FACTORS ASSOCIATED WITH UNDER-5 MORTALITY IN SOUTH AFRICA: TRENDS 1997- 2002 by

FACTORS ASSOCIATED WITH UNDER-5 MORTALITY IN
SOUTH AFRICA: TRENDS 1997- 2002
by
Peter Buwembo
SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS
FOR THE DEGREE
MASTER OF SOCIAL SCIENCE IN SOCIOLOGY
IN THE FACULTY OF HUMANITIES
UNIVERSITY OF PRETORIA
PRETORIA
SUPERVISOR: DR NJ BOMELA
MARCH, 2010
© University of Pretoria
ACKNOWLEDGEMENTS
I would like to thank my supervisor, Dr Nolunkcwe Bomela, whose guidance,
encouragement, perseverance and support made it possible for me to finish my
thesis. I am grateful to the Head of department, Professor Janis Grobbelaar who
afforded me the opportunity to finish my Master’s Programme.
Lastly I would like to thank the management of Statistics South Africa for
providing me with the resources I required for my studies and for allowing me to
use their data.
ii
DECLARATION
This mini dissertation represents original work of the author and has not been
submitted in any form to another university. Where work of others has been used
it has been acknowledged and referenced in the text. The research for this study
was performed at the University of Pretoria during the period January 2009 to
January 2010.
iii
ABSTRACT
The objective of this study is to investigate the trends in relative contribution each
selected factor makes to the chance of a child’s death over time in South Africa
for children born 5 years preceding 1997 and 5 years preceding 2002. Attention
was paid to the role played by socio-economic factors, biological and maternal
factors, environmental factors, nutrient deficiency factors and health seeking
behaviour factors. The study investigates whether the association of a specific
factor to under-5 mortality persist over time.
Data from the 1997 October Household Survey and the 2002 General Household
Survey were used. Births that occurred in the five years preceding each survey
were analysed in relation to the survival of the child and socio-economic factors,
biological and maternal factors, environmental factors, nutrient deficiency factors
and health seeking behaviour factors. Logistic regression was used to determine
the relative contribution of each factor for the two periods under review.
Under-5 mortality was significantly associated with eight factors during 19931997 period namely; mother’s education, mother’s place of residence, sex, birth
order, birth interval, mother’s age at the time of delivery of the subject child,
nutrient deficiency and place of delivery. However, during the 1998-2002 period
only five factors were significantly associated with under-5 mortality. These were
mother’s education, sex, birth interval, type of dwelling and place of delivery. This
suggests changing patterns in factors associated with under-5 mortality between
the two birth cohorts: 1993-1997 and the 1998-2002 birth cohorts.
Key words
Under-5 mortality, Socio-economic, environmental, Maternal, Nutritional, Health-seeking,
HIV/AIDS, Household survey.
iv
Table of contents
Page
ACKNOWLEDGEMENTS ................................................................................................. ii
DECLARATION.................................................................................................................iii
ABSTRACT ...................................................................................................................... iv
LIST OF TABLES............................................................................................................. ix
LIST OF FIGURES............................................................................................................ x
ABBREVIATIONS/ ACRONYMS ..................................................................................... xi
CHAPTER 1 ......................................................................................................................1
1.
INTRODUCTION................................................................................................... 1
1.1 Background and trends in under-5 mortality ............................................................... 1
1.2 Outlining the problem .................................................................................................. 5
1.3 Mortality in young children and its measurements ...................................................... 6
1.3.1 Neonatal mortality................................................................................................. 7
1.3.2 Post-neonatal mortality ......................................................................................... 7
1.3.3 Infant mortality and child mortality ........................................................................ 8
1.3.4 Under-5 mortality .................................................................................................. 9
1.4 Objective of the study.................................................................................................. 9
1.5 Assumptions................................................................................................................ 9
1.6 Limitations ................................................................................................................. 10
1.7 Rationale of the study ............................................................................................... 10
1.8 Review of chapters.................................................................................................... 11
CHAPTER 2 .................................................................................................................... 12
2.
CONCEPTUAL FRAMEWORK AND REVIEW OF LITERATURE ...................... 12
2.1 Introduction ............................................................................................................... 12
2.2 Conceptual framework for the analysis of childhood mortality .................................. 12
2.3 Socioeconomic determinants of child survival .......................................................... 14
v
2.3.1 Mother’s education ............................................................................................. 14
2.3.2 Place of residence .............................................................................................. 15
2.3.3 Labour market status of the mother.................................................................... 16
2.4 Biological and Maternal determinants of child survival ............................................. 18
2.4.1 Birth order ........................................................................................................... 18
2.4.2 Birth interval........................................................................................................ 19
2.4.3 Age of the mother ............................................................................................... 20
2.4.4 Sex of the child ................................................................................................... 21
2.5 Environmental health determinants of child survival ................................................. 22
2.5.1 Source of water and access to sanitation ........................................................... 23
2.5.2 Source of energy ................................................................................................ 24
2.5.3 Type of dwelling.................................................................................................. 25
2.6 Nutrient deficiency as a determinant of child survival ............................................... 25
2.7 Healthy seeking behaviour as a determinant of child survival .................................. 26
2.8 Summary................................................................................................................... 27
CHAPTER 3 .................................................................................................................... 28
3.
METHODOLOGY ................................................................................................ 28
3.1 Introduction ............................................................................................................... 28
3.2 Background into the data .......................................................................................... 28
3.3 The sample ............................................................................................................... 29
3.3.1 October household survey 1997 sample ............................................................ 29
3.3.2 General household survey 2002 sample ............................................................ 29
3.4 Questionnaires .......................................................................................................... 30
3.5 Training, field work operations and procedures ........................................................ 31
3.6 Data quality ............................................................................................................... 33
3.7 Analysis strategy ....................................................................................................... 34
3.7.1 Logistic regression.............................................................................................. 34
3.8
Operational definition of the dependent and independent variables .................. 39
3.8.1 Dependent variable............................................................................................. 39
3.8.2 Independent variables ........................................................................................ 39
vi
3.9 Omitted Traditional Explanatory Variables................................................................ 42
3.9.1 Impact of AIDS on under-5 mortality................................................................... 43
3.10 Summary................................................................................................................. 44
CHAPTER 4 .................................................................................................................... 45
4.
DATA ANALYSIS ................................................................................................ 45
4.1 Introduction ............................................................................................................... 45
4.2 Descriptive analysis .................................................................................................. 45
4.2.1 Under-5 survival 1993-1997 and 1998-2002 ...................................................... 45
4.2.2 Births data from auxiliary source 1993-1997 and 1998-2002 ............................. 46
4.2.3 Trends in environmental factors 1997-2002 ....................................................... 47
4.2.4 Trends in selected socio-economic indicators 1997-2002 .................................. 49
4.2.5 Trends in biological and maternal factors (1997-2002) ...................................... 50
4.2.6 Trends in health seeking behaviour 1997-2002.................................................. 52
4.2.7 Trends in nutrient deficiency factor 1997-2002................................................... 53
4.3 Trends in proportion of under-5 deaths and associated factors................................ 53
4.3.1 Proportion of children who died under-5 by environmental factors..................... 54
4.3.2 Proportion of children who died under-5 by socio-economic factors .................. 57
4.3.3 Proportion of children who died under-5 by biological and maternal factors ...... 60
4.3.4 Proportion of children who died under-5 by nutrient deficiency factors .............. 63
4.3.5 Proportion of children who died under-5 by health seeking behaviours ............. 64
4.4 logistic regression model with one independent variable.......................................... 65
4.5 Logistic regression with multiple independent variables ........................................... 68
4.6 Evaluation of the Models........................................................................................... 78
4.7 Summary................................................................................................................... 79
CHAPTER 5 .................................................................................................................... 80
5.
DISCUSSIONS AND CONCLUSION .................................................................. 80
5.1 Introduction ............................................................................................................... 80
5.2 Summary of findings ................................................................................................. 80
5.3 Discussions and conclusion ...................................................................................... 81
5.3.1 Socio-economic factors and environmental factors ............................................ 81
5.3.2 Socio-economic factors and health seeking behaviour....................................... 82
vii
5.3.3 Biological and maternal factors........................................................................... 83
5.3.4 Health seeking behaviour and HIV/AIDS ............................................................ 84
5.4 Contribution of the study ........................................................................................... 84
References...................................................................................................................... 86
APPENDIX 1 ................................................................................................................... 90
APPENDIX 2 ................................................................................................................... 91
viii
LIST OF TABLES
Page
Table 3.1
Correlation between independent variable using 1997 data
36
Table 3.2
Correlation between independent variable using 2002 data
37
Table 3.3
Collinearity statistics for OHS 1997 and GHS 2002
38
Table 4.1
Survival of children under-5 for the cohorts 1993-1997 and 1998-2002
45
Table 4.2
Table 4.3
Birth occurrences (as at end of April 2008) 1993-2002 as recorded on the population
register
Trends in environmental factors 1997-2002
46
48
Table 4.4
Trends in socio-economic factors 1997-2002
49
Table 4.5
Trends in biological and maternal factors for birth cohorts 1993-1997 and 1998-2002
51
Table 4.6
The odds of under-5 death for the birth cohorts 1993-1997 and 1998-2002: model with
one independent variable
The odds of under-5 death for the birth cohorts 1993-1997 : model with multiple
independent variables
The odds of under-5 death for the birth cohorts 1998-2002: model with multiple
independent variables
Table 4.7
Table 4.8
ix
66
71
73
LIST OF FIGURES
Page
13
Figure 2.1
Analytical Frame work for child survival
Figure 4.1
Proportions of children not delivered in hospital or clinic birth cohorts 1993-1997 and
1998-2002
52
Proportions of children in household which experienced food shortage: birth cohorts
1993-1997 and 1998-2002
53
Proportions of under-5 deaths by access to piped water: birth cohorts 1993-1997 and
1998-2002
54
Proportions of under-5 deaths by type of dwelling: birth cohorts 1993-1997 and 19982002
55
Proportions of under-5 deaths by type of sanitation facilities: birth cohorts 1993-1997
and 1998-2002
56
Proportions of under-5 deaths by access to electricity: birth cohorts 1993-1997 and
1998-2002
56
Proportion of under-5 deaths by mother’s level of education: birth cohorts 1993-1997
and 1998-2002
58
Proportions of under-5 deaths by mother's employment status: birth cohorts 1993-1997
and 1998-2002
58
Proportions of under-5 deaths by mother's residence: birth cohorts 1993-1997 and
1998-2002
59
Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8
Figure 4.9
Figure 4.10
Proportion of under-5 deaths by Mother's age at birth: birth cohorts 1993-1997 and
1998-2002
60
Figure 4.11
Proportion of under-5 deaths by birth interval: Birth cohorts 1993-1997 and 1998-2002
61
Figure 4.12
Proportions of under-5 deaths by birth order: Birth cohorts 1993-1997 and 1998-2002
62
Figure 4.13
Proportions of under-5 deaths by sex: birth cohorts 1993-1997 and 1998-2002
62
Figure 4.14
Proportions of under-5 deaths by nutritional status: birth cohorts 1993-1997 and 19982002
63
Proportions of under-5 deaths by place of delivery: birth cohorts 1993-1997 and 19982002
64
Figure 4.15
x
ABBREVIATIONS/ ACRONYMS
AIDS
Acquired Immune Deficiency Syndrome
ASSA
Actuarial Society of South Africa
CSS
Central Statistics Service
EA
Enumeration Area
GHS
General Household Survey
HIV
Human Immunodeficiency Virus
MDG
Millennium Development Goal
OHS
October Household Survey
PSU
Primary Sampling Area
SADC
Southern African Development Community
SADHS
South African Demographic and Health Survey
SAS
Statistical Analysis Software
SPSS
Statistical Package for Social Scientists
UNICEF
United Nations Children’s Fund
WHO
World Health Organisation
xi
CHAPTER 1
1. INTRODUCTION
1.1 Background and trends in under-5 mortality
It is estimated that worldwide 10.5 million children aged 0-4 years died in 1999,
about 2.2 million or 17.5% less than a decade ago. According to a UNICEF1,
WHO2, The World Bank and UN Population Division (2007) report, world wide the
number of children dying before the age of five has reached a record low, falling
below 10 million for the first time in 2006.
The decline has however not been evenly distributed. There are still some
regional differentials. The WHO (2006) estimates that on average about 15% of
newborn children in Africa are expected to die before reaching their fifth birthday.
The corresponding figures for many parts of the developing world are in the
range of 3-8% and for Europe under 2%. Infant and under five mortality rates are
by far the highest in Sub Saharan Africa, where underdevelopment, armed
conflict and the spread of HIV/AIDS have seriously undermined efforts to improve
child survival. The estimated under-five mortality rate exceeds 200 deaths per 1
000 live births3 in ten countries in this region. Infant and child mortality rates also
remain relatively high in South Asia.
Omar, et al. (2000) give a good summary of the global trend of infant and child
mortality. There have been dramatic declines in mortality in almost all countries
of the world, regardless of initial levels, socio-economic circumstances and
development strategies.
In advanced economies the declines were already
1
United Nations Children’s Fund,
World Health Organisation
3
A live birth is the complete expulsion or extraction from its mother of a product of conception, irrespective of the duration
2
of pregnancy, which, after such separation, breathes or shows any other evidence of life-such as beating of the heart,
pulsation of the umbilical cord, or definite movement of voluntary muscles-whether or not the umbilical cord has been cut
or the placenta is attached. Each product of such a birth is considered a live birth.
1
apparent at the end of the 19th century. In the period leading to 1970 under-5
mortality in these countries was 27 deaths per 1 000 live births, this declined to
10 deaths per 1 000 live births in the period 1970-1990. In the decade that
followed under-5 mortality declined to 7 deaths per 1 000 live births. The 2006
figures indicate that under-5 mortality further declined to 6 deaths per 1 000 live
births. The decline took place during the time of steady economic growth and
major improvements in nutrition, housing, and living conditions. Garenne and
Gakusi (2005) point out that during the above developments, the first benefits
were improvements in water and sanitation, hygiene, and child feeding practices
and the development of vaccinations. UNICEF (2004), identifies new, better,
more effective and costly medicines, technology, and interventions as the main
contributor to the steady decline in mortality rates in the industrialized countries
during the period 1990-2003.
In the developing countries, under-5 mortality has declined, however it is still high
compared to the developed regions. In the period 1960-1970 under-5 mortality
was 164 deaths per 1 000 live births and it further declined to 128 deaths per 1
000 live births during 1970-1980 period. In the following decade a further decline
to 103 deaths per 1 000 live births was observed. The downward trend continued
beyond the year 1990. The 2006 figures indicate that under-5 mortality in this
region declined to 79 deaths per 1 000 live births. However, East Asia and the
Pacific, Latin America and the Caribbean, and Central/Eastern Europe and the
Commonwealth of Independent States had achieved under-five mortality rates of
below 30 deaths per 1 000 live births. Achieving the Millennium Development
Goal (MDG) 4 target requires that the under-5 mortality rate declines, on
average, by 4.4 percent annually between 1990 and 2015. These regions
achieved this benchmark through 2006 or came closer to it, putting them on track
to achieve the MDG4 target (UNICEF, et al. 2007: 7).
2
A common pattern is visible in the regions mentioned above. Speculation is that
specific regional convergence has taken place of various policies and practices,
perhaps mediated through regional institutions or informal policy networks.
Murray, et al. (2007) suspect that some other factors might have been shared
across these regions, such as educational or environmental policies or the key
driver of mortality change, accumulation of stocks of household, community, and
national physical and human capital. All these put together could have driven the
mortality decline in these regions. UNICEF (2004) identified common conditions
in countries where progress has been slow. Access to clean water is low;
percentage of births not attended to by skilled personnel is high; percentage of
under-5 moderately or severely underweight is high; percentage of one year olds
who did not receive 3 doses of DPT4 is high, and; percentage of children under 6
months of age who are not exclusively breastfed is high.
Some of the developing regions have lagged behind in their overall development.
For example, in South Asia the rate at which mortality declined has been low
compared to other parts of Asia. In the period 1960-1970 under-5 mortality was
199 deaths per 1 000 live births. In the period 2000-2006 the under-5 mortality
was reported to be 83 deaths per 1 000 live births. Afghanistan is one country in
this region where under-5 mortality is still very high, at 257 deaths per 1000 live
births in 2006. Poor perinatal care is the leading reason for children under five
dying, accounting for almost one third of all their deaths. Acute respiratory
infections and diarrhoea are the other main killers. UNICEF (2004) identifies
diarrhoea as a single proximate cause of child deaths to be at its worst in the
South Asia region.
In another developing region, Middle East and North Africa, under-5 mortality
rate in 1960 was 248 deaths per 1 000 live births. During the period 1990-2000
the under-5 mortality had declined to 55 deaths per 1000 live births, it then
4
DPT is Immunization to protect against Diphtheria, Pertussis (whooping cough), and Tetanus.
3
further declined to 46 deaths per 1 000 live births in 2006. The 2006 figures show
that Djibouti and Yemen still have high under-5 mortality of 130 deaths per 1 000
live births and 100 deaths per 1 000 live births respectively. UNICEF (2004)
points out that most countries in this region have made substantial progress in
providing services to their populations through:
•
reducing the levels of malnutrition to below 10%;
•
increasing the coverage of water and sanitation to above 80% ;
•
increasing the immunization coverage to 90% of children with 3 doses of
DPT and more than 80% of children vaccinated against measles, and;
•
providing antenatal care during pregnancy and skilled attendants at
delivery
Finally, among the developing regions, mortality in Sub-Saharan Africa has
remained notoriously high. UNICEF (2004) approximates that 42% of children
who die before they are five live in Sub-Saharan Africa. In 1960 the under-5
mortality was 277 deaths per 1 000 live births and a decade later it had declined
to only 243 deaths per 1 000 live births. In the period 1970-1980 the under-5
mortality declined to 200 deaths per 1 000 live births. The pace of decline slowed
down in the following decade when under-5 mortality rate declined to 187 deaths
per 1 000 live births in 1990. During the period 1990-2000, the under-5 mortality
declined further to 170 deaths per 1 000 live births. However, West and Central
Africa showed a higher rate of 193 per 1 000 live births in the same period while
Eastern and Southern Africa reported 145 deaths per 1 000 live births. The 2006
figure indicates that the under-5 mortality rate was 160 deaths per 1 000 live
births for the region, however in West and Central Africa it was 186 deaths per
1 000 live births.
In the South African Development Community (SADC) countries, Angola and
Democratic Republic of Congo reported the highest under-5 mortality in 2006, at
260 deaths per 1 000 live births and 205 deaths per 1 000 live births respectively.
This was followed by Zambia and Swaziland which reported 182 deaths per
4
1 000 live births and 164 deaths per 1 000 live births respectively. With the
exception of Mauritius, Seychelles and South Africa, all other remaining SADC
countries reported under-5 mortality levels between 105 and 140 deaths per 1
000 live births in 2006. Small islands like The Seychelles and Mauritius have the
lowest levels at 13 deaths per 1 000 live births and 14 deaths per 1 000 live
births respectively.
There was a decline in under-5 mortality in all these
countries until 1990. However, UNICEF (2004) points out that in some countries
including South Africa mortality increased or stagnated from 1990.
Lopez (2000) warned that although there has been good progress in delivering
interventions to more and more children, failure to maintain service delivery and
expand it to control new threats like HIV/AIDS, could well see these gains
stagnate or unthinkably, decline. South Africa is in a similar situation to that
described by Lopez.
1.2 Outlining the problem
By 1980 the under-5 mortality rate in South Africa was 91 deaths per 1 000 live
births, this declined in the following decade to 60 deaths per 1 000 live births.
During the period 1990-2000 the level of under-5 mortality took a turn in an
upward direction, it increased slightly to 63 deaths per 1 000 live births. This
trend continued and the 2006 figures indicate an increase to 69 deaths per 1 000
live births. Indeed this is of great concern because these levels are relatively
high. Omar, et al. (2000) also argued that the rapid rate of decline observed
earlier was not sustainable, given the slow rate of economic development, the
impact of the AIDS epidemic, and the infusion of a very narrowly defined set of
sophisticated technology-driven public health intervention.
Nannan, et al. (2000) believe that infant and child mortality stopped declining in
South Africa in about 1992 and they believe this could be attributed to paediatric
AIDS. Garenne and Gakusi (2005) also confirm in their reconstruction of mortality
5
trends using South African Demographic and Health Survey (SADHS)5 of 1998
that under-five mortality in South Africa declined until 1992. However, mortality
started to increase rapidly after 1993. They also claim that the increase was
almost entirely due to HIV/AIDS.
However, they also point out that after
discounting for paediatric AIDS mortality, under-five death rates were stationary
in South Africa, a finding that suggests that earlier progress in treating other early
childhood causes of death were not sustained.
1.3 Mortality in young children and its measurements
The mortality risks faced by children vary with age. Deaths in certain age groups
usually have practical programme and policy implications. Mortality among the
young can be subdivided and categorized by their age at death. These can be
measured using administrative/registration data or censuses and surveys. One of
the major problems in use of registration data when estimating mortality is the
completeness of registration. Surveys collect birth histories from women and
estimate child mortality through direct techniques. Women are asked about each
of the live births and whether the child is still alive or not. Child mortality is then
directly measured.
Censuses usually use indirect methods where women are asked about the
number of live births they have ever had and the number of children that are
currently alive. These methods are based on the assumption that women’s
mortality is not correlated with their children’s mortality; otherwise there will be a
bias in the child mortality. This is because; if the mother dies then she will not be
captured in the household survey to report on the child’s death.
5
The most recent SADHS conducted in 2003 has not been released. It would have shed more light into the most recent
under-5 mortality rates in South Africa.
6
1.3.1 Neonatal mortality
Neonatal mortality includes deaths that occur during the first 28 days of life
(UNICEF, et al., 2007:9). The neonatal period begins with birth and ends 28
complete days after birth. Neonatal deaths may be subdivided into early neonatal
deaths, occurring during the first seven days of life (0-6 days) and late neonatal
deaths, occurring after the seventh day but before the 28th day of life. The WHO
(2006) shows that, neonatal deaths in developed countries are declining and this
is as a result of changing patterns in reproductive health, socioeconomic
progress and improved quality of obstetric and neonatal facilities. On the other
hand no good historical data on neonatal mortality are available for developing
countries.
Causes and determinants of neonatal deaths differ from those causing and
contributing to post neonatal and child deaths. Furthermore, WHO (2006)
suggests that neonatal deaths and stillbirths stem from poor maternal health,
inadequate care during pregnancy, inappropriate management of complications
during pregnancy and delivery, poor hygiene during delivery and the first critical
hours after birth, and lack of newborn care. The report further points out that
some babies die after birth because they are severely malformed, are born very
prematurely, suffer from obstetric complications before or during birth, have
difficulty adapting to extrauterine life, or because of harmful practices after birth
that lead to infections.
1.3.2 Post-neonatal mortality
Post-neonatal mortality includes death that occurs at ages 1 to 11 months
(UNICEF, et al., 2007:9). Post-neonatal mortality is most often caused by
infectious diseases, such as pneumonia, tetanus, and malaria. An important
factor in reducing post-neonatal mortality is adequate nutrition, particularly breast
milk, which provides babies with both the nourishment and the antibodies to fight
7
infectious diseases. Breast milk can be supplemented or substituted by mixing
formula; however, it is important that clean water is used.
The issue of HIV-infected mothers' breast-milk has become controversial. A
number of countries have instituted policies that recommend that mothers with
HIV (human immunodeficiency virus) should not breast-feed, based on some
evidence of mother-to-child transmission of HIV through breast-feeding. In
contrast there are policies that promote breast-feeding in areas with high HIV
prevalence. Because breast-feeding protects against the infectious diseases that
take the lives of millions of infants every year, there is a policy debate about the
best course of action to take. Researchers do not know if the protection against
infectious diseases afforded by breast-feeding outweighs the risks of HIV
transmission to children, so it is not possible to make a definitive conclusion
about the risks and benefits of breast-feeding by mothers with HIV. However,
Brahmbhatt and Gray (2000) suggest that the breast-fed babies of mothers with
HIV had six times the protection against diarrheal deaths in the first few months
of life than babies who were not breast-fed. In the second half-year of life,
protection against both diarrheal and acute respiratory infections was about
double that for non-breast-fed babies.
1.3.3 Infant mortality and child mortality
Infant mortality is defined as the death of a live born infant between birth and
exact age 1 (UNICEF, et al. 2007:9). Infant mortality rate is the probability of a
child born in a specific year or period dying before reaching the age of one, if
subjected to current age – specific mortality rates of that period.
Infant mortality is a potentially important indicator. This is because mortality tends
to decline more slowly among infants than among children aged 1 to 4. Child
mortality includes deaths that occur at ages 1 to 4 years.
8
1.3.4 Under-5 mortality
Under-5 mortality includes deaths that occur between birth and exact age 5
(UNICEF, et al. 2007:9). Generally all deaths in childhood occur before age 5,
thus the probability of dying by age 5 can be regarded as a good index of overall
level of child mortality.
1.4 Objective of the study
The objective of this study is to investigate the trends in relative contribution of
each factor to the chance of a child’s death over time. The study will investigate
whether the association of a specific factor to under-5 mortality persist over time.
1.5 Assumptions
The following assumptions were made.
•
The two surveys whose data was used, were conducted according to
acceptable standards, that proper procedures were followed and
interviewers were well trained in data collection.
•
The South African government is interested in monitoring the progress
toward achieving the millennium development goal number 4.
•
Short recall period may have advantage of providing better data quality,
thus the decision to study only births that occurred in the five years
preceding the survey.
•
If any child in a household went hungry in the last 12 months because
there was no food then it is an indication of possible nutrient deficiency for
both children and adults in that household.
9
1.6 Limitations
•
The study did not isolate deaths due to HIV/AIDS because of lack of data.
However, it was assumed that after controlling for all possible proximate
variables the unexplained deaths might be due to HIV/AIDS.
•
The study only considered under-5 mortality as a group and no further
disaggregation was done because the sample size is small to
disaggregate to lower levels.
•
The information on child survival was obtained from mothers. This
technique has a potential selection bias, because in order for a child to be
reported the mother must be a member of the study population at the time
of the survey. Thus, either death or emigration of the mother can affect the
reporting coverage.
•
Incorrect dating of the births can distort the data, particularly if the errors
vary with the survival status of the child.
1.7 Rationale of the study
The government has invested a lot in providing water, sanitation, housing,
electricity and education to the previously disadvantaged population. Since 1994,
life circumstances of South Africans have been improving. For example the
proportion of households living in formal dwellings increased from 65,8% in 1995
to 73,8% in 2002. Proportion of households with access to clean water increased
from 78,5% to 84,4% while access to electricity for lighting increased from 63,
5% to 76, 3% and an improvement was observed in sanitation as well over the
same period.
In light of this, one would expect under-5 mortality to have
declined.
The government would like to measure the impact of these improvements on
under-5 mortality.
However, this can only be done if the government
understands the impact of each service with time. Some services could be
having a better effect than the others or the influence of one service could have
10
shifted with time. With limited resources, not all services can be provided at the
same time. There could be a need to evaluate existing policies and change them.
1.8 Review of chapters
Chapter 2 of this study presents the conceptual framework for the analysis which
was used for this study. Subsequently, the chapter focuses on the review of the
relevant literature especially linking it to the key determinants of under-5 mortality
as described in the analytical frame work. Chapter 3 gives the background of the
secondary data which was used for the analysis. It presents how the samples for
the two surveys were drawn. It also provides the methods and procedures which
were followed during data collection. The chapter also presents quality
assurance initiatives which were put in place during the training of fieldworkers
and also during data collection. Furthermore, the chapter points out data quality
issues which were identified during analysis. It also discusses the analysis
strategy and the rationale including the operational definitions of the variables
used in the analysis. The chapter concludes by pointing out the omitted
traditional explanatory variables which are not included in the study. Chapter 4
presents the results from the univariate, bivariate and multivariate analysis.
Finally chapter 5 presents the key summary and subsequent discussions as well
as the policy implications.
11
CHAPTER 2
2. CONCEPTUAL FRAMEWORK AND REVIEW OF LITERATURE
2.1 Introduction
The purpose of this chapter is to present the conceptual framework, including the
review of literature used to explain socioeconomic, environmental, biological and
health determinants of under-5 mortality. This framework formed the basis of the
analysis in this study.
2.2 Conceptual framework for the analysis of childhood mortality
According to Mosley and Chen (1984), all social and economic determinants of
child mortality necessarily operate through a common set of biological
mechanisms, or proximate determinants, to exert an impact on mortality. In this
framework, a set of proximate determinants or intermediate variables that directly
influence the risk of morbidity and mortality are identified. All social and economic
determinants must operate through these variables to affect child survival. This
study adopted the Mosley and Chen (1984) approach to the analysis of child
mortality.
Figure 2.1 illustrates the path to a healthy child or a sick child and eventual
death. The socioeconomic factors operate through maternal, biological
environmental, nutritional and health seeking behaviour factors to leading to a
healthy child or sick child. However, with modern medical intervention (through
prevention or treatment), a child may remain healthy, the sick child could recover
and become healthy or treatment may fail and the child dies. Each of the factors
is discussed below.
12
Figure 2.1: Analytical Framework for child survival
Social Economic determinants
•
•
•
Biological and
Maternal Factors
•
•
•
•
Age of the mother
Birth interval
Birth order
Sex of the child
Healthy
Child
Mothers Education
Place of residence
Labour Market Status of the mother
Environmental Health
Factors
•
•
•
•
Access to sanitation
Source of water
Source of energy
Type of dwelling
Nutrient Deficiency
•
Children going hungry
because there is no food
Sick Child
Death
Adapted from Mosley and Chen (1984)
13
Health seeking
behaviour
•
Place of delivery
2.3 Socioeconomic determinants of child survival
The relationship between socioeconomic factors and childhood mortality has
been well established by several studies, namely: Cleland (1990), Hobcraft et al.
(1984), Hobcraft (1993) and Machado and Hill (2005). The framework adopted
from Mosley and Chen (1984) in this study uses mother’s education, type/place
of residence, and labour market status of the mother as socioeconomic factors
which might influence child survival. These factors will be discussed in more
detail below.
2.3.1 Mother’s education
Mother’s education level can affect child survival by influencing her choices and
increasing/limiting her skills in health care practices related to contraception,
nutrition, hygiene, preventative care and disease treatment. On the other hand
the educational level of the father usually correlates strongly with occupation and
therefore with household income. In many cases correlation between the health
effect and the educational level of the father or other non-childbearing,
economically productive adult members in a household largely occur because of
operations on the proximate determinants through the income effect (Mosley and
Chen 1984:34).
The relationship between mother’s education and child survival has received a lot
of attention and a number of studies have been conducted on this relationship. In
Hobcraft, et al. (1984) the association of mother’s education and child survival
usually survived controls of other socioeconomic variables. Furthermore,
Hobcraft, et al. (1984) suggested that there was no threshold level of maternal
education that needed to be reached before advantages in child survival began
to accrue; even a small amount of education was usually associated with
improved chances of child survival. However, some studies have shown that the
associations between mother’s education and child survival were weaker in Sub-
14
Saharan Africa than in Asia or particularly Latin America, where socioeconomic
differentials were generally higher. The reason for this kind of association is
unknown; however Hobcraft (1993) has tried to explain this association. He
suggests that perhaps health infrastructures are weaker in sub-Saharan Africa,
thereby inhibiting the ability of more educated mothers to take advantage of their
human capital in the health environment. Different researchers suggest pathways
whereby mother’s education might enhance child survival. Cleland (1990)
concluded that education may have a modest effect on health knowledge and
beliefs.
Madise, et al. (1999) in their study of several African countries found higher
levels of education i.e. secondary schooling and beyond to be important for child
health. However, Magadi (1997) suggests that father’s not mother’s education is
significantly associated with child health in Kenyan communities where the status
of women is low. Mosley and Chen (1984) also indicated that father’s education
may influence attitudes and thus preference in choice of consumption goods.
They pointed out that this effect is likely to be most significant for child survival
when a more educated father is married to a less educated mother. Mother’s
education can also be linked to other factors that shape and modify the economic
choices and health-related practices of individuals according to cultural traditions
and norms of society.
2.3.2 Place of residence
Place of residence of the mother affects the survival status and nutritional status
of the living children in developing countries. This relationship is well established
by several studies; Mahmood (2002), Sastry (2004), Nannan, et al. (2007). The
urban areas usually have better infrastructure for health services compared to
non-urban areas. They are usually more developed. Machado and Hill (2005)
showed that having a mother who lives in the highest developed community
reduced the odds of neonatal deaths. They concluded that community
infrastructure may improve hygienic practices. Furthermore, interactions between
15
friends and neighbours in the communities may lead to changes in behaviour
regarding infant care and in this case better off communities may benefit from the
overall level of community education (Machado and Hill, 2005:207). Kanaiaupuni
and Donato (1999) even suggested that paved roads and female labour force
participation were also important.
In South Africa, there has been an increase in rural-urban migration of the black
population since 1994. This was because the apartheid laws which restricted the
movement of the black population were abolished. However, a substantially large
number of people move to informal settlements next to big cities. The informal
settlements do not enjoy similar infrastructure as other formal urban areas. A
child living in an informal settlement has totally different living conditions
compared to the one living in a formal area. Both could be classified as living in
an urban area or in the same city, which will give biased results. Sastry (2004)
concluded that in Sao Paulo, children from disadvantaged families were worse
off in urban areas because the deleterious effects of being disadvantaged were
much larger in urban areas than they were in rural areas.
Amouzou and Hill (2004) conclude that the weak effect of urbanization they
observed could be due to the rapid increase of urban poverty in such a way that
urban poor are losing their health advantages compared to rural residents. This
is likely to be the case in South Africa. Thomas (2007), in his study of mortality
differentials in South Africa by migration status suggests that native-born internal
migrants had a steeper socioeconomic gradient in child mortality than native-born
non-migrants.
2.3.3 Labour market status of the mother
Labour market or work status of the mother is likely to affect child survival in both
directions. The need to work, especially outside the home, may affect survival
chances directly, simply by preventing the mother from caring for the infant. This
may have substantial effects through lack of proper feeding and particularly
16
breastfeeding early in life (Hobcraft, et al. 1984). However, a working mother can
also be associated with high family income which can increase a child’s survival.
Ibrahim, et al. (1994) observed that non-farming mothers in a household with
fewer children were more active than farming mothers in using oral rehydration
therapy (ORT). They concluded that mothers who had more time to give to child
care were more likely to use ORT.
Kishor and Parasuraman (1998) found that mother’s employment had a negative
effect on the child survival, if the mother works away from home for cash, lives in
urban area, or lives in the South of India. In my view this could mean that it is the
mother’s absence rather than employment status which affects the child. If the
mother works, but commutes from where the child resides, the outcome might be
different. Short, et al. (2002) identified that both work compatibility and work
intensity reduce women's involvement in child care in China. However, they also
pointed out that, if women with intensive work demands provide less child care,
this does not necessarily hinder children's physical and psychological
development. This is because in China, relatives or other members of the
household assist in child care. Child care is not exclusively left to the mother.
Alternative child caregivers such as grandmothers can reduce a mother's burden
greatly.
In India, Krishnaji (1995) showed that working mothers experience a greater child
loss than non-working mothers in respect of both male and female children.
Generally, a narrower gender differential in child mortality among working
mothers was observed in most of the states, however in the north and the northwest, the work status of women had a greater impact on male children than on
girls. To explain the case in the north and north west, Krishnaji argued that it is
because in general there is a strong bias against girls in these states. The male
children of non-working mothers are the best protected among all categories so
that the withdrawal of this protection by working mothers - if what is observed can
be described so - has a greater impact for boys.
17
He concluded that the narrower gender differential in child mortality among
working mothers could be due to the exposure women get and thus changed
attitudes towards girls.
2.4 Biological and Maternal determinants of child survival
Mosley and Chen (1984) identified birth order, birth interval and age of the
mother as factors which influence child survival. Studies conducted by Hobcraft,
et al. (1985), Rutstein (2000) and Davanzo, et al. (2004) show the association of
these factors to child survival. In addition to the above mentioned factors, the sex
of the child was also considered in this study. Each factor is discussed below.
2.4.1 Birth order
High mortality has been associated with being the first born and with high birth
order. Hobcraft, et al. (1985), showed a clear excess of neonatal mortality for the
first births and first born children continued to be at a disadvantage during the
remainder of infancy. However, contrary to the general belief, there was no clear
evidence of excess mortality for children of birth order four to six, nor even for
those of order seven and higher, once the other factors in the regression model
were controlled. This could suggest that mortality associated with births of high
orders may be predominantly caused by other factors like birth intervals.
However, it should be noted that the outcome of the first birth could be
associated with the age of mother rather than the order. Hobcraft (1991)
concludes that delaying the first birth until a woman is at least 18 years of age
might reduce the risk of death for first born children by up to 20 percent on
average and up to 30 percent in a few countries.
Other researchers like
Mohamed, et al. (1998) linked the death of the first born to low birth weight.
18
2.4.2 Birth interval
A number of studies have demonstrated increased mortality risks among children
born after short birth intervals. Some of these studies have investigated possible
pathways through which preceding birth intervals may affect childhood survival.
Boerma and Bicego (1992) provided possible pathways through which the
relationship between preceding birth intervals and child survival might be
affected, identifying prenatal and postnatal mechanisms. As far as prenatal
mechanisms are concerned, it is believed that women with a short interval
between two pregnancies have insufficient time to restore their nutritional
reserves, which might affect foetal growth. These researchers mentioned several
studies which revealed increased risk of intrauterine growth retardation for
shorter inter-pregnancy intervals. Both intrauterine growth retardation and
prematurity lead to low birth weight, which is a strong determinant of infant
mortality.
Postnatal mechanisms include poor nutrition of the mother, which may lead to
impaired lactation and the inability to provide adequate care for the children.
Sibling competition may also have an effect on the survival of the child. The
results of Boerma and Bicego’s (1992) study suggest that prenatal factors are
more significant than postnatal factors. Hobcraft, et al. (1985) conclude that short
child spacing could be the dominant source of most of the apparent increase in
risks at high birth orders and higher ages of the mother. Children born at very
short intervals after preceding births (1 to 17 months) are about twice as likely to
die as those born after intervals of 24 to 47 months: those born after 18-23
months experience an excess risk of about one-third (Hobcraft,1991).
Davanzo, et al. (2004) summarize mechanisms that have been hypothesized to
possibly contribute to the detrimental effect of a short birth interval on childhood
survival as; (a) behavioural effect associated with competition among siblings, (b)
the inability (or lack of desire) to give a child adequate attention if his or her birth
19
came sooner than desired; and, (c) disease transmission among closely spaced
siblings. Hobcraft, et al. (1985) in their quest to answer whether child spacing
effects are real or artifactual, discussed the complex web of potential
associations between breastfeeding, mortality and subsequent pregnancy. They
concluded that the most plausible mechanism for the deleterious effect of short
previous interval is maternal depletion. This results in a small baby, perhaps with
increased risk of prematurity. Low birth weight is associated with very poor
survival chances.
Some studies showed that the effects of birth spacing disappear if women attend
prenatal care. For example Mahmood (2002), showed that for mothers with
shorter previous birth intervals who have used prenatal care, their babies are
significantly more likely to have better survival chances during the neonatal
period than those mothers with the same short birth interval who did not receive
prenatal care for the index child. This was earlier suggested by Boerma and
Bicego (1992).
2.4.3 Age of the mother
Some studies like those conducted by Hobcraft, et al. (1985), Rutstein (2000),
and Machado and Hill (2005) have shown some association between the age of
the mother at birth and child survival. Hobcraft, et al. (1985) showed that
mortality was clearly higher among children of teenage mothers. However, in
their study there was nothing to suggest increased risks for children born to
mothers at older ages, even those with mothers who were aged 35 or above
after controlling for birth spacing. Mahmood (2002) on the contrary, observed
that children of older women (30-39 years) were exposed to significantly higher
neonatal and post-neonatal mortality.
20
2.4.4 Sex of the child
A number of studies have shown mortality differential by sex. Male mortality
usually exceeds female mortality in the neonatal period, but this differential is
reversed in the post-neonatal period. Higher female than male mortality
continued through childhood and this is supported in studies by Chen, et al.
(1981), Bhuiya and Streatfield (1991) and Arokiasamy (2002).
Chen, et al. (1981) point out that the reversal of the sex differential of mortality,
markedly so during childhood and persisting through adolescence, was
postulated to be reflective of sex-biased health and nutrition-related behaviour
favouring male children.
Furthermore, they conclude that son preference in
parental care, intra family food distribution, feeding practices, and utilization of
health services are some of the behavioural mechanisms by which sex-biased
attitudes may have led to the observed mortality pattern.
Son preference is most prevalent in East Asia, South Asia, Middle East and
North Africa. Hesketh and Xing (2006) point out that son preference is manifest
prenatally, through sex determination and sex selective abortion, and post-natally
through neglect and abandonment of female children, which leads to higher
female mortality.
One would expect mother’s education to intervene in sex discrimination.
However, Bhuiya and Streatfield (1991) showed that the positive effect of
mother’s education on child survival is not similar for boys and girls in
Bangladesh. They showed that for boys a change in mother’s education from no
schooling to 1-5 years of schooling resulted in a reduction in the predicted risk of
45 percent, while for girls it was only 7 percent. Furthermore, a change from no
schooling to 6 or more years reduced the risk of dying by 70 percent for boys and
by only 32 percent for girls. However, Eswaran (2002) concluded that the
empowerment of women, which increases the bargaining power of wives relative
21
to their husbands, results in a decline in fertility and in the mortality rate of
children.
Although most studies show discrimination bias towards girls, Pande (2003)
identified sex composition of siblings as a factor in selective discriminatory
practices that affect the health of surviving children. He identified that in rural
India all girls do not face the same level of discrimination; the first girl born after
two or more boys may face less discrimination than a boy who has two or more
older brothers. On the other hand, girls who were born into a family that already
has two or more surviving daughters and no surviving sons are among the most
likely to be severely stunted (38%) and are less likely to be immunized than are
first daughters.
2.5 Environmental health determinants of child survival
Environmental conditions have long been considered to have a significant
influence on mortality. These include access to sanitation, source of drinking
water, source of energy and type of dwelling. Some of these factors are so
interlinked that they will be discussed together rather than individually. For
example Ezzati and Kammen (2002) argued that to understand the health effects
of exposure to indoor smoke so that appropriate interventions and policies can
be designed and implemented is a complex phenomenon. You have to isolate
factors which determine human exposure, and their relative contributions of each
factor to personal exposure. These factors include energy technology (stove-fuel
combination), housing characteristics (e.g., the size of the house and the material
it is built from, the number of windows, and the arrangement of rooms), and
behavioural factors (e.g., the amount of time spent indoors or near the cooking
area).
22
Studies conducted by Anderson, et al. (2002) and Wichmann and Voyi (2006)
have shown a strong association with access to clean water, sanitation, clean
source of energy and with infant and child mortality.
The South African Demographic and Health Survey (SADHS) report of 1998
showed
childhood
mortality
differentials
caused
by
socio-economic,
demographic, environmental and high-risk fertility behaviour. For environmental
factors, source of drinking water, sanitation, housing materials and source of
energy were investigated. Child mortality rates, more than doubled where the
source of drinking water was other than piped water. Where poor sanitation
existed child mortality rates are higher. The report also showed that there was a
relationship between material used for the dwelling and source of energy with
child mortality. Child mortality increased more than three times where other
materials other than block/bricks are used for housing and also other sources of
energy other than electricity were being used.
2.5.1 Source of water and access to sanitation
Increased risk of potentially fatal diarrhoeal diseases is expected among
households with no clean drinking water and/or with no safe sanitation. Some
studies like Mahmood (2002) have shown a relationship between access to clean
water and sanitation to under-5 mortality. Anderson, et al. (2002) in their study of
black and coloured populations showed a hierarchy of needs in which without
clean water, sanitation matters little. In their analysis they considered household
social economic characteristic, access to and use of health care, environmental
conditions and age of the mother.
However, the 1998 SADHS report showed that children born after a very short
interval suffer a significantly higher mortality. The study by Anderson, et al.
(2002) never took birth spacing into account when actually 5% of children born in
the five years preceding the demographic and health survey fell in this category.
This study included birth spacing as a control variable. Mahmood (2002) also
23
found that families living in households with piped water connected in their
houses have a significantly lower post neonatal mortality than those families
which depend on wells for drinking water. However, the results did not show
evidence of improved child survival in households that had flush toilets compared
to those that did not have.
2.5.2 Source of energy
Cooking and heating with solid fuels on open fires or traditional stoves in poorly
ventilated indoor environments leads to health hazards. Wichmann and Voyi
(2006) suggested that exposure to cooking and heating smoke from polluting
fuels is significantly associated with 1-59 months mortality in South Africa, after
controlling for mother’s age at birth, water source, asset index and household
overcrowding. As mentioned earlier the 1998 SADHS report showed that there
was a relationship between sources of energy and child mortality.
Indoor pollution affects children more than it affects adults. Fitzgerald, et al.
(1998) explain why children are more vulnerable than adults. They argue that
infants and young children have much greater surface-area to volume ratios than
adults, thereby increasing the potential exposure through the skin. Infants and
young children engage in oral exploratory behaviour and often play on the
ground, thereby increasing potential ingestion of contaminants in soil and dust.
Exposure through respiration may be increased because infants and children
inhale air closer to the ground than adults do, increasing the potential intake of
contaminants from the soil and dust. In addition, children are also more exposed
to dietary sources of pollution.
24
2.5.3 Type of dwelling
A relationship between type of dwelling and child mortality has been established
in a number of studies, namely: Anderson et al. 2002; Jacobs et al. 2009; and
Shehzad 2006. This is to be expected: brick houses are likely to be more
hygienic than those built from informal material or scrap, as is often the case in
informal settlements in South Africa. A house that is small and inadequately
ventilated will have an adverse effect on a child’s health. The situation becomes
even worse where there is overcrowding: children become more prone to
communicable diseases. Shehzad (2006) found that, in Pakistan, child illnesses
such as diarrhoea, acute respiratory infections and fever are affected by family
size, housing and parental education.
2.6 Nutrient deficiency as a determinant of child survival
This proximate determinant relates to intake of the three major classes of
nutrients calories, protein and micronutrients. Mosley and Chen (1984) pointed
out that the survival of children is influenced by nutrients available not only to the
child but also to the mother. Nutrient availability to the infant or to the mother
during pregnancy and lactation can be measured directly by the weighing of all
foods before consumption, accompanied by the biochemical analysis of food
samples. The three indicators of nutritional status are stunting, which indicates
chronic under nutrition in children, wasting which indicates acute under-nutrition,
and finally the proportion of children who are under weight.
According to
Bomela (1999) stunting or chronic malnutrition is the most prevalent form of
malnutrition amongst the under-5 in South Africa. Malnutrition is one of the
important risk factors for mortality due to acute respiratory infections.
25
2.7 Healthy seeking behaviour as a determinant of child survival
Unlike other determinants which affect the rate at which children move from
health to sickness, this group influences this rate (through prevention) and rate of
recovery (through treatment), (Hill, 2003:139). For preventive measures this
variable is commonly assessed by reported use of such preventive services as
immunization, malaria prophylaxis, or antenatal care. For curative measures the
providers of care and types of therapy taken for specific conditions are assessed
(Mosley and Chen, 1984:33).
Rutstein (2000), in his comparison of DHS data from 62 developing countries,
showed that increases in the percentage of births that received medical care at
delivery were associated with decreasing mortality during the first year of life. An
increase in prenatal care was associated with decreases in mortality among
those under-5 years as well. Boerma and Bicego (1992) even linked prenatal
care and birth intervals, in that they hypothesised that unlike pregnant women
with short birth interval, pregnant women with longer birth intervals are more
likely to attend prenatal care services which ultimately results in a healthy child
birth.
Rutstein (2000), pointed out that an increase in the percentage of children
vaccinated against measles was associated with a decline in infant mortality and
with mortality at ages > 1. He went further to show that increases in the
percentage of children receiving medical attention for diarrhoea; acute respiratory
illness and fever were associated with the declines in mortality.
26
2.8 Summary
This chapter proposed a conceptual framework for use in the analysis and
reviewed various studies dealing with under-5 mortality rates. The conceptual
framework considers socioeconomic factors, environmental, biological and
maternal factors, nutrient deficiency factors and health-seeking behaviour
factors. A review of the literature dealing with each of the proxy indicators for the
above factors was conducted.
.
27
CHAPTER 3
3. METHODOLOGY
3.1 Introduction
This chapter will focus on the background of the data, the sample and the
research instruments used in the study. Data quality issues are also discussed.
The chapter also presents the strategy used in the analysis of the data, including
the operational definitions of the independent and dependent variables.
3.2 Background into the data
Secondary data from two national household surveys below was used.
a) October Household Survey 1997 (OHS 1997)
b) General Household Survey 2002 (GHS, 2002)
While challenging to mount, well-designed and well-implemented household
surveys produce high quality data on mortality levels and trends (UNICEF, et al.,
2007:33). The strength of household surveys is that they collect birth histories
and also data on socio-economic status, health and education. Thus the use of
household based survey data for this study.
Both surveys were annual national surveys and were conducted by Statistics
South Africa.
The October Household Survey was conducted annually from
1994 to 1999. However, this study used the 1997 round because the study is
focusing on trends in under-5 therefore it was important to select a cohort of
under-5 children which was before the last cohort which could be obtained from
the GHS 2002 i.e 1998-2002 cohort.
28
The General Household Survey replaced the October Household survey and it
started in the year 2000. It is conducted annually and for this study the 2002
round was used. Birth histories were last collected in GHS 2002. That is why this
dataset and not the later ones have been used.
3.3 The sample
Both surveys had a complex multi stage sample. The target population for both
surveys was private households in South Africa. The data base of Enumeration
Areas (EAs), as established during the demarcation phase of Census ’96,
constituted the sampling frame for both surveys. Special dwellings such as
prisons, hospitals, boarding houses, hotels, guesthouses (whether catering or
self-catering), schools and churches were excluded from the sample. The sample
for each survey is described below.
3.3.1 October household survey 1997 sample
The sampling procedure involved explicit stratification by province and
transitional metropolitan and district councils. The smaller provinces were given a
disproportionately larger number of Enumeration Areas (EAs) than the bigger
provinces.
Altogether, 3 000 EAs were drawn by means of probability
proportional to size principles in each stratum and a systematic sample of 10
households was selected in each EA. This means that 3 000 EAs were identified
as primary sampling units, and approximately 30 000 households were visited as
ultimate sampling units.
3.3.2 General household survey 2002 sample
During the 2002 GHS the sample was improved. Some small EAs were pooled
together to form a primary sampling unit (PSU). A PSU is either one EA or
several EAs depending on the number of dwellings in an EA. When the number
of dwelling units in the base or originally selected EA was found to have less than
29
100 dwelling units, this EA was combined with an adjacent EA to form a PSU.
Explicit stratification of the PSUs was done by province and area type
(urban/rural). Within each explicit stratum, the PSUs were implicitly stratified by
District Council, Magisterial District and, within the magisterial district, by average
household income (for formal urban areas and hostels).
Altogether 3 000
primary sampling units (PSUs) were included in the sample.
The allocated number of PSUs was systematically selected with “probability
proportional to size” in each stratum. Once the PSUs included in the sample
were known, their boundaries had to be identified on the ground. After boundary
identification, the next stage was to accurately list all the dwelling units in the
PSUs.
The second stage of the sample involved selecting a systematic sample of 10
dwellings from each PSU. As a result, approximately 30 000 households (units)
were interviewed.
3.4 Questionnaires
Both questionnaires went through rigorous tests and consultations. The OHS
1997 questionnaire was an improved version of OHS 1996. There was a review
panel consisting of different stakeholders under the leadership of Statistics South
Africa (StatsSA) known at the time as Central Statistical Service (CSS). The
questions were reviewed and some were dropped or improved on, some where
retained and some new questions were introduced. The new questionnaire was
tested through the behind the glass test6. This was to test if there were any
questions which were not clear or which might offend the respondent. This also
tested the flow of the interview and the length.
6
During the behind the glass test, the interviewer interviews the respondent, while the questionnaire design experts are
observing and listening through a one-way mirror.
30
After the behind the glass test the questionnaire was modified depending on
what was observed. The second test was in the field. A mini pilot study was
organised to specifically test the questionnaire. Trained field workers were
deployed in the field to test the questionnaire and report back on any problems
they encountered during the test interview. After the pilot the questionnaire was
finalized.
During the development of the GHS 2002 questionnaire, a similar process was
followed. However, during this process the lay out of the questionnaire changed
to cater for changes in data processing.
The full GHS 2002 questionnaire is attached as Appendix 1. However, since the
two questionnaires are almost similar only extracts of the OHS 1997 are included
as Appendix 2.
3.5 Training, field work operations and procedures
Field work was conducted in October 1997 and June/July in 2002. The
questionnaire was administered through face-to-face interviews. Both surveys
were conducted using almost similar procedures in the field, except for a few
which will be pointed out below.
Statistics South Africa had a national office, 9 provincial offices and a few
regional offices. The national office is responsible for the planning and
development of the survey instruments, while the provincial and regional offices
are responsible for operations in the field.
The field staff comprised of about 600 interviewers and 150 supervisors for each
survey. Each field supervisor was responsible for four interviewers. Training of
field workers was conducted in a cascaded manner. Trainers were trained at the
national level and these were Statistics South Africa national and provincial office
31
permanent staff members. The training included the questionnaire, concepts and
definitions, procedures to follow in the field, role plays and administrative
procedures.
After national training of trainers, the provincial training of fieldworkers was
conducted in each province. In provinces where the sample was big and thus
more fieldworkers, training was conducted in groups of not more than 30
trainees. Their training included all aspects of the national training and also a
field trip. Fieldworkers and provincial permanent staff who speak the same
language went through the questionnaire and agreed on the translations of key
questions.
After the four days of training field workers were required to write a test and
those who were identified as weak were retrained on some aspects. During the
training process the fieldworkers were constantly evaluated and those who
showed some leadership qualities and also passed the test well were appointed
as supervisors and they were given further training on the administrative and
supervisory tasks they were required to perform. These included identification of
the EA or the PSU boundaries, identification of the selected dwelling unit and
allocation of work to respective fieldworkers.
A vehicle was allocated to each supervisor and his/her team of four fieldworkers.
Each team was allocated between 18 and 20 EAs/PSUs to be enumerated over
a period of two weeks. The team would follow the description of the location of
the PSU, in case of the 1997 OHS the descriptions were done during the census
1996 demarcation process, while in the GHS 2002 the census 1996 descriptions
were confirmed and at times modified during the listing process as mentioned
earlier under the sample.
After the PSU was identified and boundaries established, a full count of private
dwellings was conducted and a systematic sample of 10 was selected. However,
32
this procedure was improved on during the 2002 GHS. Since there was prior
listing of all dwellings in a PSU, the sample was drawn in a more controlled
manner in the office and the fieldworkers did not have to do any sampling in the
field. They were given addresses of the selected dwelling or a description of the
house in cases where there was no address. Their role was to establish the
number of households at the selected dwelling unit. All households at a selected
dwelling were enumerated during the 2002 GHS. In the 1997 OHS multiple
households were handled differently in that if more than one household was
identified at a selected dwelling unit, only one was enumerated and was selected
randomly using probability proportional to size. A household with more people
was given a higher probability of being selected.
Each fieldworker was allocated a household to interview and face to face
interview was conducted. If the occupants of a selected dwelling were not at
home, a fieldworker was allowed to go back at least three times before he
declares a dwelling a non-contact7.
3.6 Data quality
A number of quality assurance procedures were implemented during the survey.
Coupled with the rigorous training described earlier, fieldworkers were instructed
to interview only adults in the household. When it came to birth histories which
mostly form a bigger part of this study, fieldworkers were instructed to interview
mothers because they usually have the correct information about their children.
Another quality measure which was performed was to take a sample of 5% of the
selected dwelling which were revisited by the assistant survey managers. They
conducted a control interview to confirm whether a fieldworker had visited that
household.
7
Non contact includes households which refuse to participate in the survey and households which are occupied but the
occupants are not always at home during the interviewers visit.
33
Although much was put into limiting both sampling and non-sampling errors, the
following inconsistencies were observed during the analysis.
•
In the 1997 OHS, the file which contains birth histories contained 104 877
records; however 36 records were duplicates so they were deleted.
•
About 121 records in the birth histories file could not be linked to the
person file which contained the demographic information of the mother so
they were deleted.
•
In the 2002 GHS, 11 records in the birth histories’ file could not be linked
to the person file which contains the demographic information of the
mother so they were deleted.
Deleting of records could cause a shift in birth orders and birth intervals.
However, these were relatively few to cause a significant change in the results.
3.7 Analysis strategy
Since the purpose of the study was to assess trends in the factors associated
with under-5 mortality, the approach adopted in this study was to first analyse
trends of each variable which might influence under-5 mortality between the two
points i.e 1997 and 2002. Univariate and bivariate analysis was conducted on
the variables identified in the analytical framework.
3.7.1 Logistic regression
Logistic regression estimates the odds of a certain event occurring. In this study
it was used to predict under-5 deaths. Logistic regression can be used to predict
whether an event will occur or not using a set of independent predictor variables.
Furthermore, it can be used to explain the percent of variance in the dependent
variable which is explained by a specific predictor variable.
34
This is usually explained in terms of an odds ratio. The logistic equation may be
written as follows;
eα + β1 x1 + β1 x1 +...+ βi xi
π ( x) =
1 + eα + β1 x1 + β1 x1 +...+ βi xi
Where π (x) is the probability that the response y=1
α is the equation constant and
β i is the coefficient of the predictor xi
The advantage of a logistic regression model is that the independent variables
don't have to be normally distributed. Secondly, it does not assume a linear
relationship between the independent and dependent variables. However,
logistic regression is sensitive to high correlations among the predictor
variables. This is referred to as multicollinearity. Pallant (2005) recommends
that multicollinearity problems should be checked before logistic regression
analysis. This was tested and the results are presented in Tables 3.1 and 3.2
Correlation results for both data sets indicate that, although there some
significant relationships between some of the variables none of them is very
high (i.e more than 0.7) to suggest multicollinearity. The highest registered
correlation between independent variables was between place of residence
and source of water for both 1997 and 2002 data points. The correlation
coefficients between these two predictor variables for the two data points were
0.487 and 0.483 respectively.
35
Table 3.1: Correlation between independent variables in 1997 data
Mother's education
Place of residence
Mother's
education
1.000
Place of
residence
**
.224
Mother's
labour market
status
**
.246
1.000
Mother's labour market status
sex
Birth order
Birth Interval
.175
**
1.000
sex
-.010
Birth
order
**
.238
.014
.127
**
Birth
Interval
**
.104
Mother's
age at
birth
**
.200
.065
**
.064
**
.012
.067
**
Source of
water
**
.138
Type of
sanitation
**
.172
.487
**
.189
**
.155
**
.097
**
-.002
-.002
1.000
.014
.008
.004
.014
-.032
1.000
**
.418
**
.097
**
.075
**
.248
**
.034
**
.046
.320
1.000
Mother's age at birth
1.000
Source of water
.018
*
1.000
Type of sanitation
**
Type of dwelling
**
.167
**
.013
**
**
*
**
1.000
Access to electricity
.442
.119
.020
.246
Access to
electricity
**
.251
Type of
dwelling
**
.083
-.245
**
.011
-.013
-.026
**
.069
**
.037
**
.414
**
.289
**
1.000
Place of delivery
Place of
delivery
**
.187
.059
**
.193
**
.080
**
.103
**
-.002
-.004
.067
**
.251
**
-.030
**
.037
**
.248
**
-.023
**
.075
**
.270
**
-.143
**
.050
**
.142
**
.073
**
.105
**
.160
**
.078
**
.125
**
.191
**
.075
**
-.039
1.000
.091
1.000
Nutrition
Nutrition
**
.133
**
**
1.000
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
36
Table 3.2: Correlation between independent variable in 2002 data
Mother's education
Place of residence
Mother's
education
1.000
Place of
residence
**
.231
Mother's
labour market
status
**
.193
1.000
Mother's labour market status
sex
.121
**
1.000
sex
*
-.023
Birth
order
**
.168
-.002
.094
-.011
1.000
Birth order
**
.043
**
-.111
.003
1.000
Birth Interval
Birth
Interval
**
.038
**
-.001
Source of water
**
**
-.040
*
.019
**
1.000
Mother's age at birth
.052
**
-.023
.232
Mother's
age at
birth
**
.151
Source of
water
**
.145
.483
**
.105
**
-.022
*
.092
**
-.009
.060
**
1.000
.038
**
.273
1.000
Type of sanitation
Type of
sanitation
**
.193
.263
**
.088
**
-.012
Type of dwelling
.339
**
.108
**
.021
-.188
**
-.021
-.004
**
.020
**
.046
**
*
.030
**
.050
.024
**
1.000
Access to electricity
Type of
dwelling
**
.095
.095
.076
.321
Access to
electricity
**
.223
.080
**
.101
**
.000
.181
**
.025
*
-.007
.166
**
-.010
-.005
.069
**
**
-.021
.059
**
.051
**
.387
**
**
.402
**
1.000
.054
.129
**
.174
**
.082
**
.152
**
.162
**
.112
**
.190
**
.211
**
.088
**
-.016
-.140
1.000
Place of delivery
Place of
delivery
**
.133
**
1.000
Nutrition
Nutrition
**
.175
.106
**
1.000
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
37
Furthermore, SPSS8 package can perform collinearity diagnostics. This can
pick up problems with multicollinearity that may not be evident in a correlation
matrix (Pallant, 2005). Results of collinearity diagnostics for both data points
are presented in the table below.
Table 3.3: Collinearity statistics for OHS 1997 and GHS 2002
OHS 1997
Independent variables
Mother's education
GHS 2002
Tolerance
.809
VIF
1.237
Tolerance
.838
VIF
1.193
Place of residence
.623
1.606
.685
1.460
Mother's labour market status
.902
1.109
.923
1.084
sex
.998
1.002
.998
1.002
Birth order
.750
1.333
.789
1.268
Birth Interval
.853
1.173
.935
1.070
Mother's age at birth
.773
1.294
.868
1.152
Source of water
.691
1.447
.678
1.476
1.267
Type of sanitation
.869
1.150
.789
Access to electricity
.671
1.490
.701
1.426
Type of dwelling
.867
1.153
.893
1.119
Nutrition
.959
1.042
.926
1.079
Place of delivery
.827
1.209
.911
1.098
Tolerance and variance inflation factor (VIF) are indicators of how much of
variability of a specified independent variable is not explained by the other
independent variables in the model. If the tolerance value is very small (less
than 0.1), or the VIF (which is the inverse of tolerance) is above 10, it indicates
that the multiple correlation with other variables is high (Pallant, 2005). Results
from both tests do not suggest any major multicollinearity problems, so all
variables were retained in the multivariate analysis.
First, a bivariate logistic analysis was conducted on each predictor variable
without any extra control. This was done to explore the relationship between
each predictor variable and the under-5 mortality.
8
Statistical Package for Social Scientists
38
After the bivariate logistic analysis, a multivariate logistic analysis was done to
assess the impact of a set of predictors on the dependent variable. This was
done using a hierarchical approach. Socio-economic variables where entered
in the model first as a block. These were followed by biological and maternal
factors, environmental factors and nutrition. Finally, the health seeking factor
was also entered. This allowed the assessment of each block of variables or a
variable in terms of what they add or it adds to the prediction of the dependent
variable after the previous variable or variables have been controlled for.
3.8
Operational definition of the dependent and independent
variables
Both independent and dependent variables as identified in the framework were
derived using SAS9. Each variable is discussed below. Some variables were
recoded because some options had very few responses to get sensible analysis.
3.8.1 Dependent variable
a) Under-5 mortalityThe under-5 mortality was the dependent variable. Since this is a
dichotomous variable it was treated as such. However, 1 was allocated for a
death and 0 for the survival of child.
3.8.2 Independent variables
b) Socio-economic
Mother’s education:
Two categories were created for this dummy variable. Mothers who had
completed secondary school (matric) or tertiary education were coded 1 and
those who did not complete secondary school were coded 0 and they were the
reference group.
9
Statistical Analysis Software
39
Place of residence:
This was a dichotomous variable where urban area is the reference and nonurban is the second category.
Labour market status of the mother:
The three labour market status categories, i.e. employed, unemployed and not
economically active, were recoded into two categories: employed, which is the
reference, and not employed which included both the unemployed and the not
economically active.
c) Biological and maternal-independent variables
Sex:
Sex is dichotomous: a girl was the reference and thus coded 0 and a boy,1.
Birth order:
Birth order was classified as firstborn, second order, third order, or fourth or
higher. The reference group was the first order child.
Birth interval:
This variable was treated as dichotomous by classifying firstborn and those with
birth interval of 24 months or more in one category and those with a birth interval
of fewer than 24 months in another category. The former category was the
reference group
Mother’s age at birth:
This is mother’s age at the time of birth of the subject child. It was derived by
using the current age of mother and the child’s year of birth. Three categories
40
were created i.e. less than 18 years, 18 – 34 years, and 35 years or older. The
reference was 18-34 years. These categories were created in this manner
because below 18 and above 34 age categories are considered to be risky for
child survival.
d) Environmental- independent variables
Source of water:
This is the source of water used by the household. The thirteen categories in the
questionnaire (see Appendix) were recoded into two categories piped and other.
Piped was the reference category.
Access to sanitation:
This is the type of sanitation the household used. There were fifteen categories
in the questionnaire and these were recoded into three categories: toilet on-site,
toilet off-site and bucket/no toilet. Toilet on-site was the reference category.
Source of energy:
Source of energy was treated as a dichotomous variable by assigning code 0 to
households with access to electricity from the mains and code 1 to households
without electricity.
Type of dwelling:
This was type of main dwelling used by the household. The eleven categories in
the questionnaire were recoded into a dichotomous variable, with formal dwelling
and informal dwelling. The formal dwelling category was the reference.
e) Nutrient deficiency-independent variable
No direct measurement of nutrients was done in both surveys. However, during
the 2002 GHS 2002 household was asked if in the past 12 months, there had
41
been any child (17 years or younger) in the household who went hungry because
there wasn’t enough food. During the 1997 OHS each household was asked if in
the past year, there was ever a time when the children in the household could not
be fed. These questions were used as a proxy for nutrient deficiency.
Households which reported that they could not feed a child were assigned a code
1 otherwise other households were assigned code 0. The reference category
was households which could feed their children during the reference period.
f) Health seeking behaviour-independent variable
Place of delivery:
Mothers reported their place of delivery for each child they have ever given birth
to. This was used as a proxy for health seeking behaviour of the mother.
Children who were delivered in a hospital or clinic were assigned code 0 and
those who were delivered somewhere else, were assigned code 1. The reference
category was children who were delivered in a hospital or clinic.
3.9 Omitted Traditional Explanatory Variables
There are some variables which are known to have a strong influence on child
survival but they were omitted from this analysis. This is because they were not
collected in the surveys. These are:
a)
Weight at birth
b)
Pre-natal care
c)
Breastfeeding
d)
HIV/AIDS
However, weight at birth and breastfeeding are both correlated with nutrient
deficiency and pre-natal care is likely to be correlated to the healthy seeking
behaviour variables in the study.
42
3.9.1 Impact of AIDS on under-5 mortality
A number of studies have linked the increase in under-5 mortality to the AIDS
pandemic. According to Walker (2003), vertical transmission of HIV occurs in 32
percent of births to HIV infected mothers in countries where breastfeeding is
prevalent. Adetunji’s study (2000) suggests that about 25-30% of children born
to infected mothers become infected with HIV and almost all of them die before
they are 5 years of age in most developing countries that have high HIV
prevalence. One of the main findings in his study was that under-5 mortality rates
increased in most countries with an adult HIV prevalence of =>5% while
decreases were observed in lower prevalence countries. Dorrington and others
(2004) using the ASSA200210 model estimated a prevalence of 11% for South
Africa in 2004.
Besides the direct effects that operate through vertical or perinatal transmission,
Adetunji (2000) cited indirect ways in which adult HIV/AIDS could affect the level
of under-5 mortality. These included the death of or frequent illness of the care
giver and unexplained trauma. Adetunji concludes that while it is customary to
attribute almost all the reversals and stagnation in under-5 mortality rates,
especially in Sub-saharan Africa to HIV/AIDS epidemic, this may not be as large
as they have generally been thought to be. The assumption for this study is that
after controlling for all possible proximate variables the unexplained deaths might
be due to HIV/AIDS.
10
Actuarial Society of South Africa AIDS model of 2002
43
3.10 Summary
This chapter justified the selection of the sources of data used in the study and
provided the background to the data sources and their limitations. Two criteria
were used to select variables for inclusion in the models. The first criterion was
what other studies, as reviewed in chapter two, had revealed about their
influence on under-5 mortality rates. The second criterion was availability of the
variable in the two datasets used.
44
CHAPTER 4
4. DATA ANALYSIS
4.1 Introduction
This chapter will describe the two types of analysis that were performed. Firstly,
the descriptive results are presented and discussed, followed by the results from
the logistic regression analysis.
4.2 Descriptive analysis
4.2.1 Under-5 survival 1993-1997 and 1998-2002
Table 4.1 shows the unweighted and weighted figures from the two data sets, i.e.
October Household Survey (OHS) 1997 and the General Household Survey
(GHS) 2000. A cohort of children born in 1993 and after was considered from
the OHS 1997, and a cohort of children born in 1998 and after from the GHS
2002.
Table 4.1: Survival of children under-5 for the cohorts 1993-1997 and
1998- 2002
Unweighted
Weighted
1993-1997
1998-2002
1993-1997
1998-2002
Alive
13,162
8,555
3,894,000
4,420,000
Dead
993
244
240,000
106,000
Total
14,156
8,799
4,135,000
4,526,000
58.0
23.4
Under 5 mortality rate
The sample in the 1997 OHS yielded 14,156 children of which 993 had died by
October 1997. This converts into 4,135 million births and 240 thousand deaths
respectively after weighting the data. On the other hand, the 2002 GHS sample
45
yielded 8,799 children born in 1998 and after, of which 244 had died by July
2002. This converts into 4,526 million births and 106 thousand deaths
respectively after weighting the data.
The 1997 OHS estimate under-5 mortality for the 5 year period preceding the
survey to be 58 deaths per 1 000 live births. This is computed as the ratio of
number of children who died before the age of 5 to the total children born during
the same period multiplied by 1 000. This figure looks plausible if compared to
what was estimated from the 1998 South African Demographic and Health
Survey which was 59,4 deaths per 1 000 live births. However, the estimate from
the 2002 GHS of 23,4 deaths per 1000 live births is very low compared to what
was reported in the 2003 SADHS preliminary report, which is 58 deaths per 1
000 live births. This needs to be investigated because the figure of 23,4 deaths
per 1 000 live births does not look plausible. This suggests under reporting of
deaths in the 2002 GHS. However, the objective of the study was not to estimate
levels of mortality but rather factors associated with under-5 mortality. The
assumption was that the under reporting was random and there was no under
reporting bias.
4.2.2 Births data from auxiliary source 1993-1997 and 1998-2002
The population register is another source of data which can be used to validate
the survey results. The total number of births from the survey can be compared
with total registered births from the population register.
Table 4.2: Birth occurrences (as at end of April 2008) 1993-2002 as
recorded on the population register
1993-1997
Year
1993
1994
1995
1996
1997
Total
1998-2002
Number
934,148
952,509
930,818
959,463
947,076
4,724,014
Year
1998
1999
2000
2001
2002
Total
Source: Statistics South Africa 2008
46
Number
931,357
946,918
957,634
941,664
950,439
4,728,012
Table 4.2 above shows that a total of approximately 4,724 million births occurred
between 1993 and 1997 while OHS 1997 reported 4,135 million. A similar pattern
is observed for the period 1998-2002 when the two sources are compared. The
survey reported slightly lower figures than the population register.
The expectation is not to get similar results from these sources because both are
subject to different source errors. On one hand, data from household based
surveys like OHS and GHS is subjected to age miss-reporting, missing date of
birth especially if the mother is not the one reporting. Secondly, the data is
collected through birth history of the mother, if the mother is dead or for some
reason is missed during the survey; then the child will not be reported. This can
lead to under reporting of births. On the other hand, the administrative data from
the population register is subject to late or complete lack of registration. Age
mis-reporting can also happen in administrative data.
The survey reported lower figures than the register which could suggest either
age mis-reporting or under-reporting. Another reason could be the reference
period of the surveys. The OHS 1997 was conducted in October so the births
reported do not cover November and December of 1997. GHS 2002 was
conducted in June/July so the births reported do not cover the months of August
to December of 2002.
4.2.3 Trends in environmental factors 1997-2002
After apartheid rule ended in 1994 there was a concerted effort by the new
government to extend services like water, sanitation housing to the previously
disadvantaged populations. With improved sanitation, water sources and housing
the expectation would be some improvement in under-5 mortality.
47
Table 4.3: Trends in environmental factors 1997-2002
OHS-1997
(%)
GHS-2002
(%)
Proportion of households with piped water
81.6
85.1
Proportion of households with a toilet On-site
79.8
84.4
Proportion of households with a toilet Off-site
6.1
2.6
Proportion of households with no toilet or a bucket
14.0
13.0
Proportion of households with access to electricity
61.9
77.5
Proportion of households with formal houses
88.2
86.3
Services variables
Table 4.3 shows the proportions of households with access to environmental
services which could have an impact on child survival. The figures indicate that
there was an increasing trend in access to piped water, access to electricity and
also access to toilets on-site. However, much as there was improvement in
sanitation with the middle group (i.e. those who had access to toilets but off-site),
there was virtually no improvement among the group which had no toilet or using
a bucket. The proportion of households which did not have a toilet or using a
bucket was 13.0% in 2002 compared to the 14.0% in 1997.
Access to formal houses showed some sign of stagnation during this period. The
proportion of households living in formal dwellings declined from 88.2% in 1997
to 86.3% in 2002. This could be attributed to internal rural-urban migration.
As can be seen above, the table gives a mixed picture with regards to
environmental factors. There was an improvement in some services while there
was no improvement in others. The impact of this on under-5 mortality will be
discussed later in this chapter and in chapter 5.
48
4.2.4 Trends in selected socio-economic indicators 1997-2002
Table 4.4 shows trends of selected social economic indicators by sex. The table
suggests that the proportion of South Africans living in urban areas increased
from 49.3 % in 1997 to 54.5 % in 2002. This is likely to have put pressure on
housing. This supports the observation in Table 4.3 of the decline in access to
formal dwellings. This period shows an increase of 5.6 percentage points in the
proportion of women living in urban areas.
Table 4.4: Trends in socio-economic indicators 1997-2002
Socio-economic variables
Population in urban area
Among men proportion in urban area
Among women proportion in urban area
1997
(%)
49.3
50.3
48.6
2002
(%)
54.5
54.8
54.2
Population 15-49, proportion with matric or higher
Among men 15-49 proportion with matric or higher
Among women 15-49 proportion with matric or higher
32.3
40.9
25.2
37.8
44.0
32.1
Employment population ratio 15-64
Employment population ratio Men (15-64)
Employment population ratio Women (15-64)
32.6
42.4
24.7
38.7
45.8
32.4
Employment population ratio 15-49
Employment population ratio Men (15-49)
Employment population ratio Women (15-49)
32.3
40.9
25.2
37.8
44.0
32.1
Labour Participation rate (15-64)
Labour force participation rate for men (15-64)
Labour force participation rate for women (15-64)
41.8
51.3
34.0
55.5
62.7
49.1
Labour Participation rate (15-49)
Labour force participation rate for men (15-49)
Labour force participation rate for women (15-49)
42.5
50.5
35.8
56.7
62.5
51.4
The proportion of those with matric or higher increased among women in the
child bearing age group (15-49) by 6.9 percentage points to 32.1% between 1997
and 2002. However, this is still low, compared to men in the same age category.
49
Better educated women are more likely to be better informed and this usually
transforms into reduced under-5 mortality among their children. This was
discussed in earlier chapters.
Another indicator of social economic status is employment: women who are
employed are more likely to be financially independent to be able to look after
their children. Employment population ratio is the proportion of those who are
employed to the population of the working age. It is also known as labour
absorption rate. In other words it gives the probability of getting a job.
The
employment population ratio among women of the working age group (15-64)
and among women of child bearing age (15-49) increased by 7.7 percentage
points and 6.9 percentage points respectively between 1997 and 2002.
Labour force participation is the ratio of the employed plus the unemployed to the
working age population. The table above shows that among the working age
population, the participation rate increased from 41.8% to 55.5%. The female
participation rate was still lower than the national figure in 2002. However, there
was an increase in the female participation rate by a massive 15.1 percentage
points to 49.1 between 1997 and 2002. Almost a similar increase was observed
among women of child-bearing age during the same period. This could indicate
improvement in women empowerment.
Improvements among women were observed in almost all socio-economic
indicators discussed above during the period 1997 to 2008. Improvements in
socio-economic status of women are expected to bear fruit in terms of under-5
mortality.
4.2.5 Trends in biological and maternal factors (1997-2002)
Trends in biological and maternal factors are presented in Table 4.5. The table
shows that 4.9% of children born between 1993 and 1997 belonged to mothers
who were less than 18 years old at the time of delivery. This increased to 7.1%
50
in the 1998-2002 birth cohort. This suggests an increase in the number of young
mothers.
Table 4.5: Trends in biological and maternal factor for birth cohorts
1993-1997 and 1998-2002
Variables
Proportion born by mothers age at the time of
delivery
< 18
18-34
35+
Proportion with birth interval of less than 24 months
Proportion born by birth order
1
2
3
4+
Proportions by sex
Girls
Boys
Birth cohort
1993-1997
(%)
Birth cohort
1998-2002
(%)
4.9
67.0
28.1
7.1
76.9
16.0
100.0
100.0
15.8
7.6
27.9
21.3
15.9
34.8
100.0
40.1
25.2
14.6
20.1
100.0
49.4
50.6
100.0
49.2
50.8
100.0
There was a decrease in the proportion of children born to older mothers, i.e.
those aged 35 and above at the time of delivery. This suggests a decline in
children born by the high-risk aged mothers by 12.1 percentage points.
A decline in children with a high risk birth interval is also observed in Table 4.5
above. Among the 1993-1997 birth cohort, 15.8% had a birth interval of less than
24 months compared to the 7.6% among the 1998-2002 birth cohort who had the
same birth interval.
Birth order shows a very interesting pattern. Among the 1993-1997 birth cohort
27.9% were first order children while among the 1998-2002 birth cohort the first
order children represented 40.1%. This could suggest a decrease in fertility.
51
There is no observed change in sex composition of the two birth cohorts in the
table above.
4.2.6 Trends in health seeking behaviour 1997-2002
Figure 4.1 shows a declining trend in the proportion of children who were not
delivered at a hospital or clinic. For children in the 1993-1997 birth cohort
approximately 18% were not delivered in a hospital or clinic. This proportion
declined by approximately 10 percentage points to 8% in the 1998-2002 birth
cohort as expected.
Figure 4.1: Proportion of children not delivered in hospital or clinic:
birth cohorts 1993-1997 and 1998-2002
%
20.0
18.0
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
1993-1997
1998-2002
This suggests fewer children were delivered without the help of medical
professional
among
the
1998-2002
birth
cohort.
This
improvements in health seeking behaviours among mothers.
52
could
suggest
4.2.7 Trends in nutrient deficiency factor 1997-2002
Figure 4.2 suggests a higher proportion of children among the 1998-2002 birth
cohorts who were from households which experienced food shortage compared
to the 1993-1997 birth cohort. However, these results should be interpreted with
caution because the question in GHS 2002 questionnaire was different from the
one in the OHS 1997. The change could also be due to the way the question was
phrased.
Figure 4.2: Proportion of children in households which experienced food
shortage: Birth cohorts 1993-1997 and 1998-2002
%
45.0
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
1993-1997
4.3
1998-2002
Trends in proportion of under-5 deaths and associated
factors
The dependent variable (under-5 mortality) was analysed by each independent
variable to establish the relationship between the dependent variable and each of
the independent variables for the two periods under review.
53
4.3.1 Proportion of children who died under-5 by environmental factors
Figures 4.3 - 4.6 above show the proportion of children who died before reaching
their 5th birthday by environmental factors.
Figure 4.3: Proportions of under-5 deaths by access to piped water: birth
cohorts 1993-1997 and 1998-2002
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Piped water
7.4
2.4
Unpiped
7.0
4.4
Surprisingly among the 1993-1997 birth cohort, the proportion of under-5 deaths
is slightly higher among children in households with access to piped water
compared to those without. The pattern changes among the 1998-2002 birth
cohort, the proportion of under-5 deaths by access to piped water decreased and
the proportion of under-5 deaths is lower among children in households with
access to piped water compared to those without.
54
Figure 4.4: Proportions of under-5 deaths by type of dwelling:
birth cohorts 1993-1997 and 1998-2002
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Formal
7.3
2.6
Informal
7.4
4.1
Figure 4.4 suggests that the proportion of under-5 deaths is slightly higher
among children living in informal dwellings compared to those living in formal
dwellings in the 1993-1997 birth cohort. However, the 1998-2002 cohort shows a
huge difference in the proportion of under-5 deaths between children who live in
formal and informal dwellings.
The proportion of under-5 deaths is higher in households without a toilet or using
a bucket toilet as can be seen in Figure 4.5. This is an indication of a relationship
between child survival and sanitation. A similar pattern is observed in both birth
cohorts under review, although the 1998-2002 cohort shows lower proportions
compared to the 1993-1997 birth cohort
55
Figure 4.5: Proportions of under-5 deaths by type of sanitation facilities:
birth cohorts 1993-1997 and 1998-2002
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Toilet on-site
7.0
2.3
Toilet off-site
6.8
3.1
Bucket or no toilet
9.0
4.8
.
Figure 4.6: Proportions of under-5 deaths by access to electricity: birth
cohorts 1993-1997 and 1998-2002
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Connected to the mains
6.4
2.2
Not connected
8.3
4.3
56
There also seems to be a relationship between access to clean energy, i.e.
electricity and child survival. Figure 4.6 above, show that 6.4 percent of children
from households which had access to electricity died before the age of 5 as
compared to 8.3 percent of those from households without electricity among the
1993-1997 birth cohort and 2.2 percent and 4.3 percent respectively among the
1998-2002 birth cohort.
All cases show very drastic declines between the two birth cohorts. For example,
the proportion of under-5 deaths among children in households with no piped
water declined from 7.0% to 4.4% between the two birth-cohorts.
Among
children who came from households with a bucket toilet, the proportion of under5 deaths declined from 9.0% for the 1993-1997 birth-cohort to 4.8% among the
1998-2002 birth-cohort.
4.3.2 Proportion of children who died under-5 by socio-economic factors
Figures 4.7 to 4.9 show the proportions of children who died before reaching the
age of 5 by socio-economic factors.
Figure 4.7 above suggests that the proportion of under-5 deaths is higher among
children born to mothers with less than matric compared to those with mothers
who completed matric or higher. This is true for both birth cohorts under review.
Unlike children born to mothers with less than matric, the proportion of deaths
among children born to women with matric or higher remained unchanged during
the two periods under review
57
Figure 4.7: Proportion of under-5 deaths by mother’s level of education:
birth cohorts 1993-1997 and 1998-2002
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Matric and above
1.6
1.6
Less than matric
8.5
3.2
Figure 4.8: Proportions of under-5 deaths by mother's employment status:
birth cohorts 1993-1997 and 1998-2002
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Employed
4.9
2.5
Not employed
8.0
2.9
58
Figure 4.8 shows, that there seems to be a relationship between the employment
status of the mother and child survival. The proportion of under-5 deaths is lower
among children born to mothers who are employed as compared to those born to
non working mothers. Proportions of under-5 deaths declined for both working
and non-working mothers.
A drastic decline was observed among under-5
deaths of children whose mothers were not employed.
The proportion of under-5 deaths is higher among children born to mothers who
live in non-urban areas (7.5 and 3.2 percent) as compared to those, whose
mothers live in urban areas (7.0 and 2.4 percent).
However, the difference
between urban and non-urban is minimal. Secondly the proportion of under-5
deaths declined for both urban and non-urban between the two birth-cohorts.
Figure 4.9: Proportions of under-5 deaths by mother's residence:
birth cohorts 1993-1997 and 1998-2002
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Urban
7.0
2.4
Non-urban
7.5
3.2
59
4.3.3 Proportion of children who died under-5 by biological and maternal
factors
The four biological and maternal factors are presented in the Figures 4.10 to
4.13. There was a huge variation among the 1993-1997 birth cohort in respect of
mother’s age at birth as compared to the 1998-2002 birth cohort. For example,
in the 1993-1997 birth cohort, 19.8 percent of the children born to mothers who
were aged 35 or above, died before their fifth birthday as compared to 2.4
percent among children born to mothers between the age of 18 and 34.
Figure 4.10: Proportion of under-5 deaths by Mother's age at birth:
birth cohorts 1993-1997 and 1998-2002
25.0
20.0
15.0
10.0
5.0
0.0
1993-1997
1998-2002
below 18
2.8
2.4
18-34
2.4
2.6
19.8
3.6
35+
However, the huge disparity observed between other age groups and those aged
35 years and over at birth in the 1993-1997 birth cohort declined sharply among
the 1998-2002 birth cohort. In the 1998-2002 birth cohort 3.6 percent of children
born to mothers aged 35 years and over died before the age of 5.
Children born after a long birth interval i.e. 24 months or more appear to have
better survival chances. Figure 4.11 shows that in the 1993-1997 birth cohort
21.3 percent of the children born after a short interval died before they turned 5
years as compared to 4.7 percent among children born after a long birth interval.
60
The levels in the 1998-2002 cohort reduced drastically in that only 5.5 percent of
children born after a short interval died before their fifth birthday. The gap
between the survival rate of those born after a short interval and those born after
24 months is not as huge as in the 1993-1997 cohort.
Figure 4.11: Proportion of under-5 deaths by birth interval: Birth cohorts
1993-1997 and 1998-2002
25.0
20.0
15.0
10.0
5.0
0.0
1993-1997
1998-2002
First born or interval = >
24 months
4.7
2.5
< 24 months
21.3
5.5
Figure 4.12 suggests that the proportion of children who died before the age of 5
is higher among the fourth or higher birth order children in both birth cohorts
under review. There is a decline in proportions of children who died before the
age of 5 in all birth orders in the 1998-2002 birth cohort, however, the most
noticeable change was in the fourth and higher birth order. Children of the fourth
or higher birth order were more vulnerable, in that 14.9 percent died before the
age of 5 in the 1993-1997 birth cohort. This proportion declined to 4.2 percent
among the 1998-2002 birth cohort.
61
With regard to sex, the proportion of children who died before the age 5 is higher
among the boys in both birth cohorts under review, although the levels are lower
in the 1998-2002 birth cohort.
Figure 4.12: Proportions of under-5 deaths by birth order: Birth cohorts
1993-1997 and 1998-2002
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
1993-1997
1998-2002
First
2.6
2.3
Second
3.0
2.2
Third
4.9
3.0
14.9
4.2
Fourth and above
Figure 4.13: Proportions of under-5 deaths by sex: birth cohorts
1993-1997 and 1998-2002
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Girl
6.9
2.4
Boy
7.7
3.1
62
4.3.4 Proportion of children who died under-5 by nutrient deficiency factors
As expected, nutrition seems to play a role in the child’s survival. In the 1997
OHS, households were asked if in the past year, there was ever a time when
children could not be fed because the household could not afford to buy enough
food. In GHS 2002 the question was slightly different; the households were
asked if in the past 12 months, any child in the household went hungry because
there wasn’t enough food.
Figure 4.14: Proportions of under-5 deaths by nutritional status: birth
cohorts 1993-1997 and 1998-2002
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1993-1997
1998-2002
Household without food
shortage
6.3
2.3
Household with food
shortage
9.4
3.5
These questions were used as a proxy to determine the nutritional level of the
children.
The proportion of under-5 deaths is higher among children from
households which had at least a child who went hungry because there was no
food (9,4 and 3,5 percent ) as compared to household from which no child went
hungry (6,3 and 2,3 percent).
63
4.3.5 Proportion of children who died under-5 by health seeking behaviours
The figure below suggests that, the proportion of under-5 deaths is higher among
the children born elsewhere other than in a hospital or clinic. Place of birth is
used as a proxy to determine health seeking behaviours of the mother. For the
1993-1997 birth cohort a very high proportion (22.5%) of children who were born
elsewhere other than hospitals died before the age of 5. This proportion declined
to 6.1 percent among the 1998-2002 birth cohort.
Figure 4.15: Proportions of under-5 deaths by place of delivery: birth
cohorts 1993-1997 and 1998-2002
25.0
20.0
15.0
10.0
5.0
0.0
1993-1997
1998-2002
Hospital or clinic
4.0
2.5
Elsewhere
22.5
6.1
64
4.4 Logistic regression model with one independent variable
Logistic regression was conducted for each independent variable for the two birth
cohorts. This was done to assess the impact of each predictor variable to the
dependent variable in this case the under-5 mortality. The probability that a death
will occur was regressed i.e. child dead=1 and child alive=0. Results of each
variable are presented in Table 4.6.
The results suggest that mother’s education has an impact on under-5 mortality.
The odds ratio show that children belonging to mothers who have not completed
matric are 5.8 times likely to die before their 5th birthday than those belonging to
mothers who have completed matric among the 1993-1997 birth cohort without
controlling for any other variable. This effect reduces to 2.1 times among the
1998-2002 birth cohort.
The results also suggest that place of residence influence under-5 mortality.
While the results were not statistically significant among the 1993-1997 birth
cohort, they were statistically significant among the 1998-2002 birth cohort.
Children whose mothers reside in non-urban areas are 1.3 times likely to die
before age 5 in the 1998-2002 birth cohort compared to children whose mothers
live in urban areas without controlling for any other factor.
Mother’s labour market status also has an impact on under-5 mortality for both
birth cohorts. However, the results are only statistically significant among the
1993-1997 birth cohort. The odds ratio show that children belonging to mothers
who are not working are approximately 1.7 times likely to die before turning 5
years compared to those born to employed mothers without controlling for any
other variable.
Sex of the child showed some relationship with under-5 mortality but in both birth
cohorts it was not significant.
65
Table 4.6: The odds of under-5 death for the birth cohorts 1993-1997 and
1998-2002: model with one independent variable
1993-1997 birth cohort
B
Mother's education (Ref: Matric and
above)
Below matric
S.E.
Sig.
1998-2002 birth cohort
Exp(B)
B
S.E.
Sig.
Exp(B)
1.755
.169
.000
5.782
.754
.178
.000
2.127
.083
.068
.218
1.087
.291
.131
.026
1.338
.503
.092
.000
1.654
.162
.146
.268
1.176
.127
.067
.058
1.136
.236
.131
.071
1.267
Second
.162
.151
.286
1.176
-.050
.183
.782
.951
Third
.656
.145
.000
1.926
.275
.197
.163
1.317
1.888
.112
.000
6.604
.607
.163
.000
1.836
1.704
.070
.000
5.496
.813
.183
.000
2.254
<18
.154
.249
.536
1.167
-.094
.272
.731
.911
>34
2.315
.081
.000
10.129
.328
.162
.042
1.389
-.042
.078
.589
.959
.622
.141
.000
1.862
Mother's place of residence (Ref:
Urban)
Non-urban
Employment status of the mother (Ref:
Employed)
Not employed
Sex of the child (Ref:Girl)
Boy
Birthorder (Ref: First born)
Fourth or above
Birth Interval (Ref: First born or interval
of more than 24 months)
Less than 24 months
Mother's age at birth (Ref:18-34)
Water (Ref: Piped)
Unpiped
Sanitation (Ref: Toilet on the site)
Toilet off-site
Bucket or no toilet
.
-.027
.137
.842
.973
.298
.368
.418
1.347
.293
.082
.000
1.340
.743
.146
.000
2.102
.293
.067
.000
1.340
.696
.132
.000
2.006
.019
.104
.853
1.019
.467
.168
.006
1.595
.417
.068
.000
1.517
.438
.130
.001
1.550
1.930
.069
.000
6.892
.941
.171
.000
2.562
Electricity (Ref: Connected to mains)
Not connected
Type of dwelling (Ref: Formal)
Informal
Nutrition Ref: Household with no child
who went hungry
Household with at least a child went
hungry because no food was available
Place of delivery (Ref: Hospital or clinic)
Else where
66
Birth order shows a mixed picture for the second birth. However, none of the
results were statistically significant. There is some consistency with the third
birth, and the fourth and higher, although the results are also not statistically
significant in both cohorts. The fourth and the subsequent births stand out in the
1993-1997 birth cohort. The fourth or later children are 6.6 likely to die before the
age of 5 compared to the first born child in the 1993-1997 birth cohort without
controlling for any other variable. This effect is reduced to 1.8 times among the
1998-2002 birth cohort.
Among the biological and maternal factors, birth interval shows the expected
results on its impact on under-5 mortality. Children whose birth interval is less
than 24 months are approximately 5.5 times likely to die before age 5 as
compared to the first born baby or those with birth interval of more than 24
months in the 1993-1997 birth cohort without controlling for any other factor.
This is reduced to approximately 2.3 times in the 1998-2002 birth cohort.
Among the 1993-1997 birth cohort, the children born to mothers aged above 34
years of age are 10 times more likely to die before the age of 5 compared to
those born to mothers between 18-34 years without controlling for any other
factor. This strong effect is reduced substantially to approximately 1.4 in the
1998-2002 birth cohort.
The impact of age on under-5 mortality shows conflicting results for children born
to young mothers (below 18 years of age). However, results for both periods
under review are not statistically significant.
Among the environmental factors source of water also gives conflicting results for
the two periods under review. Among the 1993-1997 birth cohort, the results
suggest that water has no impact on under-5 mortality while among the 19982002 it shows children in households with no access to piped water are 1.8 times
likely to die before the age of 5 compared to those in household with piped water
without controlling for any other factor. The unexpected results in the 1993-1997
birth cohort are not statistically significant.
67
The results suggest that sanitation has an impact on under-5 mortality. However,
for the children from households with a toilet facility which is off-site, a mixed
message for the two birth cohorts under review is noted. However none of the
results are statistically significant.
Table 4.6 shows the expected results on the effect of source of energy, nutrition,
type of dwelling and place of delivery on under-5 mortality. However, the results
for the type of dwelling among the 1993-1997 birth cohort are not statistically
significant.
4.5 Logistic regression with multiple independent variables
Tables 4.7 and 4.8 show the results of the 5 models generated from the
hierarchical logistic regression analysis for both the 1993-1997 and 1998-2002
birth cohorts respectively. In each case model 1 evaluates the effect of socioeconomic variables to under-5 mortality without controlling for any other factor.
After that biological and maternal factors were entered into the model and the
results are presented in model 2. Model 2 evaluates the impact of biological and
maternal factors while controlling the socio-economic factors.
Environmental
factors were entered in the third model (model 3), then nutrition in model 4 and
finally in model 5 place of delivery or health seeking factors.
Model 1 shows that mother’s education has an impact on under-5 mortality. After
controlling for place of residence and labour market status of the mother, children
born to mothers who have not completed matric among the 1993-1997 birth
cohort are 5.5 times more likely to die before turning 5 years as compared to
those born to mothers who have completed matric. Although the impact of
education is reduced among the 1998-2002 birth cohort a similar pattern is
observed. For example, children born to mothers without matric are 2.0 times
likely to die before 5 years as compared to those born to mothers who completed
matric among the 1998-2002 birth cohort, after controlling for place of residence
and labour market.
68
Type of residence gives a mixed message in that, among the 1993-1997 birth
cohort, it shows that living in a non-urban area decreases the risk of under-5
deaths while among the 1998-2002 birth cohort children born to mothers in a
non-urban area where 1.2 times more likely to die compared to those in urban
areas without controlling for any other variable. However, these results are not
significant in both cases.
In the 1993-1997 birth cohort employment children born to mothers who are not
employed are 1.3 times more likely to die compared to those born to employed
mothers after controlling for mothers education and place of residence. This
pattern is maintained in the 1998-2002 birth cohort but the results are not
significant.
Model 2 shows the impact of adding the biological and maternal factors to the
model. In the 1993-1997 birth cohort, the influence of mother’s education to
under-5 mortality decreased after controlling for biological and maternal factors.
For example, children born to mothers who did not complete secondary school
(matric) are now 2.4 times likely to die before turning 5 years as compared to
those born to mothers who completed matric. This is also true in the 1998-2002
birth cohort. In essence controlling for biological factors reduces the effect of
education on under-5 mortality.
Type of residence continued to give a mixed message between the two birth
cohorts. However, the results were statistically significant in 1993-1997 after
controlling for biological and maternal factors and socio-economic variables.
The impact of the labour market status of the mother increased slightly after
controlling for biological and maternal factors. While the impact of labour market
status is significant in the 1993-1997 birth cohort, it is not significant in the 19982002 birth cohort.
69
After controlling for socio-economic factors, birth interval and mothers age show
a significant influence on under-5 mortality in the 1993-1997 birth cohort. In the
1998-2002 only sex and birth interval show a significant impact on under-5
mortality.
Among the 1993-1997 birth cohort, the children with birth interval of less than 24
months are approximately 2.8 times likely to die before turning 5 years as
compared to those with higher birth interval. In the 1998-2002 this figure reduces
to 2.0. In both birth cohorts these effects are significant.
Children born to mothers aged 35 and above were approximately 6.0 times more
likely to die before the age of 5 years compared to children born to mothers
between 18-34 years of age after controlling for socio-economic variable.
Surprisingly this effect disappears in the 1998-2002 birth cohort. It is not even
statistically significant among this group.
The results also show that sex of the child has an impact on under-5 mortality.
The results were not statistically significant in the 1993-1997 birth cohort,
however, they were statistically significant in the 1998-2002 birth cohort. During
the 1998-2002 period a boy child was 1.3 times more likely to die before age 5
compared to a girl child after controlling for socio-economic variables and other
maternal and biological factors.
The impact of birth order for both birth cohorts was not statistically significant
after controlling for socio-economic factors, mothers age, sex and birth interval.
70
Table 4.7: The odds of under-5 death for the birth cohorts 1993-1997: model with multiple independent variable
Model 1
B
Exp(B)
Model 2
Sig.
B
Model 3
Exp(B)
Sig.
B
Exp(B)
Model 4
Sig.
B
Exp(B)
Model 5
Sig.
B
Exp(B)
Sig.
Mother's education (Ref:
Matric and above)
1.709
5.523
.000
.877
2.404
.000
.832
2.299
.000
.812
2.253
.000
.663
1.941
.000
-.132
.877
.058
-.247
.781
.001
-.252
.777
.006
-.257
.773
.005
-.393
.675
.000
.311
1.365
.001
.333
1.394
.001
.333
1.396
.001
.330
1.391
.001
.166
1.180
.111
.134
1.143
.066
.138
1.147
.059
.139
1.149
.056
.168
1.183
.024
Second
-.299
.741
.069
-.295
.744
.073
-.296
.744
.073
-.287
.751
.085
Third
-.169
.844
.299
-.169
.845
.301
-.170
.844
.297
-.143
.866
.383
.271
1.311
.058
.276
1.317
.053
.267
1.307
.061
.299
1.349
.035
Birth Interval (Ref: First born
or interval of more than 24
months)
Less than 24 months
1.029
2.798
.000
1.013
2.754
.000
1.013
2.753
.000
.770
2.159
.000
Mother's age at birth (Ref:1834)
<18
.124
1.132
.636
.131
1.140
.618
.113
1.120
.667
.205
1.228
.437
1.784
5.952
.000
1.786
5.963
.000
1.779
5.924
.000
1.523
4.586
.000
Below matric
a
Mother's place of residence
(Ref: Urban)
Non-urban
Employment status of the
mother (Ref: Employed)
Not employed
Sex of the child (Ref:Girl)
Boy
Birthorder (Ref: First born)
b
Fourth or above
>34
a)
Social economic factors
b)
Biological and maternal factors
c)
Environmental factors
d)
Nutrient deficiency
e)
Health seeking behaviour
71
Table 4.7 (cont.) The odds of under-5 death for the birth cohorts 1993-1997: model with multiple independent variable
Model 1
B
Exp(B)
Model 2
Sig.
B
Exp(B)
Model 3
Sig.
B
Exp(B)
Model 4
Sig.
B
Exp(B)
Model 5
Sig.
B
Exp(B)
Sig.
.830
.065
Water (Ref: Piped)
Unpiped
-.198
.821
.046
-.193
.825
.051
-.186
Sanitation (Ref: Toilet on-site)
Toilet offsite
c
Bucket or no toilet
-.037
.964
.803
-.036
.965
.808
-.126
.882
.409
.077
1.080
.404
.059
1.060
.528
-.095
.909
.321
.191
1.211
.025
.181
1.198
.035
.102
1.108
.245
-.026
.974
.831
-.050
.951
.680
.018
1.018
.887
.190
1.209
.012
.161
1.175
.036
1.211
3.356
.000
-4.636
.010
.000
Electricity (Ref: Connected to
mains)
Not connected
Type of dwelling (Ref: Formal)
Informal
Nutrition Ref: Household with
no child who went hungry
d
e
Household with at least a
child went hungry because no
food was available
Place of delivery (Ref:
Hospital or clinic)
Else where
constant
Degree of freedom
x²
-4.262
.000
-4.783
.008
.000
-4.800
.008
.000
-4.826
.008
.000
10
15
16
17
196.280
1351.237
1359.605
1365.903
1582.992
.119
1154.957
.043
8.368
.017
6.298
.126
217.089
.000
x² change from model n-1
Hosmer and Lemeshow Test
.014
3
72
Table 4.8: The odds of under-5 death for the birth cohorts 1998-2002: model with multiple independent variable
Model 1
B
a
Model 2
Exp(B)
Sig.
B
.705
2.024
.000
Mother's place of residence (Ref:
Urban)
Non-urban
.165
1.180
Employment status of the mother
(Ref: Employed)
Not employed
.029
1.030
Model 3
Exp(B)
Sig.
B
.484
1.623
.011
.430
-.125
.883
1.087
.581
.031
.260
1.296
.048
-.169
.845
Third
.103
Fourth or above
.280
.710
Mother's education (Ref: Matric and
above)
Below matric
Exp(B)
Sig.
B
.633
1.882
.001
.217
.107
1.113
.845
.083
Model 4
Model 5
Exp(B)
Sig.
B
Exp(B)
Sig.
.460
1.584
.017
.441
1.555
.022
.462
-.118
.889
.486
-.150
.861
.380
1.032
.837
.023
1.023
.880
.023
1.023
.882
.274
1.315
.038
.269
1.309
.041
.270
1.310
.041
.383
-.160
.852
.409
-.159
.853
.412
-.185
.831
.340
1.109
.625
.084
1.324
.177
.181
1.088
.692
.088
1.092
.678
.050
1.051
.816
1.198
.388
.167
1.182
.426
.091
1.095
.667
2.035
.000
.676
1.966
.001
.686
1.986
.000
.665
1.945
.001
Sex of the child (Ref:Girl)
Boy
Birth order (Ref: First born)
Second
b
Birth Interval (Ref: First born or
interval of more than 24 months)
Less than 24 months
Mother's age at birth (Ref:18-34)
<18
-.156
.856
.591
-.132
.876
.649
-.141
.868
.626
-.144
.866
.620
>34
.062
1.064
.743
.114
1.121
.548
.113
1.120
.553
.122
1.129
.524
73
Table 4.8 (cont.) The odds of under-5 death for the birth cohorts 1998-2002: model with multiple independent variable
Model 1
B
Exp(B)
Model 2
Sig.
B
Exp(B)
Model 3
Sig.
B
Model 4
Exp(B)
Sig.
B
1.475
.035
.373
Model 5
Exp(B)
Sig.
B
1.453
.043
.346
Exp(B)
Sig.
1.413
.062
Water (Ref: Piped)
Unpiped
.389
Sanitation (Ref: Toilet on-site)
Toilet offsite
c
Bucket or no toilet
-.053
.948
.888
-.039
.962
.918
-.021
.979
.955
.323
1.381
.053
.304
1.356
.069
.275
1.317
.102
.317
1.373
.048
.293
1.340
.069
.243
1.275
.134
.427
1.532
.023
.411
1.509
.028
.430
1.538
.022
.199
1.220
.142
.181
1.199
.184
.553
1.739
.003
.011
.000
Electricity (Ref: Connected to mains)
Not connected
Type of dwelling (Ref: Formal)
Informal
d
e
Nutrition Ref: Household with no
child who went hungry
Household with at least a child
went hungry because no food was
available
Place of delivery (Ref: Hospital or
clinic)
Else where
constant
Degree of freedom
x²
-4.221
.015
.000
-4.436
.012
.000
-4.487
.011
.000
-4.531
.011
.000
3
10
15
16
4.501
17
22.57
48.93
77.32
79.47
87.61
.039
.848
.006
.176
.518
x² change from model n-1
Hosmer and Lemeshow Test
74
Model 3 assesses the impact of environmental factors after controlling for socioeconomic variables, biological and maternal variables. During the 1993-1997
period, it is only the source of water and source of energy which has a significant
impact on the under-5 mortality. However, during the 1998-2002 period the
impact of type of dwelling is also statistically significant. Children in households
without piped water were 1.5 times more likely to die before the age of 5
compared to those children from households with access to piped water after
controlling for socio-economic variables, biological, maternal and other
environmental variables.
Children from households without access to electricity were approximately 1.4
times more likely to die before their fifth birthday as compared to those from
households with access to electricity after controlling for socio-economic
variables, biological, maternal and other environmental variable. This influence
was lower among the 1993-1997 birth cohorts.
The influence of mother’s education decreased as more factors were controlled.
In the 1998-2002 birth cohort the risk of the child dying before age 5 increased by
62% for a mother without matric compared to the one with matric in model 3.
However, in model 2 where only socio-economic variables, biological and
maternal factors were controlled this increase was 88%.
The impact of type of dwelling on under-5 mortality shows conflicting results for
the two periods under review. However, during the 1998-2002 period the risk of
under-5 death increased by 53% for children living in informal dwelling as
compared to those living in formal dwelling after controlling for socio-economic,
biological, maternal and other environmental variables.
75
Model 4 assesses the impact of nutrition after controlling for socio-economic,
biological, maternal and environmental variables. Children from households
which reported hunger are 1.2 times more likely to die before their fifth birth day
compared to those from households without hunger reporting after controlling for
socio-economic, biological, maternal and environmental variables. The influence
is maintained in the 1998-2002 birth cohort but not significant. The impact of both
source of water and type of dwelling decreased slightly after nutrition was
controlled. The impact of source of energy decreased and even became
insignificant during 1998-2002 period after controlling for nutrition.
Finally model 5 assesses each variable while others are controlled. The logistic
regression results in model 5 for both cohorts show the changing pattern in the
factors associated with under-5 mortality during 1993-1997 and 1998-2002
reporting periods. For example mother’s education, Mother’s age at the time of
delivery of the subject child, place of delivery, nutrition, birth order and birth
interval were among the factors which had a
significant impact on under-5
deaths during 1993-1997 reporting period. However, the pattern changed for
some factors during 1998-2002 reporting period. For the example, the odds of a
child dying before the age of 5 were 1.9 times higher for the mothers who did not
complete matric as compared to the mother who completed matric or higher for
the 1993-1997 cohort. However, for the 1998-2002 birth cohort this effect had
reduced to approximately 1.6 times and the statistical significance had reduced.
Mother’s age at the time of delivery of the subject child was the most important
factor for the 1993-1997 birth cohort. The odds of a child dying before the age of
5 were approximately 4.6 times higher for mothers who were 35 years of age and
older at the time of delivery compared to those mothers who were between the
ages 18-34. However, mother’s age at birth as a factor in explaining the under-5
death had almost disappeared for the 1998-2002 cohort.
76
Place of delivery was the second most important factor in explaining the under-5
deaths for the 1993-1997 cohort. The odds of a child dying were 3.4 times higher
for the children who were born elsewhere as compared to those who were born
in a hospital or clinic. However, for the 1998-2002 the odds of a child dying were
approximately 1.7 times higher for the same group.
Mother’s place of residence, access to clean water and sanitation gave
unexpected results for the 1993-1997 cohort. This pattern changed for the 19982002 cohort for both water and sanitation. The odds of a child dying before the
5th birthday were 1.4 times higher in households with no access to piped water
compared to those with access to piped water. Children living in households with
bucket toilet or no toilet at all were 1.3 times likely to die before the age of 5
compared to those in households with a toilet on site although the results are not
statistically significant.
Access to electricity as a predictor of under-5 death recorded an odds ratio 1.1
and approximately 1.3 for the 1993-1997 and 1998-2002 birth cohorts
respectively after controlling for other variables. The impact of access to
electricity to under-5 mortality declined when nutrition factor was controlled, it
even declined further to insignificant levels when health seeking behaviour factor
was controlled. This could mean that providing electricity without improved health
systems and health seeking behaviour will not help reduce under-5 mortality.
Type of dwelling was not very prominent in explaining under-5 mortality among
the 1993-1997 cohort, however this changed among the 1998-2002 cohort,
where children living in informal dwellings were almost 1.5 times more likely to
die before the age 5 as compared to those living in formal houses.
In both periods children living in a household where a child had gone hungry
because there was no food were approximately 1.2 times more likely to die
77
before age 5 as compare to those children living in households where no child
went hungry although not significant during 1998-2002 period
Fourth or higher birth order children were 1.4 times likely to die before the age of
5 as compared to first born for the 1993-1997 birth cohort. However for the
1998-2002 birth cohort the fourth or higher birth order children were
approximately 1.1 times more likely to die as compared to first born. Noticeably
the third order children were 1.1 times likely to die before turning 5 as compared
to the first born, which was not the case for the 1993-1997 birth cohort.
As expected children with a short birth interval, i.e. less than 24 months were
approximately 2.2 and 1.9 times likely to die before age 5 as compared to those
children who were either the first born or with birth interval of more than 24
months for both birth cohorts under review respectively after controlling for all
other variables.
Sex of the child recorded an odds ratio of approximately 1.2 and 1.3 for the two
birth cohorts respectively. A boy child was 1.3 times likely to die as compared to
a girl child for the 1998-2002 birth cohort. Among the 1993-1997 birth cohort the
influence of sex only became significant after controlling for place of delivery.
4.6 Evaluation of the Models
The two tests, (omnibus test and Hosmer-Lemeshow test) which were used to
evaluate logistic regression model send a mixed message for model 5 of the
1997 data. However, the tests suggest a good fit for the equivalent 2002 model 5
which included all the predictor variables. In the 1997 model 5 the omnibus test
suggests a good fit. It shows a high significant figure of 0.000. However, the
Hosmer and Lemeshow test does not support the omnibus test results. HosmerLemeshow results are interpreted differently from the Omnibus test. For the
Hosmer-Lemeshow Goodness of Fit Test the poor fit is indicated by a
78
significance value less than 0.05 (Pallant, 2005). In case of 1997 model 5
Hosmer-Lemeshow test show a significance value of 0.000 which suggests a
poor fit.
For the 2002 model 5 both tests suggest a good fit. The Omnibus test show a
high significant figure of 0.000 and this is supported by the Hosmer-Lemeshow
figure of 0.518. Interpretation of the 1997 model should be done with caution.
4.7 Summary
This chapter presented the results of the analyses of the data. The results from
the descriptive analysis and from the hierarchical logistic regression analysis for
both 1993-1997 and 1998-2002 birth cohorts were provided and discussed.
These results support those from studies conducted elsewhere. Mother’s
education showed a significant relationship with under-5 mortality. The effect of
mother’s education on child survival usually survived controls of other socioeconomic variables. Some results are implausible, particularly the under-5
mortality rate generated from the 2002 General Household survey.
79
CHAPTER 5
5. DISCUSSIONS AND CONCLUSION
5.1 Introduction
This chapter revisits the results presented in the previous chapter and highlights
the key findings from the study. Secondly, reference to the analytical framework
and literature used chapter 2 will be made in analysing the results in the previous
chapter. Finally, the policy implications of this study are identified and possible
future research is proposed.
5.2 Summary of findings
The results suggest changing patterns in factors associated with under-5
mortality between the two birth cohorts: 1993-1997 and the 1998-2002 birth
cohorts. In the order of importance, the factors which were predominant in the
1993-1997 birth cohort were, mother’s age at birth, place of delivery and birth
interval and mothers education, while in the 1998-2007 factors which were
predominant were, birth interval, place of delivery, mother’s education, type of
dwelling and access to piped water.
The secondary analysis of 1993-1997 and 1998-2002 data was an attempt to fill
the gap in explanation of child mortality in South Africa. Hopefully this will help in
designing relevant programmes in order to achieve the Millennium Development
Goal number 4 (MDG4).
The 1997 OHS reported the under-5 mortality rate of 58 deaths per 1 000 live
births which looks plausible compared to other sources like 59 deaths per 1 000
live births reported by UNICEF, et al. (2007) and 59,4 deaths per 1 000 live births
from the 1998 SADHS. However the rate of 24,3 deaths per 1 000 live births
reported in the 2002 GHS does not look plausible because it is not in line with
80
other sources. UNICEF, et al. (2007) reported 63 deaths per 1 000 live births for
the year 2000 and 68 deaths per 1 000 live births for 2005. The 2003 SADHS
reported 58 deaths per 1 000 live births. This suggests a serious under reporting
of deaths in the 2002 GHS so the study cannot answer the question of whether
levels of under-5 mortality changed between the two reporting periods under
review. However, the study showed the trends in the factors associated with
under-5 mortality.
In summary under-5 mortality was significantly associated with eight factors
during 1993-1997 period namely; mother’s education, mother’s place of
residence, sex, birth order, birth interval, mother’s age at the time of delivery of
the subject child, nutrient deficiency and place of delivery. However, during the
1998-2002 period only five factors were significantly associated with under-5
mortality. These were mother’s education, sex, birth interval, type of dwelling and
place of delivery.
5.3 Discussions and conclusion
The discussion focuses on the conceptual framework in the study and attention
was paid to role played by socio-economic factors, biological and maternal
factors, environmental factors, nutrient deficiency factors and health seeking
behaviour factors.
5.3.1 Socio-economic factors and environmental factors
The unexpected results of current residence of the mother, access to water and
sanitation for the 1993-1997 birth cohort could have been caused by migration of
mothers from non-urban areas to peri-urban areas. During this period there was
mass migration of especially the black population to big towns where, in some
cases water and sanitation was better. Secondly, the study uses current
residence of the mother not residence during the birth of the subject child. It is
common in South Africa for mothers to leave their children in rural areas and
81
move to big cities in search of jobs. The survey did not collect information on
whether the child was staying with the mother at the time of death.
However, it could also be true that children in urban areas of South Africa are
more susceptible to diseases and subsequent death than those in non-urban
areas. Some studies have shown similar results. Sastry (2004) concluded that in
Sao Paulo children from disadvantaged backgrounds were worse off in urban
areas. Amouzou and Hill (2004) concluded that the weak effect of urbanization
could be due to the rapid increase of urban poverty. This is a more plausible
explanation in the case of South Africa in light of the massive non-urban to urban
migration of the black population after the apartheid laws were abolished in 1990.
Although mother’s place of residence continued to give unexpected results for
the 1998-2002 birth cohort, access to piped water and good sanitation were
among the factors which were associated with the under-5 deaths. Similar results
were observed by Anderson, et al. (2002) and Mahmood (2002). This could
further strengthen the earlier argument of migration but this time mothers could
have moved to informal settlements which are classified as urban with relatively
poor access to piped water and sanitation. This is supported by the increased
influence of type of dwelling in explaining the under-5 mortality among the 19982002 birth cohort. Similar results of a relationship between type of dwelling and
child mortality were observed by Anderson, et al. (2002) and Jacobs, et al.
(2009).
5.3.2 Socio-economic factors and health seeking behaviour
An assumption could be made that the employment of women or their
participation in the labour force would empower them and this would have an
impact on their children’s survival. This does not seem to be case, especially for
the 1998-2002 birth cohort.
Among the 1993-1997 birth cohort the children
belonging to women who are not employed are 1.4 times likely to die before the
82
age of 5 after controlling for socio-economic and biological and maternal factors.
The odds did not change much after extra control of the environmental and
nutrient deficiency variables. However, the effect of employment status became
statistically insignificant after controlling for place of delivery (see 1997 model 5).
This could imply that, the employment of women in low paying jobs without
improvement in access to health facilities does not help in reducing under-5
mortality, especially in high HIV/AIDS prevalence population.
5.3.3 Biological and maternal factors
Mother’s age at the time of delivery of the subject child as a predictor of under-5
mortality was very important among the 1993-1997 birth cohort, especially
mother’s who were 35 years of age and above at time of birth. This is similar to
what Mahmood (2002) observed but contrary to what Hobcraft, et al. (1985)
reported. Although this age group is known to be a risk age for child bearing for
both, the mother and the child, the increase in the risk of a child’s death in the
1993-1997 birth cohort by approximately 4.6 times compared to a child belonging
to a mother between the ages of 18 and 34 after controlling for all other variables
needs to be investigated further. This effect almost disappeared in the 19982002 birth-cohort to agree with Hobcraft, et al. (1985). Although mother’s age
especially 35 and older at time of birth shows some minor increased risk of
under-5 mortality, the results are not statistically significant. The question is what
policy could have been responsible for this change? In fact, it seems there was
another factor rather than age which was playing a role. Further research should
be conducted in order to explain this phenomenon.
The impact of birth interval persisted in both birth cohorts even after controlling
for all other factors. This is in line with other studies conducted by Hobcraft, et al.
(1985), Hobcraft (1991) and Boerma and Bicego (1992). Models 4 and 5 of 1997
seem to suggest what Boerma and Bicego (1992) and Mahmood (2002)
concluded, that babies born to mothers with shorter previous birth intervals who
83
received prenatal care are significantly more likely have better survival chances
during neonatal period than babies born to mothers with the same short interval
who did not receive prenatal care for the index child.
5.3.4 Health seeking behaviour and HIV/AIDS
Place of delivery also featured prominently in explaining under-5 mortality. This is
in line with what Rutstein (2000) observed in his comparison of DHS data. This
could be due to the known reasons that children born elsewhere are more
susceptible to infections, but it could also be that in countries with high
prevalence of HIV like South Africa, women who deliver in hospitals or clinics are
more likely to be tested for HIV and take precautionary measures to reduce
mother to child transmission.
5.4 Contribution of the study
The study has brought to light key policy implications which government needs to
focus on. The empowerment of women through education should be encouraged
because mother’s education continues to influence under-5 mortality. Completing
high school (matric) for the mother reduced the risk of under-5 mortality by
almost 55%.
Education of the mother cannot be ignored, some studies by
Hobcraft, et al. (1984), Cleland (1990) and Hobcraft (1993) have shown the
impact of mother’s education on child mortality.
The increasing number of people moving to urban areas in search of good life
has a devastating impact on children, especially when people move to informal
settlements.
A human settlement strategy has to be developed and integrated
with services delivery especially in light of rural-urban migration. This should
also be integrated with programmes to alleviate urban poverty.
.
84
Health services should be brought nearer to the communities so that mothers
have access to them both during pregnancy and after. The impact of delivering a
child with the help of a medical professional is enormous. However, this will be
useless if mothers do not get access to medicines, especially in high HIV/AIDS
prevalence population. Administration of ARVs to pregnant mothers to prevent
mother to child transmission should be enhanced and continually improved.
All policies developed should be integrated with women empowerment
programmes especially through education. Women with low status usually do
not fully utilize the facilities provided to them. This could hamper progress even if
government improves the health care system.
Finally, the study shows changing patterns in factors associated with under-5
mortality for the period 1993-2002. Therefore, government should monitor and
evaluate existing programmes regularly in order to revise or re-design
programmes which are more relevant to the factors which are predominant in
determining child survival.
85
References
Adetunji, J., 2000. ‘Trends in under-5 mortality rates and the HIV/AIDS epidemic’ Bulletin of
the World Health Organisation, 2000, 78(10).
Amouzou, A. & Hill, K., 2004. ‘Child mortality and socioeconomic status in sub-Saharan
Africa’. African Population Studies/Etude de la Population Africaine. 2004;19(1):1-11.
Anderson, B.A., Romani, J.H., Phillips, H.E. & Van Zyl, J.A., 2002. ‘Environment, Access to
health care and Other factors affecting infant and child survival among the African and
Coloured populations of South Africa, 1989-1994’. Population and Environment vol. 23 No. 4.
Arokiasamy, P., 2002. ‘Regional Patterns of Sex Bias and Excess Female Child Mortality in
India’ Population (English Edition, 2002-), Vol. 59, No. 6 (Nov. - Dec., 2004), pp. 833-863
Bhuiya, A. & Streatfield, K., 1991. ‘Mothers' Education and Survival of Female Children in a
Rural Area of Bangladesh’. Population Studies, Vol. 45, No. 2 (Jul., 1991), pp. 253-264.
Boerma, J.T. & Bicego, G.T., 1992. ‘Preceding Birth Intervals and Child Survival: searching
for pathways of influence’ Studies in Family Planning: 23, 4:243-256.
Bomela, N. J., 1999. ‘Child Nutritional Status and Household Patterns in South Africa’ Minidissertation University of Pretoria.
Brahmbhatt, H. & Gray, R., 2000. ‘Breastfeeding and the prevention of infant mortality’. The
Lancet, Volume 355, Issue 9212, Pages 1370 – 1370.
Chen, LC., Huq, E. & D'Souza, S., 1981. ‘Sex Bias in the Family Allocation of Food and
Health Care in Rural’ Population and Development Review, Vol. 7, No. 1.
Cleland, J., 1990. ‘Maternal education and child survival: Further evidence and
explanations’. In J.C Caldwell et al (eds). What we know about Health Transition: The Cultural,
Social and Behavioural Determinants of Health. Vol. Health Transition Centre, The Australian
National University.
Davanzo, J., Razzaque, A., Rahman, M., Hale, L., Ahmed, K., Ali Khan, M., Mustafa, G. & Gausia
K., 2004. ‘The effects of birth spacing on infant and child mortality, Pregnancy outcomes,
and maternal mortality in Matlab, Bangladesh’. Working paper RAND labour and population
2004.
Department of Health, ‘South Africa Demographic and health survey – 1998’
http://www.doh.gov.za/facts/sadhs-f.html.
Dorrington, R., Bradshaw, D., Johnson, L. & Budlender D., 2004. ‘The Demographic Impact of
HIV/AIDS in South Africa: National Indicators for 2004’.Joint publication by the Centre for
Actuarial Research, the Burden of Disease Research Unit (Medical Research Council) and the
Actuarial Society of South Africa.
Eswaran, M., 2002. ‘The Empowerment of Women, Fertility, and Child Mortality: Towards a
Theoretical Analysis’. Journal of Population Economics, Vol. 15, No. 3 (Aug., 2002), pp. 433454.
86
Ezzati, M. & Kammen, D.M., 2002. ‘The Health Impacts of Exposure to Indoor Air Pollution
from Solid Fuels in Developing Countries: Knowledge, Gaps, and Data Needs’.
Environmental Health Perspectives, Vol. 110, No. 11 (Nov., 2002), pp. 1057-1068.
Fitzgerald, E.F., Schell, L.M., Marshall, E.G., Carpenter, D.O., Suk, W.A., Jan E. & Zejda, J.E.,
1998. ‘Environmental Pollution and Child Health in Central and Eastern Europe’. Environmental
Health Perspectives, Vol. 106, No. 6 (Jun., 1998), pp. 307-311.
Garenne, M. & Gakusi, E., 2005. ‘Under-five Mortality Trends in Africa: Reconstruction from
Demographic Sample Survey’. Demographic and Health Research 2005 No. 26.
Hesketh, T. & Xing, Z.W., 2006. ‘Abnormal Sex Ratios in Human Populations: Causes and
Consequences’ Proceedings of the National Academy of Sciences of the United States of
America,Vol. 103, No. 36 (Sep. 5, 2006), pp. 13271-13275.
Hill, K., 2003. ‘Frameworks for studying the determinants of child survival’. Bulletin of World
Health Organization 81(2).
Hobcraft, J.N., McDonald, J.W. & Rutstein S.O., 1984. ‘Socio-economic Factors in Infant and
Child Mortality: A Cross-national Comparison’ Population studies, 38 (2), 193-223.
Hobcraft, J.N., McDonald, J.W. & Rutstein S.O., 1985. ‘Demographic Determinants of Infant
and Early Child Mortality: A Comparative Analysis’. Population studies, 39 (1985), 363-385.
Hobcraft, J., 1991. “Child spacing and child mortality”. Paper prepared for presentation at
Demographic and Health surveys World Conference, 1991 Washington D C.
Hobcraft, J., 1993. ‘Women’s education, child welfare and child survival: a review of
evidence’. In Health transition Review Vol 3 No. 2 1993.
Ibrahim, M.M., Aden, A.S., Omar, H.M., Wall, S. & Persson, A., 1994. ‘Diarrhoea among
children in rural Somalia. Maternal perceptions, management and mortality’. Annals of
Tropical Paediatrics, 14, 215-222.
Jacobs, D.E,. Wilson, J., Dixon,S.L., Smith,J. & Evens, A., 2009. ‘ The Relationship of
Housing and Population Health: A 30-Year Retrospective Analysis’. Environmental Health
Perspectives, Vol. 117, No. 4 (Apr., 2009), pp. 597-604.
Kanaiaupuni, S.M. & Donato, K.M., 1999. ‘Migradollars and mortality: the effects of migration
on infants survival in Mexico’. Demography 36(3).
Kishor, S. & Parasuraman, S., 1998. ‘Mother’s Employment and Infant and Child Mortality in
India’. National Health Survey Subject Report No.8. International Institute for Population
Sciences, Mumbai India.
Krishnaji, N., 1995. ‘Working Mothers and Child Survival in Rural India: Insights from
Spatial Patterns’. Economic and Political Weekly, Vol. 30, No. 44 (Nov. 4, 1995), pp. 28032805+2807-2808.
Lopez, A.D., 2000. ‘Reducing child mortality’. Bulletin of the World Health Organization, 2000,
78 (10).
Machado, C.J. & Hill, K., 2005. ‘Maternal, Neonatal and Community Factors Influencing
Neonatal Mortality in Brazil’. J. Biosoc. Sci. 37, 193-208.
87
Madise, N.J., Matthews. Z. & Margetts, B., 1999. ‘Heterogeneity of Child Nutritional status
between households: A comparison of six sub-saharan Africa’ Population Studies, 53(3).
Magadi, M.A., 1997. ‘Status of Women and infant/child health in Kenya with particular
reference to high mortality zone in Nyaza Province’. UAPS Study Report.
Mahmood, M.A., 2002. ‘Determinants of Neonatal and Post-neonatal Mortality in Pakistan’.
The Pakistan Development Review 41:4 part II (Winter 2002) pp. 723-744.
Mohamed, W.N., Diamond, I., Smith, P.W.F., 1998. ‘The Determinants of Infant Mortality in
Malaysia: A Graphical Chain Modelling Approach’. Journal of the Royal Statistical Society.
Series A (Statistics in Society), Vol. 161, No. 3(1998), pp. 349-366.
Mosley, W.H. & Chen, L.C., 1984. ‘An analytical framework for the study of child survival in
developing countries’. In: Child survival: strategies for research, edited by W.H. Mosley and
Lincoln C. Chen. New York, New York, Population Council, 1984. : 25-45. (Population and
Development Review 10, Supplement, 1984).
Murray, J.L., Laakso, A., B., Shibuya, K., Hill, K. & Lopez, A.D., 2007. ‘Can we achieve
Millennium Development Goal 4? New analysis of country trends and forecasts of under-5
mortality to 2015’. Lancet 2007; Vol (370) 1040-54.
Nannan, N., Bradshaw, D., Timaeus, I.M. & Dorrington R., 2000. ‘The impact of HIV/AIDS on
infant and child mortality in South Africa’ International Conference on AIDS.
Omar, B.A., Lopez, A.D. & Inoue, M., 2000. "The Decline in Child Mortality: A Reappraisal".
Bulletin of the World Health Organization 78(10): 1,175–1,191.
Pallant, J., 2005. ‘SPSS survival manual guide’ 2nd edition. Open University Press.
Pande, R.P., 2003. ‘Selective Gender Differences in Childhood Nutrition and Immunization
in Rural India: The Role of Siblings’. Demography, Vol. 40, No. 3 (Aug., 2003), pp. 395-418.
Rutstein, S.O., 2000. Factors associated with trends in infant and child mortality in
developing countries during the 1990s. Bulletin of World Health Organization, 2000, 78(10).
Sastry, N., 2004. ‘Urbanization, development, and under-five mortality differentials by place
of residence in Sao Paulo, Brazil, 1970-1991’. Demographic Research-Special Collection 2:
Article 14. Determinant s of Diverging Trends in Mortality.
Shehzad, S., 2006. ‘The Determinants of Child Health in Pakistan: An Economic Analysis’.
Social Indicators Research, Vol. 78, No. 3 (Sep., 2006), pp. 531-556.
Short, S.E., Chen F., Entwisle, B. & Fengying, Z., 2002. ‘Maternal Work and Child Care in
China: A Multi-Method Analysis’. Population and Development Review, Vol. 28, No. 1 (Mar.,
2002), pp. 31-57.
Statistics South Africa., 2008. ‘Mortality and Causes of Deaths in South Africa, 2006:
Findings From Deaths Notification’.
Thomas. K.J.A., 2007. ‘ Child Mortality and Socioeconomic Status: An Examination of
Differentials by Migration Status in South Africa’ International Migration Review 41 (1), 40-74.
88
UNICEF, WHO, The World Bank & United Nations Population Division., 2007. ‘Levels and
trends of child mortality in 2006’ Estimates developed by the inter-agency Group for Child
Mortality Estimation.
UNICEF., 2004. Progress For Children, A Child Survival Report Card: Number 1, 2004
http://www.unicef.org/progressforchildren/2004v1/industrialized.php
United Nations., 1992. ‘Child Mortality since the 1960s: A database for developing
countries’. Department of Economics and Social Development, United Nation.
Walker, N. & Ghys, P., 2003. ‘Proportion of child mortality attributable to HIV’ presented at
Meeting on empirical Evidence for the Demographic and Socio-Economic Impact of AIDS.
Durban, South Africa, March 26-28.
Wichmann, J. & Voyi, K. V. V., 2006. ’Influence of cooking and heating fuel use on 1-59
Month old Mortality in South Africa’. Maternal Child Health J (2006) 10 :553-561.
World Health Organisation., 2006. ‘Neonatal and Perinatal Mortality’ Country, Regional and
Global Estimates 2006.
89
APPENDIX 1
GHS 2002 QUESTIONNAIRE
90
General Household Survey
RSM / QA
Particulars of the household
2002
Number
Date checked
For office use
PSU number
Response details
Dwelling unit number
Visit no
Date (actual)
Result code
Next visit (planned)
1
Physical identification of the dwelling
unit/household
2
3
4
FINAL RESULT
Telephone number of enumerated household (if any)
T t l
b
f
i th h
Comments and full details of all non-response/unusual circumstances
h ld
Questionnaire no. for this household (for persons no. 01 - 10 = 1, etc.)
Households at the selected dwelling
Household number for this household
RESULT CODES (for response details)
Total number of households at the selected dwelling
Field staff
Interviewer
Number
Interview date
Supervisor
Number
Date checked
1
2
3
4
5
6
7
8
Completed
Non-contact
Refused
Partly complete
No usable information
Vacant dwelling
Listing error
Other
Comment and give full details above
of all non-response
+
+
Questionnaire ID
FLAP
This section covers particulars of each person in the household
The following information must be obtained in respect of every person who normally resides in this household at least four nights a week.
Do not forget babies. If there are more than 10 persons in the household, use a second questionnaire.
Person (respondent) number
Ask who the head (or the acting head) of the household is
A
→ End of questions for this person
Is ...... a male or a female?
1 = MALE
2 = FEMALE
D
How old is ......? (In completed years - In whole numbers)
Less than 1 year = 00.
E
What population group does ....... belong to?
1 = AFRICAN/BLACK
2 = COLOURED
3 = INDIAN/ASIAN
4 = WHITE
5 = OTHER, specify
F
+
03
04
05
06
07
08
09
10
Surname:
Has ...... stayed here (in this household) for at least four
nights on average per week during the last four weeks?
1 = YES
2 = NO
C
02
First name and surname
First name:
Write down first name and surname of each
member of the household, starting with the
head or acting head.
If more than one head or acting head,
take the oldest.
Write sideways if necessary.
B
01
Head/ Acting
head
Is there any other person residing in this household,
than those already mentioned, who is not presently
here?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
YES
→ If “YES”, Go back to A
NO
1
+
+
+
Questionnaire ID
SECTION 1 This section covers particulars of each person in the household
Start from the left (person number 01) and complete section 1 for each person in the household separately.
01
1.1
What is ……’s present marital status?
1 = MARRIED OR LIVING TOGETHER AS HUSBAND AND WIFE
2 = WIDOW/WIDOWER
3 = DIVORCED OR SEPARATED
4 = NEVER MARRIED
1.2.b
→ Go to Q 1.3.a
Does ……’s spouse/partner live in this household?
1 = YES
2 = NO
→ Go to Q 1.3.a
1.2.c
Which person is the spouse/partner of ……?
Give person number
1.3.a
Is …… ‘s father still alive?
1 = YES
2 = NO
3 = Don’t know
+
03
04
05
06
07
08
09
10
What is ……’s relationship to the head of the
household? (I.e. to the person in column 1)
1 = Mark the head/acting head
2 = HUSBAND/WIFE/PARTNER
3 = SON/DAUGHTER/STEPCHILD/ADOPTED CHILD
4 = BROTHER/SISTER
5 = FATHER/MOTHER
6 = GRANDPARENT/GREAT GRANDPARENT
7 = GRANDCHILD/GREAT GRANDCHILD
8 = OTHER RELATIVE (E.G. IN-LAWS OR AUNT/UNCLE)
9 = NON-RELATED PERSONS
1.2.a
02
→ Go to Q 1.4.a
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
8
9
8
9
8
9
8
9
8
9
8
9
8
9
8
9
8
9
8
9
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
2
+
+
01
1.3.b
Is ……’s father part of the household?
1 = YES
2 = NO
1.3.c
Which person is ……’s father?
Give person number
1.4.a
Is …… ‘s mother still alive?
1 = YES
2 = NO
3 = Don’t know
1.4.b
+
Questionnaire ID
→ Go to Q 1.4.a
→ Go to Q 1.5.a
Is ……’s mother part of the household?
1 = YES
2 = NO
→ Go to Q 1.5.a
02
03
04
05
06
07
08
09
10
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1.4.c
Which person is ……’s mother?
Give person number
1.5.a
Can …… read in at least one language?
1 = YES
2 = NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Can …… write in at least one language?
1 = YES
2 = NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1.5.b
1.6.a
In the last seven days, did …… spend at least one hour
fetching water for home use (not for sale)?
1 = YES
2 = NO
1.6.b
+
→ Go to Q 1.7.a
How many hours did …… spend on fetching water in
the last seven days?
3
+
+
01
1.7.a
In the last seven days, did …… spend at least one hour
fetching wood/dung for home use (not for sale)?
1 = YES
2 = NO
1.7.b
1.8
1.9
→ Go to Q 1.8
02
03
04
05
06
07
08
09
10
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
How many hours did …… spend on fetching wood/dung
in the last seven days?
Do you know if there is a welfare office in your area?
1 = YES
2 = NO
3 = DON’T KNOW
1
2
3
Who is the person who usually brings in the most
money into the household?
Give person number and mark a box below
1 = If there is one person who brings in the highest
amount, give the person number of this person and
mark box 1
2 = If two persons or more bring in the same highest
amount, give person number of the oldest of them
and mark box 2
3 = If the respondent does not know, give person
number of the oldest person who brings in money
and mark box 3
4 = If no-one brings in money, give person number of
the oldest person in the household and mark box 4
+
+
Questionnaire ID
1
2
3
4
4
+
+
+
Questionnaire ID
Education
01
1.10
1.11
What is the highest level of education that …… has
completed?
00 = NO SCHOOLING
01 = GRADE R/0
02 = SUB A/GRADE 1
03 = SUB B/GRADE 2
04 = GRADE 3/STANDARD 1
05 = GRADE 4/STANDARD 2
06 = GRADE 5/STANDARD 3
07 = GRADE 6/STANDARD 4
08 = GRADE 7/STANDARD 5
09 = GRADE 8/STANDARD 6/FORM 1
10 = GRADE 9/STANDARD 7/FORM 2
11 = GRADE 10/STANDARD 8/FORM 3
12 = GRADE 11/STANDARD 9/FORM 4
13 = GRADE 12/STANDARD 10/FORM 5/MATRIC
14 = NTC l
15 = NTC II
16 = NTC III
17 = DIPLOMA/CERTIFICATE WITH LESS THAN GRADE 12/STD 10
18 = DIPLOMA/CERTIFICATE WITH GRADE 12/STD 10
19 = DEGREE
20 = POSTGRADUATE DEGREE OR DIPLOMA
21 = OTHER (specify in column)
22 = DON'T KNOW
03
04
05
06
07
08
09
10
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Is …… currently attending school or any other
educational institution?
1 = YES
2 = NO
+
02
→ Go to Q 1.13
5 +
+
01
1.12
+
Questionnaire ID
What is the main reason why …… is currently not
attending school or any other education institution?
01 = TOO OLD/YOUNG
02 = HAS COMPLETED SCHOOL/EDUCATION
03 = SCHOOL/EDUCATION INSTITUTION IS TOO FAR AWAY
04 = NO MONEY FOR FEES
05 = HE/SHE IS WORKING (AT HOME OR JOB)
06 = EDUCATION IS USELESS OR UNINTERESTING
07 = ILLNESS
08 = PREGNANCY
09 = FAILED EXAMS
10 = GOT MARRIED
11 = FAMILY COMMITMENT (CHILD MINDING, ETC.)
12 = OTHER, specify in column underneath
02
03
04
05
06
07
08
09
10
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
01
02
03
04
05
06
07
08
09
10
11
12
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
→ Go to Q 1.19
1.13
1.14
Which of the following educational institutions does
…… attend?
Include distance and correspondence education
1 = Pre-school (including day care, crèche, pre-primary)
2 = School
3 = University
4 = Technikon
5 = College
6 = Adult basic education and training/literacy classes
7 = Other adult educational classes
8 = Other than any of the above
Is it a correspondence/distance educational institution?
1 = YES
2 = NO
+
→ Go to Q 1.16
6 +
+
01
1.15
+
Questionnaire ID
How long does it take …… to get to the
school/educational institution where he/she attends?
1 = LESS THAN 15 MINUTES
2 = 15 - 30 MINUTES
3 = MORE THAN 30 MINUTES
4 = DON’T KNOW
02
03
04
05
06
07
08
09
10
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
01
02
03
04
05
06
07
08
09
10
11
12
13
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Ask for all who are attending school any educational institution
1.16
What is the total amount of tuition fees paid for ….. in a
year?
Do not include the cost of uniforms, books and
other learning materials.
01 = R1 – R100
02 = R101 – R200
03 = R201 – R300
04 = R301 – R500
05 = R501 – R1000
06 = R1001 – R2000
07 = R2001 – R3000
08 = R3001 – R4000
09 = R4001 – R8000
10 = R8001 – R12000
11 = MORE THAN R12000
12 = NONE
13 = DON’T KNOW
1.17
+
This academic year, has …… benefited from any
exemptions and/or bursaries?
1 = YES
2 = NO
3 = DON’T KNOW
7 +
+
01
1.18
+
Questionnaire ID
During the past 12 months, what problems, if any, did
…… experience at the school(or other educational
institution)?
1 = Lack of books
2 = Poor teaching
3 = Lack of teachers
4 = Facilities in bad condition
5 = Fees too high
6 = Classes too large
7 = Other, specify in column
02
03
04
05
06
07
08
09
10
YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
HEALTH
Ask for everyone
1.19
1.20
Is …… covered by a medical aid or medical benefit
scheme or other private health insurance?
1 = YES
2 = NO
3 = DON'T KNOW
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
During the past month, did …… suffer from any
illnesses or injuries?
1 = YES
01
01
01
01
01
01
01
01
01
01
02
02
02
02
02
02
02
02
02
02
2 = NO
+
→ Go to Q 1.29
8 +
+
01
1.21
1.22
What sort of illnesses or injuries did …… suffer from?
Was it ….
01 = Flu or acute respiratory tract infection
02 = Diarrhoea
03 = Severe trauma (e.g. due to violence, motor vehicle
accident, gunshot, assault, beating)
04 = TB or severe cough with blood
05 = Abuse of alcohol or drugs
06 = Depression or mental illness
07 = Diabetes
08 = High or low blood pressure
09 = HIV/AIDS
10 = Other sexually transmitted disease
11 = Other illness or injury
During the past month, did …… consult a health worker
such as a nurse, doctor or traditional healer as a result
of illness or injury?
1 = YES
2 = NO
1.23
+
+
Questionnaire ID
→ Go to Q 1.28
What kind of health worker was it?
If more than one consultation, take the most recent.
1 = NURSE
2 = DOCTOR
3 = MEDICAL SPECIALIST
4 = PHARMACIST/CHEMIST
5 = DENTIST
6 = SPIRITUAL HEALER (CHURCH RELATED)
7 = TRADITIONAL HEALER
8 = ANY OTHER HEALTH CARE PROVIDER
Including psychologist, physiotherapist, chiropractor,
homeopath, optometrist
9 = DON'T KNOW
02
NO
YES
03
NO
YES
04
NO
YES
05
NO
YES
06
NO
YES
07
NO
YES
08
NO
YES
09
NO
YES
10
NO
YES
NO
YES
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
01
02
01
02
01
02
01
02
01
02
01
02
01
02
01
02
01
02
01
02
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
9
9
9
9
9
9
9
9
9
9
9 +
+
01
1.24
+
Questionnaire ID
02
03
04
05
06
07
08
09
10
Where did the consultation take place?
If more than one consultation, ask about the
most recent
one.
1.25
+
Public sector (i.e. government, provincial or community
institution)
01 = HOSPITAL
02 = CLINIC
03 = OTHER IN PUBLIC SECTOR, specify
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
01
02
03
Private sector (including private clinics, surgery, private
hospitals and sangomas)
04 = HOSPITAL
05 = CLINIC
06 = PRIVATE DOCTOR/SPECIALIST
07 = TRADITIONAL HEALER
08 = PHARMACY/CHEMIST
09 = HEALTH FACILITY PROVIDED BY EMPLOYER
10 = ALTERNATIVE MEDICINE, E.G. HOMEOPATHIST
11 = OTHER IN PRIVATE SECTOR, specify
12 = DON'T KNOW
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
04
05
06
07
08
09
10
11
12
What problems, if any, were experienced by …… during
this particular visit to a health worker?
1 = Facilities not clean
2 = Long waiting time
3 = Opening times not convenient
4 = Too expensive
5 = Drugs that were needed, not available
6 = Staff rude or uncaring or turned patient away
7 = Incorrect diagnosis
8 = Other, specify in column
YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
10 +
+
01
1.26
1.27
+
Questionnaire ID
02
03
04
05
06
07
08
09
10
How satisfied was …… with the service he/she received?
1 = Very satisfied
2 = Somewhat satisfied
3 = Neither satisfied nor dissatisfied
4 = Somewhat dissatisfied
5 = Very dissatisfied
6 = DON’T KNOW
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
Did …… have to pay for this service?
1 = YES
2 = NO
3 = DON'T KNOW
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
→ Go to Q 1.29
Ask only if “NO” to Q 1.22
1.28
+
Why did …… not consult any health worker during the
past month?
1 = TOO EXPENSIVE
2 = TOO FAR
3 = NOT NECESSARY
4 = DON’T KNOW
5 = OTHER, specify in column underneath
11 +
+
+
Questionnaire ID
Ask for everyone in the household
Read out: I am now going to ask about disabilities experienced by any persons within the household.
01
1.29
Is …… limited in his/her daily activities, at home, at
work or at school, because of a long-term physical,
sensory, hearing, intellectual, or psychological
condition, lasting six months or more?
1 = YES
2 = NO
1.30
1.31
During the past 12 months, did …… make use of a
welfare office or services?
1 = YES
2 = NO
3 = DON’T KNOW
+
01
02
→ Go to Q1.31
What difficulty or difficulties does …… have? Is it …..
1 = Sight (blind/severe visual limitation)
2 = Hearing (deaf, profoundly hard of hearing)
3 = Communicating (speech impairment)
4 = Physical (e.g. needs wheel chair, crutches or
prosthesis; limb or hand usage limitation)
5 = Intellectual (serious difficulties in learning, mental
retardation)
6 = Emotional (behavioural, psychological problems)
7 = Other, specify in column
→ Go to section 2
02
03
01
02
04
01
02
05
01
02
06
01
02
07
01
02
08
01
02
09
01
02
10
01
02
01
02
YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO YES NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
12 +
+
01
1.32
03
04
05
06
07
08
09
10
a. Social worker
1 = YES
2 = NO
3 = DON’T KNOW
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
b. Social grant
1 = YES
2 = NO
3 = DON’T KNOW
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
3 = DON’T KNOW
+
02
Which of the following services/assistance was …… in
need of?
c. Poverty relief/Job creation project
1 = YES
2 = NO
1.33
+
Questionnaire ID
Ask only if there is a “YES” in any part of Q 1.32
How satisfied was …… with the service/assistance
rendered at the welfare office?
1 = Very satisfied
2 = Somewhat satisfied
3 = Neither satisfied nor dissatisfied
4 = Somewhat dissatisfied
5 = Very dissatisfied
6 = DON’T KNOW
13 +
+
+
Questionnaire ID
SECTION 2This section covers activities of household members aged 15 and above in the last seven days, unemployment and non-economic activities.
Ask for all household members aged 15 and above. It is very important that you try to ask these questions of each person themselves if at all possible.
Read out: Now I am going to ask some questions about activities in the last seven days for each household member aged 15 and above
01
2.0
2.1
Interviewer to answer
Is the person him/herself responding to questions?
1 = YES
2 = NO
In the last seven days, did …… do any of the following
activities, even for only one hour? Show prompt card 2.
02
1
2
03
1
2
YES NO
04
1
2
YES NO
05
1
2
YES NO
06
1
2
YES NO
07
1
2
YES NO
08
1
2
YES NO
09
1
2
YES NO
10
1
2
YES NO
1
2
YES NO
YES NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
c) Do any work as a domestic worker for a wage, salary,
or any payment in kind?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
d) Help unpaid in a family business of any kind?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
e) Do any work on his/her own or the family’s plot, farm,
food garden, cattle post or kraal or help in growing
farm produce or in looking after animals for the
household?
Examples: ploughing, harvesting, looking after livestock.
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
f) Do any construction or major repair work on his/her
own home, plot, cattle post or business or those of
the family?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
g) Catch any fish, prawns, shells, wild animals or other
food for sale or family food?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
h) Beg for money or food in public?
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
a) Run or do any kind of business, big or small, for
himself/herself or with one or more partners?
Examples: Selling things, making things for sale, repairing things,
guarding cars, brewing beer, hairdressing, crèche businesses, taxi or
other transport business, having a legal or medical practice, etc.
b) Do any work for a wage, salary, commission or
any payment in kind (excl. domestic work)?
Examples: a regular job, contract, casual or piece work for pay, work
in exchange for food or housing.
Examples: Help to sell things, make things for sale or exchange, doing
the accounts, cleaning up for the business, etc. Don't count normal
housework.
If “YES” for a person to any part of Question 2.1→ Go to Q 2.3 for that person.
+
If all “NO” for a person, continue with next question.
14 +
+
01
2.2
2.3
If “NO” to all parts of Question 2.1
Even though …… did not do any of these activities in
the last seven days, does he/she have a job, business,
or other economic or farming activity that he/she will
definitely return to?
For agricultural activities, the off season in agriculture is not
a temporary absence.
1 = YES
2 = NO
→Go to Q 2.10
+
Questionnaire ID
1
2
02
1
2
03
1
2
04
1
2
05
1
2
06
1
2
07
1
2
08
1
2
09
1
2
10
1
2
Read out:
You said …… was doing these activities during the
last seven days (or was temporarily absent).
Refer to Q 2.1 or Q 2.2
What kind of work did …… do in his/her main job
during
the last seven days (or usually does, even if he/she was
absent in the last seven days)? Give occupation or job
title.
Work includes all the activities mentioned earlier
Record at least two words: Car sales person, Office
cleaner, Vegetable farmer, Primary school teacher, etc.
For agricultural work on own/family farm/plot, state whether
for own use or for sale mostly.
2.4
What were ……'s main tasks or duties in this job?
Examples: Selling fruit, repairing watches, keeping
accounts, feeding and watering cattle.
CODE BOX FOR OFFICE USE
+
15 +
+
+
Questionnaire ID
01
2.5
What is the name of ……’s place of work?
For government or large organisations, give the name of
the establishment and branch or division: e.g. Education
Dept – Rapele Primary School; ABC Gold Mining,
Maintenance Div.
Write ‘Own house’ or ‘No fixed location’, if relevant.
2.6
What are the main goods and services produced at
……'s place of work? What are its main functions?
Examples: Repairing cars, Selling commercial real estate,
Sell food wholesale to restaurants, Retail clothing shop,
Manufacture electrical appliances, Bar/ restaurant, Primary
Education, Delivering newspapers to homes.
02
03
04
05
06
07
08
09
10
CODE BOX FOR OFFICE USE
+
16 +
+
01
2.7
What is ……’s total salary/pay at his/her main job?
Including overtime, allowances and bonus, before any tax
or deductions.
Give amount in whole figures, without any text or decimals
If “NONE”, “REFUSE” or “DON’T KNOW”→ Go to Q 2.9
2.8
Only if amount given in Q 2.7
Is this
1 = Per week
2 = Per month
3 = Annually
2.9
+
Questionnaire ID
02
03
04
05
06
07
08
09
10
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Only if “NONE”, “REFUSE” or “DON’T KNOW” in Q 2.7
Show the categories. Make sure the respondent points at
the correct income column (weekly, monthly, annually) on
prompt card 3 and mark the applicable code.
Weekly
Monthly
Annually
01
NONE
NONE
NONE
01
01
01
01
01
01
01
01
01
01
02
R1 - R46
R1 - R200
R1 - R2 400
02
02
02
02
02
02
02
02
02
02
03
R47 - R115
R201 - R500
R2 401 - R6 000
03
03
03
03
03
03
03
03
03
03
04
R116 - R231
R501 – R1 000
R6 001 - R12 000
04
04
04
04
04
04
04
04
04
04
05
R232 - R346
R1 001 - R1 500
R12 001 - R18 000
05
05
05
05
05
05
05
05
05
05
06
R347 = R577
R1 501 = R2 500
R18 001 - R30 000
06
06
06
06
06
06
06
06
06
06
07
R578 - R808
R2 501 - R3 500
R30 001 - R42 000
07
07
07
07
07
07
07
07
07
07
08
R809 - R1 039
R3 501 - R4 500
R42 001 - R54 000
08
08
08
08
08
08
08
08
08
08
09
R1 040 - R1 386
R4 501 - R6 000
R54 001 - R72 000
09
09
09
09
09
09
09
09
09
09
10
R1 387 - R1 848
R6 001 - R8 000
R72 001 - R96 000
10
10
10
10
10
10
10
10
10
10
11
R1 849 - R2 540
R8 001 - R11 000
R96 001 - R132 000
11
11
11
11
11
11
11
11
11
11
12
R2 541 - R3 695
R11 001 - R16 000 R132 001 - R192 000
12
12
12
12
12
12
12
12
12
12
13
R3 696 - R6 928
13
13
13
13
13
13
13
13
13
13
R6 929 OR MORE
R16 001 - R30
000
R192 001 - R360 000
14
R360 001 OR MORE
14
14
14
14
14
14
14
14
14
14
15
DON'T KNOW
R30 001 OR MORE
DON'T KNOW
15
15
15
15
15
15
15
15
15
15
16
REFUSE
DON'T KNOW
REFUSE
16
16
16
16
16
16
16
16
16
16
REFUSE
→ Go to Section 3
+
17 +
+
+
Questionnaire ID
The following questions cover unemployment and non-economic activities
Ask for all household members aged 15 and above who did not work and were not absent from work (i.e. for those whose answer on Q 2.2 = 2).
Read out: Now I am going to ask some questions about whether you (……) wanted and were (was) available for any of the types of work mentioned earlier
01
2.10
Why did …… not work during the past seven days?
01 = HAS FOUND A JOB, BUT IS ONLY STARTING AT A DEFINITE
DATE IN THE FUTURE
→ Go to Q 2.14
02 = LACK OF SKILLS OR QUALIFICATIONS FOR AVAILABLE JOBS
03 = SCHOLAR OR STUDENT AND PREFERS NOT TO WORK
04 = HOUSEWIFE/HOMEMAKER AND PREFERS NOT TO WORK
05 = RETIRED AND PREFERS NOT TO SEEK FORMAL WORK
06 = ILLNESS, INVALID, DISABLED OR UNABLE TO WORK
(HANDICAPPED)
07 = TOO YOUNG OR TOO OLD TO WORK
08 = SEASONAL WORKER, E.G. FRUIT PICKER, WOOL-SHEARER
09 = CANNOT FIND SUITABLE WORK (SALARY, LOCATION OF
WORK OR CONDITIONS NOT SATISFACTORY)
10 = CONTRACT WORKER, E.G. MINE WORKER RESTING
02
03
04
05
06
07
08
09
10
01
01
01
01
01
01
01
01
01
01
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
02
03
04
05
06
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
07
08
09
10
10
10
10
10
10
10
10
10
10
11
12
11
12
11
12
11
12
11
12
11
12
11
12
11
12
11
12
11
12
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
ACCORDING TO CONTRACT
11 = RECENTLY RETRENCHED
12 = OTHER REASON
2.11
If a suitable job is offered, will …… accept it?
1 = YES
2 = NO
3 = DON'T KNOW
2.12
→ Go to Q 2.14
How soon can …… start work?
1 = WITHIN A WEEK
2 = WITHIN TWO WEEKS
3 = WITHIN FOUR WEEKS
4 = LATER THAN FOUR WEEKS FROM NOW
5 = DON'T KNOW
+
18 +
+
01
2.13
+
Questionnaire ID
During the past four weeks, has …… taken any
action
1 = to look for any kind of work
02
03
04
05
06
07
08
09
10
YES NO
YES NO
YES NO
YES NO
YES NO
YES NO
YES NO
YES NO
YES NO
YES NO
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
2 = to start any kind of business
Ask for everyone who has come to Question 2.10 (all persons unemployed or not economically active)
2.14
Has …… ever worked before?
1 = YES
2 = NO
2.15
+
→ Go to Q 2.16
How long ago was it since …… last worked?
01 = 1 WEEK - LESS THAN 1 MONTH
02 = 1 MONTH - LESS THAN 2 MONTHS
03 = 2 MONTHS - LESS THAN 3 MONTHS
04 = 3 MONTHS - LESS THAN 4 MONTHS
05 = 4 MONTHS - LESS THAN 5 MONTHS
06 = 5 MONTHS - LESS THAN 6 MONTHS
07 = 6 MONTHS - LESS THAN 1 YEAR
08 = 1 YEAR - LESS THAN 2 YEARS
09 = 2 YEARS - LESS THAN 3 YEARS
10 = 3 YEARS OR MORE
11 = DON'T KNOW
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
01
02
03
04
05
06
07
08
09
10
11
19 +
+
01
2.16
+
Questionnaire ID
How does …… support him/herself?
1 = Did odd jobs during the past seven days
2 = Supported by persons in the household
3 = Supported by persons not in the household
4 = Supported by charity, church, welfare, etc.
5 = Unemployment Insurance Fund (UIF)
6 = Savings or money previously earned
7 = Old age or disability pension
8 = Other sources, e.g. bursary, study loan, specify in
YES NO
1
1
1
1
1
1
1
1
02
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
03
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
04
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
05
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
06
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
07
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
08
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
09
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
10
2
2
2
2
2
2
2
2
YES NO
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
column
If “YES” to response category 1
→ Go back to Q 2.1 for that person
+
20 +
+
+
Questionnaire ID
SECTION 3
This section covers information regarding children ever born
The following information must be obtained in respect of every woman aged between 12 and 50 years. For each woman record the total number of children ever born alive.
Include all children born alive,(i.e all those who are still living, whether or not they live in the household, and those who are dead). Do not include stillbirths and children adopted
by the mother. Start with the last born and strictly follow the birth order. Do not forget babies.
If there is no woman in the household, go to section 4.
Read out: I am now going to ask regarding mothers in this household
3.0.1
3.0.2
3.0.3
Is there any woman in this household aged between 12 and 50 years, who
has ever given birth?
1 = Yes
2 = No
End of this section. Go to Section 4
How many women in this household aged between 12 and 50 years have
ever given birth?
What are the names of the women who have ever given birth?
1
2
Person
number
1.Name of the first woman……..…………………………Give person number
2.Name of the second woman….……………………..…Give person number
3.Name of the third woman…….…………………………Give person number
4.Name of the fourth woman…….…………….…………Give person number
5.Name of the fifth woman……..…………………………Give person number
Remember: If there are more than 3 women aged between 12 and 50 years in the household, who have ever given birth, you will need another
questionnaire.
21
+
+
+
+
Questionnaire ID
Read out: I am now going to ask each woman questions regarding all the children she has ever had.
Record the name of the woman and her personal number, as indicated on the flap. Record births by each woman on a separate form.
First name of woman………………………………………………Person number
Male
3.1.1
3.1.2
3.1.3
Female
Total
How many children (live births) have you ever given
birth to?
How many of your children are still alive?
How many children (live births) have you had in the
past 12 months
Read out: Now, I am going to ask you questions regarding each of the live births you have ever had, starting with the most recent
Child number
If there are more than 10 children born to one woman,
continue on the next form and change the child numbers
(ie, 01=11 and so on) Record twins on separate columns
3.1.4
First name and surname
(Write down the first name of each child
born alive, starting with the last born.
Strictly follow the birth order)
3.1.5
Is …… still alive ?
1 = YES
2 = No
01
02
03
04
05
06
07
08
09
10
Start with
the last born
First name:
3.1.6
3.1.7
Go to 3.1.10
If alive, Is ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
How old is ......? (In completed years - In whole numbers)
Less than 1 year = 00.
22
+
+
+
01
3.1.8
+
Questionnaire ID
What was …… ‘s date of birth?
(Write down the year, month and day of birth in
the space provided for each child. The year must
be a 4 digit number).
02
03
04
05
06
07
08
09
10
YYYY
MM
DD
3.1.9
3.1.10
Is …… currently a member of this household?
1 = YES
2 = NO
Go to 3.1.13
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
If dead, Was ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3.1.11
How old was ...... when he/she died?
(In completed years - In whole numbers)
Less than 1 year = 00.
3.1.12
When did …… ‘s death occur?
(Write down the date of death as indicated)
YYYY
MM
DD
Ask for all children ever born to the woman
3.1.13
3.1.14
3.1.15
Where was …… born?
1 = IN A HOSPITAL
2 = AT A CLINIC
3 = ELSEWHERE
Was the birth of …… registered?
End of section 3 for this child
1 = YES
2 = NO
Why was the birth of …… not registered?
1 = FAR DISTANCE
2 = LACK OF KNOWLEDGE
3 = DOES NOT SEEM IMPORTANT
23
+
+
+
+
Questionnaire ID
Read out: I am now going to ask each woman questions regarding all the children she has ever had.
Record the name of the woman and her personal number, as indicated on the flap. Record births by each woman on a separate form.
First name of woman………………………………………………Person number
Male
3.2.1
3.2.2
3.2.3
Female
Total
How many children (live births) have you ever given
birth to?
How many of your children are still alive?
How many children (live births) have you had in the
past 12 months
Read out: Now, I am going to ask you questions regarding each of the live births you have ever had, starting with the most recent
Child number
If there are more than 10 children born to one woman,
continue on the next form and change the child numbers
(ie, 01=11 and so on) Record twins on separate columns
3.2.4
First name and surname
(Write down the first name of each child
born alive, starting with the last born.
Strictly follow the birth order)
3.2.5
Is …… still alive ?
1 = YES
2 = No
01
02
03
04
05
06
07
08
09
10
Start with
the last born
First name:
3.2.6
3.2.7
Go to 3.2.10
If alive, Is ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
How old is ......? (In completed years - In whole numbers)
Less than 1 year = 00.
24
+
+
+
01
3.2.8
+
Questionnaire ID
What was …… ‘s date of birth?
(Write down the year, month and day of birth in
the space provided for each child. The year must
be a 4 digit number).
02
03
04
05
06
07
08
09
10
YYYY
MM
DD
3.2.9
3.2.10
Is …… currently a member of this household?
1 = YES
2 = NO
Go to 3.2.13
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
If dead, Was ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3.2.11
How old was ...... when he/she died?
(In completed years - In whole numbers)
Less than 1 year = 00.
3.2.12
When did …… ‘s death occur?
(Write down the date of death as indicated)
YYYY
MM
DD
Ask for all children ever born to the woman
3.2.13
3.2.14
3.2.15
Where was …… born?
1 = IN A HOSPITAL
2 = AT A CLINIC
3 = ELSEWHERE
Was the birth of …… registered?
End of section 3 for this child
1 = YES
2 = NO
Why was the birth of …… not registered?
1 = FAR DISTANCE
2 = LACK OF KNOWLEDGE
3 = DOES NOT SEEM IMPORTANT
25
+
+
+
+
Questionnaire ID
Read out: I am now going to ask each woman questions regarding all the children she has ever had.
Record the name of the woman and her personal number, as indicated on the flap. Record births by each woman on a separate form.
First name of woman………………………………………………Person number
Male
3.3.1
3.3.2
3.3.3
Female
Total
How many children (live births) have you ever given
birth to?
How many of your children are still alive?
How many children (live births) have you had in the
past 12 months
Read out: Now, I am going to ask you questions regarding each of the live births you have ever had, starting with the most recent
Child number
If there are more than 10 children born to one woman,
continue on the next form and change the child numbers
(ie, 01=11 and so on) Record twins on separate columns
3.3.4
First name and surname
(Write down the first name of each child
born alive, starting with the last born.
Strictly follow the birth order)
3.3.5
Is …… still alive ?
1 = YES
2 = No
01
02
03
04
05
06
07
08
09
10
Start with
the last born
First name:
3.3.6
3.3.7
Go to 3.3.10
If alive, Is ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
How old is ......? (In completed years - In whole numbers)
Less than 1 year = 00.
26
+
+
+
01
3.3.8
+
Questionnaire ID
What was …… ‘s date of birth?
(Write down the year, month and day of birth in
the space provided for each child. The year must
be a 4 digit number).
02
03
04
05
06
07
08
09
10
YYYY
MM
DD
3.3.9
3.3.10
Is …… currently a member of this household?
1 = YES
2 = NO
Go to 3.3.13
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
If dead, Was ...... a male or a female?
1 = MALE
2 = FEMALE
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
2
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3.3.11
How old was ...... when he/she died?
(In completed years - In whole numbers)
Less than 1 year = 00.
3.3.12
When did …… ‘s death occur?
(Write down the date of death as indicated)
YYYY
MM
DD
Ask for all children ever born to the woman
3.3.13
3.3.14
3.3.15
Where was …… born?
1 = IN A HOSPITAL
2 = AT A CLINIC
3 = ELSEWHERE
Was the birth of …… registered?
End of section 3 for this child
1 = YES
2 = NO
Why was the birth of …… not registered?
1 = FAR DISTANCE
2 = LACK OF KNOWLEDGE
3 = DOES NOT SEEM IMPORTANT
27
+
+
+
+
Questionnaire ID
SECTION 4 This section covers information regarding the household.
Ask a responsible adult in the household
4.1
Indicate the type of main dwelling and other
dwelling that the household occupies?
01 = DWELLING/HOUSE OR BRICK STRUCTURE ON A
Main
dwelling
01
Other
dwelling
01
4.2
Thinking back five years ago, what type of
dwelling/dwellings did this household occupy?
01 = DWELLING/HOUSE OR BRICK STRUCTURE ON A
02 = TRADITIONAL DWELLING/HUT/STRUCTURE MADE
02
02
03
03
04
04
05
05
06
06
07
07
08
08
09
10
09
10
09 = ROOM/FLATLET
10 = CARAVAN/TENT
11
11
11 = OTHER, specify
12 = HOUSEHOLD DID NOT EXIST
OF
03 = FLAT OR APARTMENT IN A BLOCK OF FLATS
04 = TOWN/CLUSTER/SEMI-DETACHED HOUSE
(Simplex, Duplex or Triplex)
05 = UNIT IN RETIREMENT VILLAGE
06 = DWELLING/HOUSE/FLAT/ROOM IN BACKYARD
07 = INFORMAL DWELLING/SHACK IN BACKYARD
08 = INFORMAL DWELLING/SHACK NOT IN BACKYARD,
E.G. IN AN INFORMAL/SQUATTER SETTLEMENT OR ON
FARM
09 = ROOM/FLATLET
10 = CARAVAN/TENT
11 = OTHER, specify
02
02
03
03
04
04
05
05
06
06
07
07
08
08
09
10
09
10
11
11
12
12
OF
TRADITIONAL MATERIALS
TRADITIONAL MATERIALS
Other
dwelling
01
SEPARATE STAND OR YARD OR ON FARM
SEPARATE STAND OR YARD OR ON FARM
02 = TRADITIONAL DWELLING/HUT/STRUCTURE MADE
Main
dwelling
01
03 = FLAT OR APARTMENT IN A BLOCK OF FLATS
04 = TOWN/CLUSTER/SEMI-DETACHED HOUSE
(Simplex, Duplex or Triplex)
05 = UNIT IN RETIREMENT VILLAGE
06 = DWELLING/HOUSE/FLAT/ROOM IN BACKYARD
07 = INFORMAL DWELLING/SHACK IN BACKYARD
08 = INFORMAL DWELLING/SHACK NOT IN BACKYARD,
E.G. IN AN INFORMAL/SQUATTER SETTLEMENT OR ON
FARM
28
+
+
+
4.3
4.4
+
Questionnaire ID
What is the main material used for the roof and
the walls of the main dwelling?
Mark one code in each column.
01 = BRICKS
02 = CEMENT BLOCK/CONCRETE
03 = CORRUGATED IRON/ZINC
04 = WOOD
05 = PLASTIC
06 = CARDBOARD
07 = MIXTURE OF MUD AND CEMENT
08 = WATTLE AND DAUB
09 = TILE
10 = MUD
11 = THATCHING
12 = ASBESTOS
13 = OTHER, specify
14 = NOT APPLICABLE
Roof
In what condition are the roof and the walls of the
main dwelling?
1 = Very weak
2 = Weak
3 = Needs minor repairs
4 = Good
5 = Very good
Roof
Walls
01
01
02
02
03
03
04
05
04
05
06
06
07
08
07
08
09
09
10
11
10
11
12
12
13
14
13
14
Walls
1
1
2
3
2
3
4
4
5
5
4.5
Is the dwelling ….
1 = Owned and fully paid off
2 = Owned, but not yet fully paid off (e.g. with a mortgage)
3 = Rented
4 = Occupied rent-free as part of employment contract of
family member
5 = Occupied rent-free not as part of employment contract
of family member
6 = Other, specify
4.6
What is the total number of rooms in the dwelling(s) that the
household occupies?
Give the total number of rooms, including living rooms, bedrooms
and kitchens, but excluding bathrooms and toilets.
4.7
Did any member of this household receive a
government
housing subsidy, such as RDP housing subsidy,
to obtain this dwelling or any other dwelling?
Do not include housing subsidies for government
employees.
1 = YES
2 = NO
3 = DON’T KNOW
1
2
3
4
5
6
1
2
3
29
+
+
+
4.8
What is the household’s main source of water?
Mark one code only
01 = PIPED (TAP) WATER IN DWELLING
02 = PIPED (TAP) WATER ON SITE OR IN YARD
03 = NEIGHBOUR’S TAP
04 = BOREHOLE ON SITE
05 = RAIN-WATER TANK ON SITE
06 = PUBLIC TAP
07 = WATER-CARRIER/TANKER
08 = BOREHOLE OFF SITE/COMMUNAL
09 = FLOWING WATER/STREAM/RIVER
10 = DAM/POOL/STAGNANT WATER
11 = WELL
12 = SPRING
13 = OTHER, specify
Ask only if Q 4.8 = 01, 02, 03 or 06 (e.g. tap/piped water),
otherwise go to Q 4.14
01
02
→ Go to Q 4.10
4.11
03
04
05
06
07
08
09
4.12
10
11
12
13
4.10
How long does it take members of this household to get
to the water source?
1 = 0 - 14 MIN
2 = 15 - 29 MIN
3 = 30 - 44 MIN
4 = 45 - 59 MIN
5 = 60 MIN OR MORE
The water from the main source
1 = Is it safe to drink?
2 = Is it clear?
3 = Does it taste good?
4 = Is it free from odours?
How often do you get interruptions in your piped
water supply?
1 = DAILY
2 = WEEKLY
3 = MONTHLY
4 = 6 MONTHLY
5 = YEARLY
6 = ALMOST NEVER
Ask if water is not in dwelling, yard or site, otherwise go to Q 4.10.
4.9
+
Questionnaire ID
What normally causes the interruption?
1 = BURST PIPES
2 = PUMP NOT WORKING
3 = GENERAL MAINTENANCE
4 = NOT ENOUGH WATER IN THE SYSTEM (DEMAND TOO HIGH)
5 = WATER ONLY DELIVERED AT FIXED TIMES
6 = NON-PAYMENT FOR SERVICES
(CUT OFF)
7 = VANDALISM
8 = OTHER, specify
9 = DON’T KNOW
1
2
→ Go to Q 4.14
→ Go to Q 4.14
1
2
3
4
5
6
1
2
3
4
5
6
7
8
9
3
4.13
4
5
YES NO
1
2
1
2
1
1
2
2
The last time it happened, when was the problem
rectified?
1 = THE SAME DAY
2 = DURING THE SAME WEEK
3 = DURING THE SAME MONTH
4 = LONGER THAN MONTH, specify
1
2
3
4
30
+
+
+
4.14
4.15
+
Questionnaire ID
Does this household have a connection to the MAINS
electricity supply?
1 = YES
2 = NO
What is the main source of
energy/fuel for this household?
01 = ELECTRICITY FROM MAINS
02 = ELECTRICITY FROM GENERATOR
03 = GAS
04 = PARAFFIN
05 = WOOD
06 = COAL
07 = CANDLES
08 = ANIMAL DUNG
09 = SOLAR ENERGY
10 = OTHER, specify
11 = NONE
Cooking
4.16
1
2
Heating
Lighting
01
01
01
02
03
02
03
02
03
04
04
04
05
06
05
06
07
08
08
09
10
09
10
11
11
4.17
Thinking back five years ago, did this household have
a connection to the MAINS electricity supply, then?
1 = YES
2 = NO
3 = HOUSEHOLD DID NOT EXIST
4 = DON’T KNOW
What type of toilet facility is
available for this household?
Mark only one, the main toilet
1 = FLUSH TOILET CONNECTED TO
In
dwelling On site
11
12
1
2
3
4
Off site
13
A PUBLIC SEWAGE SYSTEM
2 = FLUSH TOILET CONNECTED TO A
21
22
23
32
33
42
43
52
53
62
63
73
SEPTIC TANK
07
09
10
11
3 = CHEMICAL TOILET
4 = PIT LATRINE WITH VENTILATION
PIPE
5 = PIT LATRINE WITHOUT
VENTILATION PIPE
6 = BUCKET TOILET
7 = NONE
→ Go to Q 4.20
31
+
+
+
+
Questionnaire ID
Ask if toilet is “ON SITE” or “OFF SITE”. Otherwise Go to Q 4.19
4.18
How far is the nearest toilet facility to which the household
has access?
1 = LESS THAN 2 MINUTES (LESS THAN 200M)
2 = 2 MINUTES BUT LESS THAN 5 MINUTES (200M - 500M)
3 = MORE THAN 5 MINUTES (MORE THAN 500M)
4.21
1
2
3
4.22
Ask if answer to Q 4.17 is “BUCKET TOILET”. Otherwise Go to Q 4.20
4.19
How frequently is it removed?
1 = ONCE A WEEK OR MORE OFTEN
2 = ABOUT ONCE A FORTNIGHT
3 = ABOUT ONCE A MONTH
4 = LESS OFTEN THAN ONCE A MONTH
1
2
3
4
How is the refuse or rubbish of this household taken care
of?
1 = REMOVED BY LOCAL AUTHORITY AT LEAST ONCE A WEEK
2 = REMOVED BY LOCAL AUTHORITY LESS OFTEN THAN ONCE A
Is there a cellular telephone available to this household for
regular use?
1 = YES
2 = NO
1
2
1
2
Ask if answer is “No” to both Q 4.21 and Q 4.22. Otherwise Go to
Q4.25
4.23
Ask for all households
4.20
Does this household have a landline telephone in the
dwelling?
1 = YES
2 = NO
1
2
How far does it take from here, to the nearest accessible
telephone, using your usual means of transport?
1 = 0 - 14 MIN
2 = 15 - 29 MIN
3 = 30 – 44 MIN
4 = 45 – 59 MIN
5 = 60 MIN OR MORE
1
2
3
4
5
WEEK
3 = REMOVED BY COMMUNITY MEMBERS AT LEAST ONCE A WEEK
4 = REMOVED BY COMMUNITY MEMBERS LESS OFTEN THAN ONCE
3
4
A WEEK
5 = COMMUNAL REFUSE DUMP/COMMUNAL CONTAINER
6 = OWN REFUSE DUMP
7 = NO RUBBISH REMOVAL
8 = OTHER, specify
5
6
7
8
Ask for all households
4.24
Thinking back five years ago, did this household have a
landline telephone in the dwelling then?
1 = YES
2 = NO
3 = HOUSEHOLD DID NOT EXIST
4 = DON’T KNOW
1
2
3
4
32
+
+
+
4.25
+
Questionnaire ID
How does this household receive most of its mail/post?
1 = DELIVERED TO THE DWELLING
2 = DELIVERED TO A POST BOX/PRIVATE BAG
3 = THROUGH FRIEND OR NEIGHBOUR
4 = THROUGH SHOP
5 = THROUGH SCHOOL
6 = THROUGH WORKPLACE
7 = THROUGH AUTHORITY
8 = DO NOT RECEIVE MAIL
9 = OTHER, specify
4.26
1
2
3
4
Facility
5
6
7
8
9
What means of transport are usually, or would usually be used by
members of this household to get to the nearest of each of these
facilities?
If more than one means of transport, take the one used over the longest
distance
ON
TAXI
BUS
TRAIN
(PUBLIC)
FOOT
OWN
TRANSPORT
OTHER,
specify
below
a Food market
1
2
3
4
5
6
b Public transport
1
2
3
4
5
6
c
1
2
3
4
5
6
d Primary school
1
2
3
4
5
6
e Secondary school
1
2
3
4
5
6
f
Pre-Primary/Pre-school
centre
Clinic
1
2
3
4
5
6
g Hospital
1
2
3
4
5
6
h Post office or post office
agent
1
2
3
4
5
6
i
1
2
3
4
5
6
Welfare office
If “Other” in Q 4.26, specify:
33
+
+
+
+
Questionnaire ID
4.27 How long in minutes does it take or would it take, from here to reach
the nearest ………using the usual means of transport?
Facility
4.28
0 - 14
15 - 29 30 - 44 45 - 59 60 MIN
DON’T
MIN
MIN
KNOW
MIN
MIN
OR
MORE
2 = NO
4.29
a
Food market
1
2
3
4
5
6
b
Public transport
1
2
3
4
5
6
c
Pre-Primary/Pre-school
centre
1
2
3
4
5
6
d
Primary school
1
2
3
4
5
6
e
Secondary school
1
2
3
4
5
6
f
Clinic
1
2
3
4
5
6
g
Hospital
1
2
3
4
5
6
h
Post office or post office
agent
1
2
3
4
5
6
i
Welfare office
1
4.30
2
3
4
5
6
Does this household have access to land that is, or could
be, used for agricultural purposes?
1 = YES
→ Go to Q 4.32
How many hectares of land, for agricultural purposes, if any,
does the household have access to?
Exclude communal grazing land
1 = LESS THAN 5.000 M2 (5.000 m2 is approximately one soccer
field)
2
2
2 = 5.000M - 9.999M
3 = 1 BUT LESS THAN 5 HA
4 = 5 BUT LESS THAN 10 HA
5 = 10 BUT LESS THAN 20 HA
6 = 20 HA OR MORE
7 = DON’T KNOW
On what basis does the household have access to the
land?
1 = OWNS THE LAND
2 = RENTS THE LAND
3 = SHARECROPPING
4 = TRIBAL AUTHORITY
5 = OTHER, specify
6 = DON’T KNOW
1
2
1
2
3
4
5
6
7
1
2
3
4
5
6
34
+
+
+
4.31
What farming activities, if any, take place on the land? Is
it……..?
1 = Field crops
2 = Horticulture
3 = Livestock
3 = Poultry
5 = Orchards
6 = Other, (Specify)……………………………………
4.37
YES NO
1
2
1
2
1
2
1
2
1
2
1
2
4.33
Did the household receive a government land grant to
obtain a plot of land for residence or for farming?
1 = YES
2 = NO
3 = DON’T KNOW
How many chicken, ducks, etc. are currently owned by
the household?
4.39
Does the household own any of the following?
01 = Car or truck
02 = Motorcycle
03 = Tractor
04 = Plough
05 = Television
06 = Bicycle
07 = Radio
08 = Bed
09 = Watch or clock
10 = Books
1
→ Go to Q 4.35
2
2 = NO
4.34
How many head of cattle and other large livestock are
currently owned by the household?
4.35
Does the household own any sheep, goats and other
medium size animals?
1 = YES
2 = NO
4.36
2
3
→ Go to Q 4.37
1
2
4.40
1
→ Go to Q 4.39
4.38
1
Does the household own any cattle or other large
livestock?
1 = YES
Does the household own any poultry such as chickens,
ducks, etc (but excluding chicks)
1 = YES
2 = NO
Ask for all households
4.32
+
Questionnaire ID
In the past 12 months, did any adult in this household go
hungry because there wasn’t enough food?
1 = NEVER
2 = SELDOM
3 = SOMETIMES
4 = OFTEN
5 = ALWAYS
2
YES NO
1
2
1
1
2
2
1
2
1
2
1
1
2
2
1
2
1
1
2
2
1
2
3
4
5
How many sheep, goats and other medium size animals
are currently owned by the household?
35
+
+
+
4.41
4.42
4.43
4.44
+
Questionnaire ID
In the past 12 months, did any child (17 years or
younger) in this household go hungry because there
wasn’t enough food?
1 = NEVER
2 = SELDOM
3 = SOMETIMES
4 = OFTEN
5 = ALWAYS
4.45
Does any member of this household receive any of the
following Welfare Grants?
1 = Old age pension
2 = Disability grant
3 = Child support grant
4 = Care dependency grant
5 = Foster care grant
6 = Grant in aid
7 = Social relief
What is the main source of income for this household?
1 = SALARIES AND/OR WAGES
2 = REMITTANCES
3 = PENSIONS AND GRANTS
4 = SALES OF FARM PRODUCTS
5 = OTHER NON-FARM INCOME
6 = NO INCOME
Include everything that the household and its
members spent
money on, including food, clothing, transport, rent
and rates, alcohol and tobacco, school fees,
entertainment and any other expenses.
1
2
3
4
5
In the past 12 months, is there any young person, aged
5 - 17, who has left this household to live on the streets?
1 = YES
2 = NO
3 = DON’T KNOW
What was the total household expenditure in the last
month?
1
2
3
YES
1
1
NO
2
2
1
2
1
2
1
1
2
2
1
2
01 = R 0 – R 399
02 = R 400 – R 799
03 = R 800 – R 1 199
04 = R 1 200 – R 1 799
05 = R 1 800 – R 2 499
06 = R 2 500 – R 4 999
07 = R 5 000 – R 9 999
08 = 10 000 OR MORE
09 = DON’T KNOW
10 = REFUSE
01
02
03
04
05
06
07
08
09
10
1
2
3
4
5
6
36
+
+
+
Please read as you show the prompt card
Now, I am now going to ask you questions regarding your physical safety
and that of other members of your household. In some of the questions I will
show you a prompt card, which has eleven choices “00” to “10” describing
the level of your feelings about safety or satisfaction. Kindly point out the
level that best describes your feelings.
4.46
4.47
+
Questionnaire ID
Regarding your own safety, how safe do you feel if you
are walking in this area at night?
1 = VERY SAFE
2 = RATHER SAFE
3 = RATHER UNSAFE
4 = VERY UNSAFE
Thinking about your physical safety in your
neighbourhood, how safe do you and other members of
the household feel living here?
(Ask respondent to point out the answer on a prompt card)
01 = 10 (COMPLETELY SAFE)
02 = 09
03 = 08
04 = 07
05 = 06
06 = 05
07 = 04
08 = 03
09 = 02
10 = 01
11 = 00 (COMPLETELY UNSAFE)
4.48
1
2
3
4
4.49
01
02
03
04
05
06
07
08
09
10
11
During the past 12 months, have you or any member of
this household been subjected to the following
incidents?
Have you or any member of this household ……
1 = had things stolen
2 = been harassed or threatened by a household
member
3 = been harassed or threatened by someone outside
the household
4 = been sexually molested by a household member
5 = been sexually molested by someone out side the
household
6 = been beaten up or hurt by a household member
7 = been beaten up or hurt by someone outside the
household
Taking everything into account, how satisfied are you
with public safety these days?
(Ask respondent to point out the answer on a prompt card)
01 = 10 (COMPLETELY SATISFIED)
02 = 09
03 = 08
04 = 07
05 = 06
06 = 05
07 = 04
08 = 03
09 = 02
10 = 01
11 = 00 (COMPLETELY DISSATISFIED)
YES
1
NO
2
1
2
1
2
1
1
2
2
1
2
1
2
01
02
03
04
05
06
07
08
09
10
11
37
+
+
+
Please read out
Now, in the following questions, I am going to ask you whether you agree
with several statements dealing with general problems of life. Please tell me
if you completely agree, somewhat agree, somewhat disagree or strongly
disagree with the statement.
4.50
4.51
4.52
+
Questionnaire ID
Would you agree with the statement that, you can’t do
much to change most of the difficulties we face today?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
Would you agree with the statement that, you often feel
lonely?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
Would you agree with the statement that, you don’t really
enjoy your work?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
1
4.53
4.54
2
3
4
1
2
3
4
4.55
Would you agree with the statement that, life has become
so complicated today that you almost can’t find your
way?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
Would you agree with the statement that, you are very
optimistic about the future?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
Would you agree with the statement that, in order to get
ahead nowadays you are forced to do things that are not
correct?
1 = COMPLETELY AGREE
2 = SOMEWHAT AGREE
3 = SOMEWHAT DISAGREE
4 = STRONGLY DISAGREE
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
38
+
+
+
4.56
Questionnaire ID
Please tell me how satisfied you are with your life in
general.
(Ask respondent to point out the answer on a prompt card)
01 = 10 (COMPLETELY SATISFIED)
02 = 09
03 = 08
04 = 07
05 = 06
06 = 05
07 = 04
08 = 03
09 = 02
10 = 01
11 = 00 (COMPLETELY DISSATISFIED)
+
01
02
03
04
05
06
07
08
09
10
11
End of interview.
Thank the respondent!
Interviewer to answer questions on next page.
39
+
+
+
Questionnaire ID
4.57
Indicate the column number of the person who answered
the questions in Section 5
4.58
In what language was the main part of the interview
conducted?
01 = AFRIKAANS
02 = ENGLISH
03 = ISINDEBELE/SOUTH NDEBELE/NORTH NDEBELE
04 = ISIXHOSA/XHOSA
05 = ISIZULU/ZULU
06 = SEPEDI/NORTHERN SOTHO
07 = SESOTHO/SOUTHERN SOTHO/SOTHO
08 = SETSWANA/TSWANA
09 = SISWATI/SWAZI
10 = TSHIVENDA/VENDA
11 = XITSONGA/TSONGA
12 = OTHER, specify
+
01
02
03
04
05
06
07
08
09
10
11
12
40
+
+
APPENDIX 2
EXTRACT FROM OHS 1997 QUESTIONNAIRE
91
SECTION 2
This section covers information regarding births.
This section must be completed for all women who have ever given birth
A separate form must be completed for each woman
Interviewer: Please read the instructions on this page
before you start with Question 2.1.
Record all live births starting with the first born. Do not include
still births and children adopted by the mother. Remember to include
children who have died and children who are not currently part of
the household.
First name of woman (a): ....................................................................................................................... Respondent No: .................................................
2.1 How many children (live births) have you ever given birth to?
2.2 How many of your children are still living?
2.3 How many children (live births) have you had in the past 12 months?
Now let us talk about each of your children
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
2.12
2.13
List of children
(from the eldest to the
youngest )
Interviewer:
Record twins
on separate lines and
mark with a bracket
Name of child
(optional)
BIRTH ORDER
1
2
3
4
5
6
7
8
Is/Was the child
a boy or a girl?
All children
Date of birth
All children
Where was the
child born?
In what year,
month and day
was the child
born?
All children If not registered
Was the birth
Why?
1= far distance
registered?
2= lack of
knowledge
3= Does not seem
important
All children
Is the child
still alive?
If alive:
Is the child currently
living with this
household?
If alive:
How old is he/she
Interviewer:
Record age in
completed years
less than
1 year = 0
If dead
How old was the
Child when he/she
died?
Interviewer:
Record age in
completed years
less than 1 year = 0
Age in years
Age at death in years
No
Reasons for not
Registering
Yes
No
Yes
No
2
In a in a
Else
hos- clinic where Yes
pital
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
2
1
2
Boy
Girl
1
Year Mon Day
11
Section 9
Household information
This section covers information regarding the
dwellings, services and perceived quality of life of the
household.
9.1 How many dwellings does this household occupy on
this particular site? By household we mean a person or
a group of persons who live together at least four
nights a week at the same address, eat together and
share resources.
9.2 Indicate the type of main dwelling and other dwelling(s) that the
household occupies?
You can circle more than one code for the other dwelling(s) if the
household occupies more than 2 dwellings
Type of dwelling
Main
dwelling
Other
dwelling
Dwelling/house or brick structure on a separate stand or
yard
1
1
Traditional dwelling/hut/structure made of traditional
materials
2
2
Less than one dwelling (sharing a dwelling
with other households)
1
Flat or apartment in a block of flats
3
3
4
4
One dwelling
2
Town/cluster/semi-detached house (simplex, duplex or
triplex)
Two dwellings
3
5
5
Three dwellings
4
Unit in retirement village
Dwelling/house/flat/room in backyard
Informal dwelling/shack, in backyard
More than three dwellings
5
6
7
8
6
7
8
9
9
10
11
10
11
Informal dwelling/shack NOT in back yard, e.g. in an
informal/squatter settlement
Room /flatlet
Caravan/tent
Other (specify)
0317-E
1
40
9.3 What is the MAIN material used for the roof and the
walls of the (main) dwelling?(Circle one code in each
column)
Material
Roof
Bricks
Walls
01
Cement block/concrete
02
02
Corrugated iron/zinc
03
03
Wood
04
04
Plastic
05
05
Cardboard
06
06
Mixture of mud and cement
07
07
Wattle and daub
08
08
Tile
09
Mud
9.4 What is the total number of rooms in the dwelling(s) that the
household occupies?
Total number of rooms including living
rooms, bedrooms and kitchens (excluding
bathrooms and toilets)
9.5 Is this dwelling (main dwelling, if more than one) owned by the
household (even if not yet fully paid) ?
Yes (Go to question 9.11)
1
No (Continue)
2
IF THE HOUSEHOLD DOES NOT OWN THE DWELLING(S),
ANSWER QUESTIONS 9.6 TO 9.10
9.6 If the dwelling(s) is/are not owned by the household, [Ask] Are
you required to pay rent for the dwelling(s)?
Yes
10
Thatching
11
11
Asbestos
12
12
(continue)
No (Go to question 9.10 )
1
2
9.7 What was the rent that was charged last month?
R.........................
9.8 Is this rent subsidised?
Yes
1
No
2
Do not know
3
0317-E
2
41
9.9 Do you rent this dwelling with or without furniture ?
9.11 Since this dwelling is owned by the household, [Ask] Is this
ownership:
With furniture
1
Full title (including free-hold and lease-hold)
1
Without furniture
2
Sectional title
2
Do not know
3
9.10 Is the dwelling owned by:
Employer (eg Eskom, AECI, Transnet,
Farmer)
1
If ‘Sectional title’ what was the levy paid last month?
R.........................
Government (national, provincial or local)
2
Charity organisation
3
Private owner
4
Yes
1
Other (specify)...................................................
5
No
2
9.12 Is this household presently paying off a bond on the dwelling(s)?
If ‘Yes’ how much did you pay last month?
R......................
IF THE HOUSEHOLD DOES OWN THE DWELLING(S), ANSWER
QUESTIONS 9.11 TO 9.12)
42
ASK EVERY HOUSEHOLD
Services available for the dwelling:
9.13 What is this household’s main source of water? (Circle only one
code)
9.14 If the water source is outside the dwelling(s) [Ask] How far is the
water source from the dwelling(s)?
Piped (tap) water, in dwelling
1
Less than 100 m
1
Piped (tap) water, on site or in yard
2
100 m - less than 200 m
2
Public tap
3
200 m - less than 500 m
3
Water-carrier/tanker
4
500 m - less than 1 km
4
Borehole on site
5
1 km or more
5
Borehole: off site/communal
6
Not applicable (water on site)
6
Rain-water tank on site
7
Flowing water/stream
8
Dam/pool/stagnant water
9
Well
10
Spring
11
Other (specify).
12
9.15 Does the household have to pay for its water?
Always
1
Sometimes
2
Never
3
43
9.16 If the household has to pay for its water [Ask],
How much does the household pay?
Less than R50
R50 or more
Do not know
1
2
3
9.18 From where does the household get its wood? Indicate the main
source. (Circle one code)
ASK EVERY HOUSEHOLD
9.17 What is the main source of energy/fuel for this household? (Circle
one code for each source)
Energy/fuel
source
Cooking
Heating
Lighting
Electricity
1
1
1
Gas
2
2
2
Paraffin
3
3
3
Wood
4
4
Coal
5
5
Candles
6
Animal dung
7
7
Solar Energy
8
8
8
.............
.............
............
Other (Specify)
IF WOOD IS THE MAIN SOURCE OF FUEL FOR THE
HOUSEHOLD, (FOR EITHER COOKING OR HEATING OR BOTH,
ANSWER QUESTIONS 9.18 TO 9.22)
Woodlot
1
Commercial plantations
2
Natural forest
3
Veld
4
Home yard trees
5
Merchants
6
9.19 Is the wood obtained enough for normal household
purpose?
Always
1
Mostly yes
2
Mostly no
3
No
4
44
ASK EVERY HOUSEHOLD
9.20 Does the household have to pay for the wood?
Always
1
Sometimes
2
Never
3
9.21 Does the household have to fetch wood?
Yes
1
No
2
9.22 How far is the wood if it has to be fetched?
Less than 100m
1
100m - less than 200m
2
200m - less than 500m
3
500m - less than 1km
4
1 km or more
5
Sanitation
9.23 What type of toilet facility is available for this household? (Circle
only one code)
Toilet facility
In
dwelling
On
site
Off
site
1. Flush toilet
1
1
1
2. Chemical toilet
2
2
3. Pit latrine with
ventilation (VIP)
4. Other pit latrine
3
3
4
4
5.Bucket toilet
6. None
5
5
6
5.Other
7
9.24 Is the toilet facility shared with other households?
Yes
1
No
2
45
9.25 If the toilet is not in the dwelling [Ask] How far is the nearest
toilet facility to which the household has access?
Less than 25m
1
25m- less than 50m
2
50m- less than 100m
3
100m or more
4
9.26 If the facility is a bucket toilet [Ask] How frequently is it
removed?
ASK EVERY HOUSEHOLD
Refuse disposal:
9.27 How is the refuse or rubbish of this household disposed of? (Circle
only one code)
Removed by local authority at least once a
week
1
Removed by local authority less often
2
Removed by community members at least
once a week
3
Once a week or more often
1
Removed by community members less often
4
About once a fortnight
2
Communal refuse dump/communal container
5
About once a month
3
Own refuse dump
6
Less often than once a month
4
No rubbish removal
7
Other (Specify) .............................................
46
Telecommunication
ASK EVERY HOUSEHOLD
9.28 Does anyone in this household have a cellular telephone?
Let us talk about your safety and perceived quality of life
Yes
1
No
2
9.29 Is there a telephone in this dwelling?(Please DO NOT include
cellular telephones)
Yes
1
No
2
9.31 How safe do you feel living in the neighbourhood where
you live?
Very safe
1
Rather safe
2
Rather unsafe
3
Very unsafe
4
9.32 How safe do you feel in the dwelling where you live?
9.30 If there is no telephone in the dwelling(s) [Ask]
How many minutes do you have to travel to the nearest telephone you
can use ( by your usual means of transport)?
Very safe
1
Rather safe
2
0 - 5 minutes
1
Rather unsafe
3
6 - 15 minutes
2
Very unsafe
4
16 - 30 minutes
3
31 - 60 minutes
4
9.33 Do you feel safer, about the same, or less safe, than you felt a
year ago?
1 - 2 hours
5
Safer
1
Over 2 hours
6
The same
2
Less safe
3
47
9.34 During the past 12 months, has this household experienced any
burglaries, robberies or housebreaking ?
Yes
1
No
2
9.35 During the past 12 months, has anyone been murdered while he/she
was a member of this household?
9.37 In the past year, was there ever a time when you could not afford to
feed the children in the household?
Yes
1
No
2
Not applicable (no children)
3
9.38 Taking everything into account, how satisfied is this household
with the way it lives these days?
Yes
1
Very satisfied
1
No
2
Satisfied
2
Neither satisfied nor dissatisfied
3
Dissatisfied
4
Very dissatisfied
5
ASK EVERY HOUSEHOLD
9.36 Do you have any street lighting where you live?
Yes
1
No
2
48
ASK EVERY HOUSEHOLD
9.39 Compared to one year ago, how would you say things
are for this household?
Things are better
1
Things are about the same
2
Things are worse
3
9.40 How much money did this household spend in total,
on all items (including food, clothing, housing,
transport, medical care, etc), during the past month?
9.43 If there were any unusual cash purchases during
the past month or past year, [Ask] How much did
the household spend on them all together?
Past month
R.....................
Past year (please do not include
purchases for the past month)
R.....................
ASK EVERY HOUSEHOLD
9.44 If anyone in this household gets ill or injured and decides to seek
medical help, where do they usually go first? (Circle only one code)
R.........................
9.41 How much did the household spend on food during
the past month?
Public
Sector
R.........................
9.42 Were there any unusual cash purchases (e.g. car, fridge, furniture,
etc.) during the past month and/or the past year?
Hospital
1
Clinic
2
Other
(specify)..................................
3
Hospital
Private
Sector
4
Clinic
5
Past month
Past year
Yes
1
1
Private doctor/specialist
6
No
2
2
Traditional healer
Others(specify)………………
7
8
49
ASK EVERY HOUSEHOLD
9.45 How far is the hospital/clinic/doctor/traditional healers
where the household members usually go?
(Circle only one code)
9.47 What means of transport do the members of this household mainly
use to get to the health facility? - mainly = longest distance (Circle
only one code).
Ambulance
1
Less than 1 km
1
Own transport (car, minibus, etc.)
2
1km - less than 5km
2
Train
3
5km - less than 10km
3
Taxi
4
10km - less than 15km
4
Bus (public)
5
15km or more
5
On foot
6
Other transport (specify)
7
9.46 How long does it usually take to get there?
Less than 15 minutes
1
15 minutes - less than 30 minutes
2
30 minutes - less than 1 hour
3
1 hour - less than 2 hours
4
2 hours or more
5
9.48 Where is this health care person/facility where household members
usually go? (State place name, magisterial district and province).
Town/place
name
Magisterial
district
Province (New)
50
ASK EVERY HOUSEHOLD
9.49 How far is the nearest social welfare service point?
Less than 1km
1
1km - less than 5km
2
5km or more
3
Do not know
4
9.50. Please indicate the respondent number of the person
who answered the questions in this section
You have come to the end of the interview for this household. Thank
the respondent for his/her co-operation.
51
52

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