRQ
'HIDXOWHU 7UDFLQJ &OLHQWV GR QRW DWWHQG VHUYLFHV
6\VWHP
DVVFKHGXOHG
7KH WUDMHFWRU\ RI FOLHQWV FDQ
EH H[WHQGHG WR RWKHU IDFLOLWLHV
GXH WR VKRUWDJH RI PHGLFDO
UHVRXUFHV
+HDOWK
SURYLGHUV
IDFH
GLIILFXOWLHV LQ FRPPXQLFDWLRQ
GXH WR ODUJHU JHRJUDSKLFDO
DUHDVLQXUEDQFRQWH[W
$FWLRQ
'HVLJQDQG,PSOHPHQWDWLRQ
'+,6 7UDFNHU FDQ EH DGRSWHG WR
IDFLOLWDWH WKH DFWLYLW\ WKURXJK H
WUDFLQJDQGDXWRPDWLFUHPLQGHUV
'+,6WUDFNHUFDQEHLPSOHPHQWHG
DQG
XVHG
WR
IDFLOLWDWH
FRPPXQLFDWLRQ DPRQJ KHDOWK
ZRUNHUV ZKR DUH LQ GLIIHUHQW
KHDOWKSURJUDPVDQGIDFLOLWLHV
,GHQWLI\LQJ&OLHQWV
*OREDO GHYHORSPHQW WHDP +HDOWK SURJUDPV VXPPDULHV
EHLQJ LQ (XURSH FRQWH[W KDV IXQFWLRQDOLW\ ZDV LQFOXGHG LQ WKH
OLPLWHG XQGHUVWDQGLQJ RI WKH VWDQGDUG'+,67UDFNHU
SULPDU\ KHDOWK IDFLOLWLHV¶
SUDFWLFHV DURXQG LGHQWLILFDWLRQ
GHIDXOWHU VLQFH WKHUH ZDV QR
IXQFWLRQDOLW\ LQ WKH '+,6
7UDFNHU WR IDFLOLWDWH WKH
SURFHVV
RI
LGHQWLI\LQJ
GHIDXOWHUV HJ RYHUYLHZ RI
FOLHQWV¶ DWWHQGDQFHV GURS RXW
UDWHV DQG GHIDXOWLQJ VWDWXV
ZLWKLQKHDOWKFDUHSURJUDPV
7UDFLQJ&OLHQWV
,Q UXUDO FRQWH[W IDFLOLW\ DQG
FRPPXQLW\OHDGHUVFROODERUDWH
LQ WUDFLQJ GHIDXOWHUV DQG
KDQGZULWWHQ GHIDXOWHU OLVW LV
FDUULHG E\ FRPPXQLW\ DFWRUV
HJ &+: ZKLOH FRQGXFWLQJ
KRPHYLVLWV
r
ee
rP
Fo
7KH VWDQGDUG '+,6 WUDFNHU ZDV
FKDQJHG WR KDYH IXQFWLRQDOLW\ WR
JHQHUDWH GHIDXOWHU OLVW UDWKHU WKDQ
YLHZLQJLWRQWKHVFUHHQ
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 22 of 25
)HZ FOLHQWV KDYH DFFHVV WR 5HFRQILJXUDWLRQ RI WKH V\VWHP WR
PRELOHSKRQHVLQUXUDODUHDV
FROOHFW QRQFOLHQW LQIRUPDWLRQ
HJ&+:FRPPXQLW\OHDGHU
'HVLJQ LPSOLFDWLRQ WR WKH JOREDO
WHDPWKDWWKHUHLVDQHHGWRFKDQJH
WKHVWDQGDUG'+,6V\VWHPVRWKDW
DXWRPDWHG 606 UHPLQGHUV WR
FOLHQWV ZKR PLVV WKHLU VFKHGXOHG
DSSRLQWPHQWV LQ 7DQ]DQLD FDQ
DOVR EH VHQW WR QRQFOLHQW DFWRUV
UDWKHUWKDQWRFOLHQWVRQO\
IRU 7KH VHDUFK RSHUDWLRQ LV 'HFLVLRQ WR XVH WKH V\VWHP WR
2SHUDWLRQ 6HDUFKLQJ
GHIDXOWHUQDPHV
GHSHQGLQJ RQ WKH PDQXDO LPSURYH WKH SURFHVV DQG UHPRYH
EDVHG UHJLVWU\ ZKLFK OHG WR PHGLDWLRQEUHDNGRZQV
HUURUDQGWHGLRXVSURFHVV
$OVR WKH VHDUFK RSHUDWLRQ LV
FRQVWUDLQHG E\ KLJKHU QXPEHU
RIFOLHQWVDQGEXV\VFKHGXOH
%H\RQGWKLVGLUHFWFRQWULEXWLRQWRGHVLJQRIDQ,&7VROXWLRQWKHVWXG\KDVDOVRFRQWULEXWHGWR
WKHGLVFRXUVHLQKHDOWKLQIRUPDWLRQV\VWHPV+,6LQGHYHORSLQJFRXQWULHVE\HPSKDVL]LQJWKH
ϮϮ
Page 23 of 25
FUXFLDOUROHRIWKHIURQWOLQHKHDOWKZRUNHUV¶RZQSUREOHPVROYLQJLQYHQWLRQDQGDGDSWDWLRQRI
LQIRUPDWLRQWRROV
$FNQRZOHGJHPHQW
:HWKDQNKHDOWKZRUNHUVFOLHQWVDQGRWKHULQIRUPDQWVQDPHVRIFROOHDJXHVWREHDGGHGDIWHUUHYLHZ
DQGWKHDQRQ\PRXVUHYLHZHUVIRUWKHLUFRQVWUXFWLYHFRPPHQWV
‡ˆ‡”‡…‡•
$OWHU67KHRU\RI:RUNDURXQGV&RPPXQLFDWLRQVRIWKH$VVRFLDWLRQIRU,QIRUPDWLRQ6\VWHPVSS
%DQQRQ /- DQG %¡GNHU 6 %H\RQG WKH LQWHUIDFH (QFRXQWHULQJ DUWLIDFWV LQ XVH '$,0, 5HSRUW
6HULHV
%DUGUDP - DQG 'RU\DE $ $FWLYLW\ $QDO\VLV ± $SSO\LQJ $FWLYLW\ 7KHRU\ WR $QDO\]H &RPSOH[ :RUN LQ
+RVSLWDOV&6&:0DUFK±+DQJ]KRX&KLQD
%OLQ ) DQG 0XQUR 0 :K\ KDVQ¶W WHFKQRORJ\ GLVUXSWHG DFDGHPLFV¶ WHDFKLQJ SUDFWLFHV" 8QGHUVWDQGLQJ
UHVLVWDQFHWRFKDQJHWKURXJKWKHOHQVRIDFWLYLW\WKHRU\&RPSXWHUVDQG(GXFDWLRQ
%RHU1, YDQ %DDOHQ3 -DQG.XPDU . $Q DFWLYLW\ WKHRU\ DSSURDFKIRU VWXG\LQJWKHVLWXDWHGQHVV RI
NQRZOHGJH VKDULQJ ,Q6\VWHP 6FLHQFHV +,&66 3URFHHGLQJV RI WKH WK $QQXDO +DZDLL ,QWHUQDWLRQDO
&RQIHUHQFHRQSS,(((
%UDD-0RQWHLUR(DQG6DKD\61HWZRUNVRI$FWLRQ6XVWDLQDELOLW\+HDOWK,QIRUPDWLRQ6\VWHPVDFURVV
'HYHORSLQJ&RXQWULHV0,64XDUWHUO\SS
%URZQ-6'XJXLG32UJDQL]DWLRQDOOHDUQLQJDQGFRPPXQLWLHVRISUDFWLFH7RZDUGDXQLILHGYLHZRI
ZRUNLQJOHDUQLQJDQGLQQRYDWLRQ2UJDQL]DWLRQVFLHQFHSS
%U\PDQ$6RFLDOUHVHDUFKPHWKRGVUG(G2[IRUG8QLYHUVLW\3UHVV1HZ<RUN
&DVVHQV - DQG .RIRG3HWHUVHQ $ 8VLQJ $FWLYLW\ 7KHRU\ WR 0RGHO &RQWH[W $ZDUHQHVV $ 4XDOLWDWLYH
&DVH6WXG\,Q)/$,56&RQIHUHQFHSS
&KHQ < /LX = DQG 9DLQLR - $FWLYLW\EDVHG FRQWH[WDZDUH PRGHO ,Q'HVLJQ 8VHU ([SHULHQFH DQG
8VDELOLW\'HVLJQ3KLORVRSK\0HWKRGVDQG7RROVSS6SULQJHU%HUOLQ+HLGHOEHUJ
&OXWV 00 7KH HYROXWLRQ RI DUWLIDFWV LQ FRRSHUDWLYH ZRUN FRQVWUXFWLQJ PHDQLQJ WKURXJK DFWLYLW\
,Q3URFHHGLQJVRIWKHLQWHUQDWLRQDO$&06,**5283FRQIHUHQFHRQ6XSSRUWLQJJURXSZRUNSS
$&0
&UHVZHOO -: 5HVHDUFK 'HVLJQ 4XDOLWDWLYH 4XDQWLWDWLYH DQG 0L[HG 0HWKRGV $SSURDFKHV &DOLIRUQLD
6DJH3XEOLFDWLRQV
'DPWHZ=$FFHOHUDWLQJ+HDOWK6HUYLFHDQG'DWD&DSWXULQJWKURXJK&RPPXQLW\+HDOWK:RUNHUVLQ5XUDO
(WKLRSLD $ 3UHUHTXLVLWH WR 3URJUHVV 3XEOLVKHG LQ WKH SURFHHGLQJV RI WKH ,QWHUQDWLRQDO &RQIHUHQFH RQ
.QRZOHGJH0DQDJHPHQWDQG,QIRUPDWLRQ6KDULQJ2FWREHU%DUFHORQD6SDLQSS
'H\ $. $ERZG * ' DQG 6DOEHU ' $ FRQFHSWXDO IUDPHZRUN DQG D WRRONLW IRU VXSSRUWLQJ WKH UDSLG
SURWRW\SLQJRIFRQWH[WDZDUHDSSOLFDWLRQV+XPDQFRPSXWHULQWHUDFWLRQ
'RU\DE $ DQG %DUGUDP - &RQWH[W DQG $FWLYLW\DZDUHQHVV LQ &ROODERUDWLYH (QYLURQPHQWV
6$*$ZDUH¶6HSWHPEHU%HLMLQJ&KLQD
(PHUVRQ50)UHW]5,DQG6KDZ//:ULWLQJHWKQRJUDSKLFILHOGQRWHV8QLYHUVLW\RI&KLFDJR3UHVV
(QJHVWU|P</HDUQLQJE\H[SDQGLQJ$QDFWLYLW\WKHRUHWLFDODSSURDFKWRGHYHORSPHQWDOUHVHDUFK+HOVLQNL
)LQODQG2ULHQWD.RQVXOWLW
(QJHVWU|P<([SDQVLYH/HDUQLQJDW:RUN7RZDUGDQDFWLYLW\WKHRUHWLFDOUHFRQFHSWXDOL]DWLRQ-RXUQDORI
(GXFDWLRQDQG:RUN
(QJHVWU|P<3XRQWL$DQG6HSSlQHQ/6SDWLDODQGWHPSRUDOH[SDQVLRQRIWKHREMHFWDVDFKDOOHQJHIRU
UHRUJDQL]LQJZRUN.QRZLQJLQRUJDQL]DWLRQV$SUDFWLFHEDVHGDSSURDFK
)ROFKHU 9 $SSURSULDWLQJ DUWLIDFWV DV LQVWUXPHQWV ZKHQ GHVLJQIRUXVH PHHWV GHVLJQLQXVH,QWHUDFWLQJ
ZLWKFRPSXWHUV
)RUVWHU0%DLOH\&%ULQNKRI0:*UDEHU&%RXOOH$6SRKU0%DOHVWUH(0D\0.HLVHU2-DKQ
$ DQG (JJHU 0 (OHFWURQLF PHGLFDO UHFRUG V\VWHPV GDWD TXDOLW\ DQG ORVV WR IROORZXS VXUYH\ RI
DQWLUHWURYLUDO WKHUDS\ SURJUDPPHV LQ UHVRXUFHOLPLWHG VHWWLQJV %XOOHWLQ RI WKH :RUOG +HDOWK 2UJDQL]DWLRQ KWWSZZZQFELQOPQLKJRYSXEPHG
r
ee
rP
Fo
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
Ϯϯ
Information Technology & People
*DVVHU / 7KH LQWHJUDWLRQ RI FRPSXWLQJ DQG URXWLQH ZRUN$&0 7UDQVDFWLRQV RQ ,QIRUPDWLRQ 6\VWHPV
72,6
*L]DZ$$0XNKHUMHH$/HZLV-DQG6DKD\68QGHUVWDQGLQJGHVLJQXVHLQWHUDFWLRQWKHFDVHRIQDPH
EDVHG KHDOWK LQIRUPDWLRQ V\VWHP LQ ,QGLD 3URFHHGLQJV RI ,$',6 ,QWHUQDWLRQDO &RQIHUHQFH RQ ,QIRUPDWLRQ
6\VWHPV%HUOLQ
+HHNV 5 ,QIRUPDWLRQ V\VWHPV DQG GHYHORSLQJ FRXQWULHV )DLOXUH VXFFHVV DQG ORFDO LPSURYLVDWLRQV 7KH
,QIRUPDWLRQ6RFLHW\
+HHNV5+HDOWKLQIRUPDWLRQV\VWHPV)DLOXUHVXFFHVVDQGLPSURYLVDWLRQ,QWHUQDWLRQDOMRXUQDORIPHGLFDO
LQIRUPDWLFV
+HOOH0'LVWXUEDQFHVDQGFRQWUDGLFWLRQVDVWRROVIRUXQGHUVWDQGLQJZRUNLQWKHQHZVURRP6FDQGLQDYLDQ
-RXUQDORI,QIRUPDWLRQ6\VWHPV
+XDQJ+DQG*DUWQHU*8VLQJDFWLYLW\WKHRU\WRLGHQWLI\UHOHYDQWFRQWH[WSDUDPHWHUV,Q/RFDWLRQ%DVHG
6HUYLFHVDQG7HOH&DUWRJUDSK\,,SS6SULQJHU%HUOLQ+HLGHOEHUJ
,JLUD )7 7KH VLWXDWHGQHVV RI ZRUN SUDFWLFHV DQG RUJDQL]DWLRQDO FXOWXUH LPSOLFDWLRQV IRU LQIRUPDWLRQ
V\VWHPVLQQRYDWLRQXSWDNH-RXUQDORI,QIRUPDWLRQ7HFKQRORJ\
-RQHV6$6KHUPDQ**DQG9DUJD&$([SORULQJVRFLRHFRQRPLFFRQGLWLRQVDQGSRRUIROORZXSUDWHV
RI+,9H[SRVHGLQIDQWVLQ-RKDQQHVEXUJ6RXWK$IULFD$,'6FDUH
.DDVE¡OO- &KDZDQL 06 +DPUH *$DQG6DQGYDQG - &RPSHWHQFLHVDQG /HDUQLQJIRU 0DQDJHPHQW
,QIRUPDWLRQ6\VWHPV,QIRUPDWLRQWHFKQRORJ\DQG2UJDQL]DWLRQV-RXUQDOSS
.DOHPER ): DQG =JDPER0 /RVVWR )ROORZ XS$ 0DMRU&KDOOHQJHWR6XFFHVVIXO ,PSOHPHQWDWLRQ RI
3UHYHQWLRQ RI 0RWKHUWR&KLOG 7UDQVPLVVLRQ RI +,9 3URJUDPV LQ 6XE6DKDUDQ $IULFD ,651 $,'6
KWWSG[GRLRUJ
.DQMR & 3UDJPDWLVP RU 3ROLF\ ,PSOLFDWLRQV RQ +HDOWK ,QIRUPDWLRQ 6\VWHPV 6XFFHVV7KH (OHFWURQLF
-RXUQDORI,QIRUPDWLRQ6\VWHPVLQ'HYHORSLQJ&RXQWULHV
.DSWHOLQLQ9DQG1DUGL%$FWLYLW\7KHRU\LQ+&,)XQGDPHQWDOVDQG5HIOHFWLRQV0RUJDQDQG&OD\SRRO
3XEOLVKHUV
.DSWHOLQLQ 9 1DUGL % $ DQG 0DFDXOD\ & 0HWKRGV WRROV 7KH DFWLYLW\ FKHFNOLVW D WRRO IRU
UHSUHVHQWLQJWKH³VSDFH´RIFRQWH[WLQWHUDFWLRQV
.DUDUD*9HUEHNH)DQG1\VVHQ0+RVSLWDO,QIRUPDWLRQ0DQDJHPHQW8VLQJ2SHQ6RXUFH6RIWZDUH
5HVXOWVRIWKH0,'$3URMHFWLQ+RVSLWDOVLQ5ZDQGD-RXUQDORI+HDOWK,QIRUPDWLFVLQ$IULFD
.RIRG3HWHUVHQ $ DQG &DVVHQV - 8VLQJ DFWLYLW\ WKHRU\ WR PRGHO FRQWH[W DZDUHQHVV ,Q0RGHOLQJ DQG
5HWULHYDORI&RQWH[WSS6SULQJHU%HUOLQ+HLGHOEHUJ
.RUSHOD 0 GH OD +DUSH 5 DQG /XXNNRQHQ , 'HSLFWLQJ WKH ODQGVFDSH DURXQG LQIRUPDWLRQ IORZV
0HWKRGRORJLFDOSURSRVLWLRQV
.XXWWL.$FWLYLW\WKHRU\DVDSRWHQWLDOIUDPHZRUNIRU KXPDQFRPSXWHULQWHUDFWLRQUHVHDUFK&RQWH[WDQG
FRQVFLRXVQHVV$FWLYLW\WKHRU\DQGKXPDQFRPSXWHULQWHUDFWLRQ
/HRQWHY$$FWLYLW\FRQVFLRXVQHVVDQGSHUVRQDOLW\(QJOHZRRG&OLIIV3UHQWLFH+DOO
/HRQWHY$7KHSUREOHPRIDFWLYLW\LQSV\FKRORJ\,Q-DPHV9:HUWVFK7KH&RQFHSWRI$FWLYLW\LQ6RYLHW
3V\FKRORJ\SS1HZ<RUN0(6KDUSH,QF
/XQG ++ DQG 6XWLQHQ ( &RQWH[WXDOLVHG ,&7' D ERWWRPXS DSSURDFK ,Q3URFHHGLQJV RI WKH WK
,QWHUQDWLRQDO&RQIHUHQFHRQ$SSOLHG&RPSXWHU6FLHQFH:6($6-DSDQ
0DUDLV0$QDO\VLVRIWKHIDFWRUVDIIHFWLQJWKHVXVWDLQDELOLW\RI,&7'LQLWLDWLYH3URFHHGLQJVRI,',$
³,&7IRUGHYHORSPHQWSHRSOHSROLF\DQGSUDFWLFH´
0LOHV0DQG+XEHUPDQ04XDOLWDWLYH'DWD$QDO\VLV6DJH3XEOLFDWLRQ8QLWHG6WDWHVRI$PHULFD
0LOQH & 6FDQWOHEXU\ . %ORQVWHLQ - DQG *OHDVRQ 6 &RWHDFKLQJ DQG GLVWXUEDQFHV %XLOGLQJ D EHWWHU
V\VWHPIRUOHDUQLQJWRWHDFKVFLHQFH5HVHDUFKLQ6FLHQFH(GXFDWLRQ
0R\R &0 $Q $VVHVVPHQW RI WKH 4XDOLW\ RI +HDOWK 0DQDJHPHQW ,QIRUPDWLRQ 6\VWHP 'DWD LQ 6HOHFWHG
+HDOWK )DFLOLWLHV LQ /LORQJZH 'LVWULFW 0DVWHU 7KHVLV 'HSDUWPHQW RI 3XEOLF +HDOWK &ROOHJH RI 0HGLFLQH
8QLYHUVLW\RI0DODZL
0SKDWVZH : 0DWH .6 %HQQHWW % 1JLGL + 5HGG\ - %DUNHU 30 DQG 5ROOLQV 1 ,PSURYLQJ
SXEOLF KHDOWK LQIRUPDWLRQ D GDWD TXDOLW\ LQWHUYHQWLRQ LQ .ZD=XOX±1DWDO 6RXWK $IULFD %XOOHWLQ RI WKH :RUOG
+HDOWK2UJDQL]DWLRQ$UWLFOH,'%/7
0\HUV 0' 4XDOLWDWLYH 5HVHDUFK LQ ,QIRUPDWLRQ 6\VWHPV 0DQDJHPHQW ,QIRUPDWLRQ 6\VWHPV4XDUWHUO\
±
r
ee
rP
Fo
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 24 of 25
Ϯϰ
Page 25 of 25
1DUGL%$(G&RQWH[WDQGFRQVFLRXVQHVV$FWLYLW\WKHRU\DQGKXPDQFRPSXWHULQWHUDFWLRQ&DPEULGJH
0$0,73UHVV
1JOD]L 0' .DSODQ 5 :RRG 5 %HNNHU / DQG /DZQ 6' ,GHQWLILFDWLRQ RI ORVVHV WR IROORZXS LQ D
FRPPXQLW\EDVHGDQWLUHWURYLUDOWKHUDS\FOLQLFLQ6RXWK$IULFDXVLQJDFRPSXWHUL]HGSKDUPDF\WUDFNLQJV\VWHP
%0&,QIHFWLRXV'LVHDVHVKWWSZZZELRPHGFHQWUDOFRP
1JRPD & DQG ,JLUD )7 &RQWUDGLFWLRQV LQ &ROOHFWLQJ DQG 5HFRUGLQJ 0DWHUQDO +HDOWK 'DWD DW WKH
&RPPXQLW\ /HYHO $&DVHVWXG\IURP WZR 'LVWULFWVLQ 7DQ]DQLD-RXUQDO RI +HDOWK ,QIRUPDWLFVLQ'HYHORSLQJ
&RXQWULHV
1LFROLQL ' 3UDFWLFH DV WKH 6LWH RI .QRZLQJ ,QVLJKWV IURP WKH )LHOG RI 7HOHPHGLFLQH2UJDQL]DWLRQ
6FLHQFH
1LFROLQL ' 3UDFWLFH 7KHRU\ :RUN DQG 2UJDQL]DWLRQ $Q ,QWURGXFWLRQ 2[IRUG 8QLYHUVLW\ 3UHVV 8QLWHG
.LQJGRP
3LWXOD . DQG 5DGKDNULVKQDQ 7 2Q HOLFLWLQJ UHTXLUHPHQWV IURP HQGXVHUV LQ WKH ,&7'
GRPDLQ5HTXLUHPHQWV(QJLQHHULQJ
6 E¡-,.RVVL(.7LWOHVWDG2+7RKRXUL55%UDD-&RPSDULQJVWUDWHJLHVWRLQWHJUDWHKHDOWK
LQIRUPDWLRQ V\VWHPV IROORZLQJ D GDWD ZDUHKRXVH DSSURDFK LQ IRXU FRXQWULHV,QIRUPDWLRQ 7HFKQRORJ\ IRU
'HYHORSPHQW
6DXHUERUQ5DQG/LSSHYHOG7,QWURGXFWLRQ,Q7/LSSHYHOG56DXHUERUQDQG&%RGDUWHGV'HVLJQ
DQG,PSOHPHQWDWLRQRI+HDOWK,QIRUPDWLRQ6\VWHP:RUOG+HDOWK2UJDQLVDWLRQ*HQHYD
6DXJHQH =% &XVWRPL]DWLRQ RI *HQHULF 2SHQ 6RXUFH 6RIWZDUH IRU +HDOWK 6HFWRU LQ 'HYHORSLQJ
&RXQWULHV'RFWRUDOGLVVHUWDWLRQ8QLYHUVLW\RI2VOR1RUZD\
6KR]L 1$ 3RWWDV ' DQG 0RVWHUW3KLSSV 1 3HUFHLYHG %HQHILWV 2I 5HPRWH 'DWD &DSWXULQJ ,Q
&RPPXQLW\+RPH%DVHG&DUH7KH&DUHJLYHUV¶3HUVSHFWLYH7KH-RXUQDORI&RPPXQLW\,QIRUPDWLFV
6LPED'2DQG0ZDQJX0$4XDOLW\RIDURXWLQHGDWDFROOHFWLRQV\VWHPIRUKHDOWKFDVHRI.LQRQGRQL
GLVWULFWLQWKH'DUHV6DODDPUHJLRQ7DQ]DQLD6RXWK$IULFDQ-RXUQDORI,QIRUPDWLRQ0DQDJHPHQW
6LQQLJ ' 3LWXOD . %HFNHU 5 5DGKDNULVKQDQ 7 DQG )RUEULJ 3 6WUXFWXUHG 'LJLWDO 6WRU\WHOOLQJ IRU
(OLFLWLQJ 6RIWZDUH 5HTXLUHPHQWV LQ WKH ,&7' 'RPDLQ ,Q+XPDQ&RPSXWHU ,QWHUDFWLRQSS 6SULQJHU
%HUOLQ+HLGHOEHUJ
6WDU 6/ %RZNHU *& DQG 1HXPDQQ /- 7UDQVSDUHQF\ EH\RQG WKH LQGLYLGXDO OHYHO RI VFDOH
&RQYHUJHQFHEHWZHHQLQIRUPDWLRQ$UWHIDFWVDQGFRPPXQLWLHVRISUDFWLFH'LJLWDOOLEUDU\XVH6RFLDOSUDFWLFHLQ
GHVLJQDQGHYDOXDWLRQ
7KRPVRQ.$&KHWL(2DQG5HLG7,PSOHPHQWDWLRQDQGRXWFRPHVRIDQDFWLYHGHIDXOWHUWUDFLQJV\VWHP
IRU +,9 SUHYHQWLRQ RI PRWKHU WR FKLOG WUDQVPLVVLRQ RI +,9 307&7 DQG 7% SDWLHQWV LQ .LEHUD 1DLUREL
.HQ\D7UDQVDFWLRQRIWKH5R\DO6RFLHW\RI7URSLFDO0HGLFLQHDQG+\JLHQH
7LHGHPDQQ -3 1HZ OLWHUDFLHV QHZ FRQWH[WV" $ WKHRUHWLFDO GHILQLWLRQ RI UHDGLQJ FRQWH[W'RFWRUDO
GLVVHUWDWLRQ9DQGHUELOW8QLYHUVLW\
7LLKRQHQ7,QIRUPDWLRQV\VWHPVLQFRQWH[W8QLYHUVLW\RI(DVWHUQ)LQODQG
7LLKRQHQ7.RUSHOD0DQG0XUVX$&UHDWLQJDIUDPHZRUNWRUHFRJQL]HFRQWH[WRULJLQDWHGIDFWRUVLQ,6
LQRUJDQL]DWLRQV,Q6RFLDO,QIRUPDWLFV$Q,QIRUPDWLRQ6RFLHW\IRUDOO",Q5HPHPEUDQFHRI5RE.OLQJSS
6SULQJHU86
7LLKRQHQ7/XXNNRQHQ,DQG.RUSHOD00RGHOOLQJWKHFRQWH[WRILQIRUPDWLRQV\VWHPV,Q,67$IULFD
SS,(((
7RQJLD5DQG6XEUDKPDQLDQ(,QIRUPDWLRQDQG&RPPXQLFDWLRQV7HFKQRORJ\IRU'HYHORSPHQW,&7'
$ GHVLJQ FKDOOHQJH" ,Q,QIRUPDWLRQ DQG &RPPXQLFDWLRQ 7HFKQRORJLHV DQG 'HYHORSPHQW ,&7'
,QWHUQDWLRQDO&RQIHUHQFHRQSS,(((
:DOVKDP*,QWHUSUHWLYHFDVHVWXGLHVLQ,6UHVHDUFKQDWXUHDQGPHWKRG(XURSHDQ-RXUQDORILQIRUPDWLRQ
V\VWHPV
:DOVKDP*'RLQJLQWHUSUHWLYHUHVHDUFK(XURSHDQMRXUQDORILQIRUPDWLRQV\VWHPV
r
ee
rP
Fo
ew
vi
Re
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Information Technology & People
Ϯϱ
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement