Microbial risk assessment and its implications for risk Therese Westrell

Microbial risk assessment and its implications for risk Therese Westrell
Linköping Studies in Arts and Science • 304
Microbial risk assessment and its implications for risk
management in urban water systems
Therese Westrell
Linköping Studies in Arts and Science
In the Faculty of Arts and Science at Linköping University research is pursued
and research training is conducted within seven broad problem areas known as
themes, in Swedish tema. These are: Child Studies, Communication Studies, Gender
Studies, Health and Society, Food Studies, Technology and Social Change and Water and
Environmental Studies. Jointly they publish the series Linköping Studies in Arts
and Science.
Distributed by:
Department of Water and Environmental Studies
Linköping University
S-581 83 Linköping
Sweden
Therese Westrell
Microbial risk assessment and its implications for risk management in urban water
systems
Edition 1:1
ISBN 91-85295-98-1
ISSN 0282-9800
© Therese Westrell
Printed by UniTryck, Linköping, 2004.
ii
List of papers
This thesis is based on the following papers that will be referred to in the text by
their Roman numerals.
I.
Westrell, T., Bergstedt, O., Heinicke, G. and Kärrman, E. (2002). A systems
analysis comparing drinking water systems - central physical-chemical treatment
and local membrane filtration. Water Science and Technology: Water Supply 2(2): 1118.
II. Westrell, T., Bergstedt, O., Stenström, T. A. and Ashbolt, N. J. (2003). A
theoretical approach to assess microbial risks due to failures in drinking water
systems. International Journal of Environmental Health Research 13(2): 181-197.
III. Westrell, T., Teunis, P., van den Berg, H., Lodder, W., Ketelaars, H.,
Stenström, T. A. and de Roda Husman, A. M. Short and long term fluctuations
of norovirus concentrations in surface water and their implications for public
health. (Submitted to Water Research).
IV. Westrell, T., Andersson, O. and Stenström, T. A. Drinking water consumption
patterns in Sweden. (Submitted to Journal of Water and Health).
V. Schönning, C., Westrell, T., Stenström, T. A., Arnbjerg-Nielsen, K., Hasling,
A. B., Hansen, L. and Carlsen, A. Microbial risk assessment of local handling
and reuse of human faeces. (Submitted to Journal of Water and Health).
VI. Westrell, T., Schönning, C., Stenström, T. A. and Ashbolt, N. J. (2004).
QMRA (quantitative microbial risk assessment) and HACCP (hazard analysis
critical control points) for management of pathogens in wastewater and sewage
sludge treatment and reuse. Water Science and Technology 50(2): 23-30.
Papers I, II and IV are reproduced with the kind permission from the publishers
(IWA publishing http://www.iwaponline.com/ and Taylor and Francis Journals
http://www.tandf.co.uk/journals).
iii
Table of contents
1
Introduction.............................................................................................. 1
1.1
Environmental transmission of infectious diseases.......................................................3
1.2
What is risk and how is it estimated?........................................................................7
1.3
Microbial risk assessment (MRA) ............................................................................8
2
Aims, Rationale and Approaches ............................................................ 11
2.1
Aims ......................................................................................................................11
2.2
Rationale ................................................................................................................11
2.3
Approaches.............................................................................................................13
3
Quantitative microbial risk assessment (QMRA) ...................................17
3.1
Hazard identification ..............................................................................................17
3.2
Exposure assessment...............................................................................................23
3.3
Dose-response assessment.........................................................................................33
3.4
Risk characterisation...............................................................................................36
4
Epidemiological concerns .......................................................................41
4.1
Susceptibility and immunity.....................................................................................41
4.2
Sensitive subpopulations ..........................................................................................41
4.3
Secondary transmission............................................................................................42
4.4
Dynamic modelling..................................................................................................42
4.5
Health indices .........................................................................................................43
5
Risk management of water systems ....................................................... 45
5.1
Guidelines and regulations.......................................................................................45
5.2
Hazard analysis and critical control points (HACCP) ...........................................45
5.3
Tolerable risk .........................................................................................................47
6
Short summary of results ........................................................................ 49
6.1
Paper I – Centralised versus decentralised drinking water treatment.........................49
6.2
Paper II – Failures in drinking water treatment and distribution ............................50
6.3
Paper III – Norovirus fluctuations in surface water .................................................50
6.4
Paper IV – Drinking water consumption patterns ..................................................51
6.5
Paper V – Risks associated with local handling of faeces .........................................52
6.6
Paper VI –HACCP for safe handling and reuse of wastewater and sludge..............52
7
Discussion .............................................................................................. 55
7.1
Use of appropriate data and models.........................................................................55
7.2
The importance of different agents ............................................................................56
7.3
Fluctuating pathogen concentrations in surface waters ...............................................57
7.4
Use of epidemiology in QMRA...............................................................................59
7.5
Microbial health risks in decentralised systems .........................................................60
7.6
Outbreaks versus sporadic cases...............................................................................61
v
7.7
7.8
7.9
Sensitive sub-populations......................................................................................... 62
Application of HACCP in the management of wastewater and biosolids treatment and
reuse ....................................................................................................................... 63
Risk communication – experts and users................................................................. 64
8
Conclusions ............................................................................................ 67
9
Acknowledgements ................................................................................ 69
10
References ...............................................................................................71
vi
Microbial risk assessment in urban water systems
1 Introduction
Sustainable development was defined in the ‘Bruntland Report’ (WCED, 1987) as
“development that meets the needs of the present without compromising the ability
of future generations to meet their own needs”. Within the water sector
sustainability could be described as the ability to plan and manage water resources in
such a way that they can be sustained for use by future generations.
All nine Millennium Development Goals set by UN member states in 2000 can be
related to water and sanitation, as described in the report from the Joint Monitoring
Programme for Water and Sanitation by the World Health Organization (WHO)
and the United Nations Children’s Fund (UNICEF) (2004). Ensuring
environmental sustainability, for example, demand adequate treatment and disposal
of wastewaters to contribute to better ecosystem conservation and less pressure on
scarce freshwater resources. Reduced child mortality can be achieved by improved
sanitation and drinking water sources. The goal is to halve the proportion of people
that are without sustainable access to safe drinking water and basic sanitation by
2015. While the progress in drinking water is good so far, with 83 percent coverage
today, the action on sanitation is slow in most developing regions. An estimated 2.6
billion people were still without improved sanitation facilities in 2002 (WHO and
UNICEF, 2004).
Today many efforts towards urban water sustainability are being made, for example
by reusing treated wastewater for watering of golf courses, irrigation in agriculture,
etc. However, more sustainable options require water recycling and demand
management. One example is NeWater in Singapore where advance-treated
wastewater will be returned to raw water reservoirs in what is called indirect potable
use (Guendert, 2004). Reuse of treated wastewater for toilet flushing and garden
irrigation is practised for example in Tokyo and Kobe, Japan (Ogoshi et al., 2001)
and in Sydney, Australia (Anderson, 1996). In order to sustain urban water supply
for future generations planning for future population growth and change must be
attended to, which will influence the choice of alternative water sources.
It has become clear that sanitation and sanitary systems should not only safely
dispose of human residuals, but also provide the option of reusing nutrients in
agriculture. Simultaneously, processing of human waste should minimise risks to the
human population and the natural environment. Within the Swedish research
program Sustainable Urban Water Management Program (hereafter called Urban
Water) the objective is to evaluate water and wastewater systems adapted to urban
environments, taking into account various stakeholders and interactions (Figure 11).
1
Therese Westrell
The aim of the research in Urban Water is to develop support for strategic decisions
on future sustainable urban water systems in Sweden. The approach consists of a
common conceptual framework, a number of sustainability aspects or groups of
criteria, indicators and assessment tools.
FIGURE 1-1. The conceptual framework of Urban Water
with its components: users, organisation and technology and
the five sustainability aspects used to evaluate systems (Urban
Water, 2002)
The urban water system is more than just the technical system. Users and
organisations (e.g. water companies, municipal boards, authorities) form the system
(Figure 1-1) and the interactions need to be studied, where five sustainability aspects
are recognised as central, namely:
•
•
•
•
•
Health and hygiene
Environment and resource use
Technology and function
Economy
Socio-culture
These aspects must be taken into consideration when assessing urban water systems
with respect to sustainable development even though some aspects may be more
decisive in each specific application. The criterion for the health and hygiene aspect
is ‘risk of infection’ and the method chosen for evaluation of this criterion is
quantitative microbial risk assessment, QMRA. The method of QMRA will be
developed into a toolbox for use in comparison of water and wastewater systems
(Ashbolt et al., 2004) within Urban Water.
2
Microbial risk assessment in urban water systems
This thesis deals with how to assess hygienic risks in urban water systems as a
means to evaluate the future sustainability of the systems, also in comparison with
other sustainability aspects. The work is mainly focused on Sweden however the
methodology is also applicable to other regions of the world.
1.1
Environmental transmission of infectious diseases
Infectious diseases are transmitted from one person to another and may include
various environmental pathways (Figure 1-2). The pathogens excreted in the faeces
of an infected person will, in a conventional system, end up in sewage. In Sweden
the wastewater from all urban residents (7.7 out of 9 million inhabitants) is treated
in municipal wastewater treatment plants (Svenskt Vatten, 2000). Although
treatment normally reduces the levels of microorganisms by 90-99.9%, given the
potentially high numbers in sewage, substantial loads of pathogens may still remain.
When treated wastewater is discharged into receiving waters the pathogens can be
transmitted to humans via waters used for recreation or food production. If the
water source is used for drinking water production, municipal water treatment must
be sufficient to prevent disease transmission to the consumer.
FIGURE 1-2. Circulation of pathogens in the environment
Many pathogens are concentrated in the sewage sludge. When sludge is used in
agriculture as a fertiliser, transmission of pathogens via crops could occur if the
treatment or storage time has not been sufficient. Some pathogens can be
transmitted between animals and humans, so called zoonoses, and can be found in
both domestic and wild animals (Wahlström et al., 2003; Hutchison et al., 2004).
Heavy rains or snowmelt can cause run-off from pastures and agricultural land
3
Therese Westrell
fertilised with animal slurry, manure or sludge, which can contribute substantial
loads of pathogens to water courses (Hansen and Ongerth, 1991; Kistemann et al.,
2002; Ferguson et al., 2003). Heavy rains can also cause combined sewer overflows
(CSO), which means that wastewater mixed with stormwater is discharged,
untreated or partially-treated, to the recipient.
Other diffuse sources of microbial pollution to watercourses are also common, for
example leaking septic tanks, waterfowl etc. In some places crops are irrigated with
wastewater and transmission of pathogens to humans (and animals) can occur
directly or via crops, aerosols and potentially via groundwater. Transmission of
pathogens via the environment can clearly result in infections and diseases and also
secondary person-to-person transmission can occur. In this way the circulation of
pathogens in the environment will continue. The design of our water and
wastewater systems will affect how and where this transmission can occur and how
it can be restricted.
1.1.1
Disease outbreaks associated with drinking water
The largest waterborne outbreak of disease in recent time for a developed country
occurred in Milwaukee, USA, in 1993 where an estimated 403 000 people fell ill
with gastroenteritis caused by the parasitic protozoa Cryptosporidium (Mac Kenzie et
al., 1994). (The microbial agents are further described in Section 3.1.) About 4 400
people were hospitalised and an uncertain number of people died as a result of the
outbreak (Kramer et al., 1996). The reason for the outbreak was deterioration in
water quality and decreased effectiveness of the coagulation-filtration process in the
waterworks. Although the treated water had turbidity levels more than four times
the highest recorded during ten years, it still fulfilled all requested water quality
standards at that time (Kramer et al., 1996). Another very severe outbreak occurred
in Walkerton, Canada, in 2000 where 2 300 people fell ill and seven died from
exposure to the bacteria enterohaemorrhagic Escherichia coli (EHEC) and
Campylobacter in the drinking water (Hrudey et al., 2003). The likely cause of the
outbreak was surface water contamination of one of the groundwater wells after
heavy rains accompanied by flooding.
Sweden has a long history of surveillance of waterborne disease outbreaks. Every
year a few outbreaks are reported, resulting in a few hundred up to several thousand
disease cases (Andersson and Bohan, 2001). The two largest outbreaks occurred in
1988 and 1995. In the first of these, approximately 11 000 people became ill due to
a chlorination failure in the water treatment plant. The etiological agent, i.e. the
microorganism that caused the outbreak, could not be established. In 1995
approximately 10 000 people fell ill with Campylobacter when a change in pipeline
introduced stagnant raw water to the distribution network (Andersson and Bohan,
2001).
4
Microbial risk assessment in urban water systems
More than 70% of Swedish waterborne disease outbreaks are due to unknown
agents (Andersson and Bohan, 2001). The most commonly identified agents
between 1980 and 1999 were Campylobacter and the protozoa Giardia lamblia. During
the last few years the number of waterborne outbreaks involving noroviruses has
increased nationally (Andersson and Bohan, 2001; Carrique-Mas et al., 2003; Nygård
et al., 2003), in other Nordic countries, such as Finland (Kukkula et al., 1997;
Kukkula et al., 1999) and Norway (Nygård et al., 2004b), and also internationally
(Lopman et al., 2003; Blackburn et al., 2004). Surface water systems have been
responsible for the largest waterborne outbreaks. Nonetheless most outbreaks occur
in groundwater systems where the most common cause is contamination of the
water source through (surface) wastewater infiltration (Andersson and Bohan,
2001).
1.1.2
Disease outbreaks associated with recreational waters
In the USA 65 outbreaks in recreational waters occurred during 2001-2002 (Yoder et
al., 2004) affecting a total of approximately 2 500 persons. One third occurred in
fresh waters. The major agents involved in the outbreaks were Cryptosporidium,
noroviruses, EHEC, Giardia and Shigella. In 25% of the outbreaks in fresh waters
the etiologic agent was unknown. In Europe the same agents as in the USA are
thought to be involved in most recreational waterborne outbreaks.
Outbreaks with Escherichia coli O157:H7 and other Shiga toxin-producing E. coli have
been associated with recreational swimming in lakes (Keene et al., 1994; Ackman et
al., 1997; Barwick et al., 2000; McCarthy et al., 2001; Feldman et al., 2002; Samadpour
et al., 2002; Bruce et al., 2003). These outbreaks have mainly affected children and
the attack rates have been higher when swallowing water and submerging the head.
The most likely source has often been infected bathers or run-off from cattle
grounds.
The most recent outbreak associated with recreational waters in Sweden occurred in
the summer of 2004 when more than 200 people fell ill from swimming in lakes in
the Gothenburg area. The causative agent was found to be norovirus. The lake
where most people got ill serves as the water source for about half of the city. The
performance of the associated drinking water system was evaluated in Papers I and
II. Other norovirus outbreaks in recreational waters have been reported by Hoebe et
al. (2004) and Maunula et al. (2004).
1.1.3
Illness associated with the reuse of wastewater and sludge
Several epidemiological studies have revealed an increased risk for parasitic
infestations and other enteric diseases associated with raw wastewater reuse in
agricultural irrigation (Katzenelson et al., 1976; Fattal et al., 1986; Cifuentes, 1998;
5
Therese Westrell
Srikanth and Naik, 2004). Melloul et al. (2002) showed that the incidence of
protozoan infections and infections with Salmonella were over represented amongst
children living in areas with wastewater irrigation compared to control areas (72%
compared to 45% and 21% compared to 1%, respectively). Blumenthal et al. (2001)
noticed in a similar study an increased risk of parasitic helminth infections.
Treatment of wastewater, e.g. in storage lagoons, however seems to be efficient in
reducing the transmission of pathogens (Shuval, 1991; Blumenthal et al., 2001).
Several foodborne outbreaks of disease have been associated with the irrigation of
crops with sewage-impacted water (Colley, 1996; Hardy, 1999; Doller et al., 2002).
Amahmid et al. (1999) investigated the contamination of vegetables after raw sewage
irrigation and found high levels of pathogens, e.g. 254 cysts of Giardia and 2.7 ova of
the parasitic roundworm Ascaris per kilogram of coriander. Similarly in Israel,
Armon et al. (2002) demonstrated the presence of Cryptosporidium in soil and crops
irrigated with wastewater effluents.
There have not been any recorded outbreaks or evidence of transmission of
pathogens via wastewater irrigation or sludge application to agricultural land in
Sweden (Carlander, 2002). Carlander (2002) however, detected significant numbers
of faecal indicator bacteria in groundwater following wastewater irrigation on energy
crops, which could imply that transmission of pathogens could occur via this
pathway. Furthermore, in Norway, supermarket samples were shown to have
Cryptosporidium oocysts in 8% of the sprout samples and Giardia cysts in 2% of the
samples (Robertson et al., 2002).
1.1.4
Epidemiological limitations
Epidemiological tools are often not sensitive enough to detect a few cases arising
from exposure to pathogens transmitted via the environment (Eisenberg et al.,
2002). As shown in Figure 1-3 infectious diseases may be endemic in the
population, i.e. a few cases are always present. Large outbreaks are more frequently
detected since they draw considerable attention and outbreak investigations can thus
be initiated, which can establish the link between the disease and a certain source.
For example, the Milwaukee outbreak was only recognised after widespread absence
among hospital employees, students, and school teachers, increased numbers of
emergency room visits for diarrhoeal illness, and a shortage of anti-diarrhoeal drugs
(Kramer et al., 1996). The etiological agent and the waterborne nature of the
outbreak were not identified until at least two weeks after its onset. Not until then
was a boil-order advisory issued. Minor outbreaks or sporadic cases are unlikely to
be reported. Few data are available for evaluation of endemic transmission, and
specifically designed blinded randomised studies have given contradictory results,
proposing that less than 5% (Hellard et al., 2001; Sinclair, 2003b) or up to 40%
(Payment et al., 1997) of community gastrointestinal illness may be waterborne.
6
Microbial risk assessment in urban water systems
Quantitative microbial risk assessment (QMRA) can here serve a purpose for
estimating infection risks from low exposure to hazards transmitted via the
environment.
Number of cases
Detected outbreak
Threshold of detection
Undetected outbreak
Sporadic cases
Endemic level
Time
FIGURE 1-3. Difference between endemic background rates,
sporadic cases and outbreaks. Adapted from Haas et al. (1996).
1.2
What is risk and how is it estimated?
Risk is the likelihood and consequence that something with a negative impact will
occur. The ‘agent’ that causes an adverse effect is a hazard. When considering the
hazardous agent Salmonella, for example, a measure of risk is the probability of
getting disease symptoms after exposure to a certain dose of that hazard. Lidskog et
al. (1997) describe the scientific idea about risk as containing: a cause – effect
relationship, negative consequences, probability of occurring and ability to affect.
Risk incorporates the probability that an event will occur with the effect it will have
on the society and the environment, and also in which socio-political context it
takes place (Cutter, 1993).
There are different classifications of risk. One is the separation between dreaded
risk, such as nuclear power plant accidents, and more everyday, known risk that
most people do not experience as a real danger, e.g. accidents with bikes and
lawnmowers. Another is the separation between risks of low probability but large
negative outcome and risks of high probability but less negative effect. A general
consensus is that voluntary risks are much more accepted than non-voluntary risks,
i.e. people are willing to accept higher risks when they have made the choices
themselves compared to risks that they have not chosen and cannot affect (Covello,
1983).
How risks are dealt with within the society is formulated within the area of risk
analysis, which includes three components: risk assessment, risk management and
7
Therese Westrell
risk communication (Haas et al., 1999). These are highly interrelated and should be
worked through together for a successful risk analysis.
Risk assessment is defined as the qualitative or quantitative characterisation and
estimation of potential adverse health effects associated with exposure of
individuals or populations to hazards (here microbial agents).
Risk management is the process for controlling risks, weighing alternatives and
selecting appropriate action, accounting for risk assessment, values,
engineering, economics, and legal and political issues.
Risk communication is the communication of risks to managers, stakeholders,
public officials, and the public. It includes public perception and the ability to
exchange information.
Risk assessment is further divided into four subsequent steps (CAC, 1999):
• hazard identification, in which the human health effects of the particular hazard
are described;
• exposure assessment, which determines the size and nature of the exposed
population and the pathways, amount and duration of the exposure;
• dose-response assessment, which characterises the relationship between
administered dose and incidence of health effect from both human and
animal studies; and
• risk characterisation, which integrates the information from the previous steps
in order to estimate the magnitude of the public health problem and to
evaluate variability and uncertainty.
1.3
Microbial risk assessment (MRA)
The first field subjected to risk assessment was that of the risks of chemicals to
human health. Microorganisms however differ from chemicals in many ways and
the concept has therefore been further developed for assessing microbial hazards.
Important characteristics for microorganisms are that they are affected by their
environment to a high degree and under unfavourable conditions can be inactivated
or die, while under favourable conditions some may multiply. Their transport in the
environment differs between different groups of microorganisms. The response in
humans and animals after ingestion of pathogenic microorganisms varies widely due
to many factors, for example strain or species of the microorganisms, health of
humans or animals, prior exposure (immunity), intake together with food etc. The
risks from microorganisms therefore have to be addressed specifically but can later
be integrated in a combined or overall assessment.
8
Microbial risk assessment in urban water systems
Haas et al. (1999) define quantitative microbial risk assessment (QMRA) as the
application of principles of risk assessment to estimate the consequences from a
planned or actual exposure to infectious microorganisms. Risk assessments have
also been developed for describing the public health consequences of exposure to
pathogens from drinking water, based on its initial use within the food and chemical
sectors. QMRA is today applied to establishing standards, guidelines and other
recommendations regarding drinking water and consumer health (Rose and Gerba,
1991b; Macler and Regli, 1993; Eisenberg et al., 2002). It has a central role in the
new drinking water guidelines from the WHO for assessment of the
accomplishment of established health targets and for the evaluation of Water Safety
Plans (WHO, 2004). In the latter, it is used to support decisions regarding barriers
and treatments necessary to safeguard public health in water supply systems.
One of the most influential QMRA studies was undertaken by Regli et al. (1991).
They described the quantitative assessment of risk from microorganisms in drinking
water, the problems associated with such an analysis as well as the monitoring
required to demonstrate that risk levels are met. They also showed how QMRA can
be used to determine the level of treatment necessary to ensure that consumers
receive a finished drinking water with risks of less than one infection per 10 000
people a year from Giardia and enteric viruses in drinking water, a benchmark set up
by the US Environmental Protection Agency (USEPA).
Other researchers have assessed the infectious risks in drinking water from viruses
(Haas et al., 1993; Gerba et al., 1996b; Crabtree et al., 1997) and protozoa (Haas et al.,
1996; Teunis et al., 1997; Gale, 2000; Medema et al., 2003; Pouillot et al., 2004).
Microbial risk assessment has both been used to qualitatively (Parkin et al., 2003)
and quantitatively (Ashbolt et al., 1997; Soller et al., 2003) assess the health risks of
recreational swimming and it is incorporated in the WHO Guidelines for Safe
Recreational Waters (WHO, 2003).
The potential adverse health effects associated with the reuse of treated wastewater
and sludge is poorly documented (Stenström and Carlander, 1999). QMRA is an
appropriate tool for estimating the associated risks and important exposure
pathways. This is especially valuable when implementing new reuse strategies. For
example, QMRA has been undertaken to assess the health risks of using reclaimed
water for irrigation in urban areas, e.g. parks and golf courses (Asano and Sakaji,
1990; Rose and Gerba, 1991a; Tanaka et al., 1998; Jolis et al., 1999), use of treated
sewage sludge (often referred to as biosolids) in agriculture (Gale, 2003; Eisenberg et
al., 2004) and irrigation of crops with raw (Shuval et al., 1997) and treated
wastewater (Petterson et al., 2001). In Sweden QMRA have been performed on
source-separating sanitary systems, namely the use of urine as fertiliser in agriculture
(Höglund et al., 2002) and local greywater treatment (Ottoson and Stenström, 2003).
9
Therese Westrell
Many variables in a QMRA are subjected to regional differences. The incidence of
diseases in the human and animal population is known to differ between countries,
resulting in differences in the occurrence and concentrations of pathogens in
surface waters, wastewaters etc. The survival of pathogens in the environment is
highly affected by climatic factors, e.g. temperature and solar irradiation, and hence
differs between tropical and temperate regions. The treatment processes used also
differ, e.g. chlorination is commonly practised on wastewater effluents in the USA
while it is seldom used in Sweden; ozonation can occur in several steps in drinking
water treatment in the Netherlands while it is still unusual in Swedish waterworks.
QMRA are often focused on a specific pathogen or pathogen group and only
consider one exposure pathway. A holistic approach is needed in order to assess the
impact on public health of a whole water and wastewater system and in order make
comparisons of different systems.
10
Microbial risk assessment in urban water systems
2 Aims, Rationale and Approaches
2.1
Aims
The general aim of this thesis was to investigate health risks from infectious
microorganisms transmitted via urban water and wastewater systems. This was
undertaken by developing models for microbial risk assessment with particular
emphasis on exposure assessment. The specific research questions were:
• Is decentralised drinking water treatment feasible and microbiologically safe
from a sustainability point-of-view? (Paper I)
• What impacts do failures in drinking water treatment and distribution have
on the health of tap water consumers? (Paper II)
• What are the concentrations of noroviruses in surfaces waters, do they
fluctuate and if so, how can such fluctuations be predicted? (Paper III)
• What does tap water consumption look like in Sweden and does it differ
between groups in the population? (Paper IV)
• What infection risks are associated with local handling and reuse of human
faeces? (Paper V)
• How can hygienic risks in handling and reuse of wastewater and sewage
sludge be controlled? (Paper VI)
2.2
Rationale
Paper I
Today the possibility of using different water qualities for different purposes in a
household is discussed and sometimes applied (McGann, 2004). Since only about 10
of the 200 litres of water used per person and day (Svenskt Vatten, 2000) are used
for drinking and food, it may be possible to supply lower quality water for other
uses. As mentioned, reuse of treated wastewater has been realised in e.g. Singapore,
Australia and Japan, and in China decentralised systems with membrane treatment
have been built in order to improve the drinking water quality for the residents (Ma
et al., 1998). The health risks however have not been assessed in relation to
conventional distribution.
11
Therese Westrell
Paper II
Waterborne outbreaks of disease are often associated with different types of failure
(Stenström et al., 1994). Teunis et al. (1997) concluded that the frequency of failure
in treatment processes rather than the average removal would determine the risks of
infection with protozoa from drinking water. Payment et al. (1997) proposed that
contamination in the distribution system was responsible for a major part of
waterborne disease. The aim therefore was to assess the type and frequency of
failures in drinking water treatment and distribution and their possible health
impacts on the population. As mentioned in Section 1.2, most of the waterborne
outbreaks in Sweden, where an etiological agent was identified, were caused by
Campylobacter. No other quantitative assessment of the risks posed by this organism
in drinking water has however been published.
Paper III
The numbers of waterborne outbreaks of disease associated with noroviruses
appear to have increased in recent years. In order to assess the risk of transmission
of noroviruses via drinking water and recreational waters it is essential to know their
occurrence and concentration in surface waters or recycled wastewaters. Since these
viruses cannot be cultured in vitro their detection is dependant on molecular
methods, which only result in presence/absence or semi-quantitative real-time PCR
data. In order to ensure safe waters for recreation and drinking water production,
risk managers are in urgent need of tools to predict the likely occurrence of peak
norovirus concentrations in surface waters.
Paper IV
A significant correlation between the risk of becoming ill and the quantity of water
consumed have been reported from investigations of waterborne outbreaks (Maurer
and Sturchler, 2000; Carrique-Mas et al., 2003). In QMRA it is essential to know the
amount of unheated tap water ingested on a daily basis in order to establish the dose
of potential pathogens reaching consumers. Water consumption varies between
individuals and demographic variables may also be important. A stochastic
distribution representing the whole population or certain groups of the population
is therefore warranted for risk assessments. Only one such study has been published
until now, based on data dating from the end of the 1970’s (Roseberry and
Burmaster, 1992). More recent and country-specific data are therefore required in
order to refine the QMRA models.
Paper V
An increased interest in nutrient recycling has resulted in source-separating systems
being implemented at ‘eco-villages’ in Sweden. In some of these the households are
responsible for management of urine and faeces. This arrangement involves new
types of exposure than would be encountered in conventional wastewater systems.
Recommendations on safe handling of excreta in such systems are still lacking,
12
Microbial risk assessment in urban water systems
especially regarding colder climates. The World Health Organization will publish
separate guidelines for the safe reuse of excreta in 2005 (T.A. Stenström, pers. comm.).
While the health risks associated with the reuse of urine have been assessed
(Höglund et al., 2002) the risks associated with storage and reuse of the faecal
fraction still need to be evaluated, with emphasis on different exposure scenarios
and for situations relevant to Nordic countries.
Paper VI
In order to ensure the safe reuse of wastewater and sludge, proper risk management
is needed. Besides posing risks to the end users of food crops or likewise, the
handling of these fractions during treatment and reuse should be considered.
Proposed here is the application of Hazard Analysis and Critical Control Points
(HACCP) for the management of microbial hazards in wastewater and sludge
handling and reuse. This has recently also been recommended by Water UK for
biosolids treatment and use in agriculture (Water UK, 2004). Other transmission
routes besides the consumption of crop however need to be studied, both on an
individual basis and on a population basis.
2.3
Approaches
Paper I
A systems analysis comparing the existing drinking water system in the city of
Gothenburg with two scenarios with membrane treatment was undertaken (the
system structures are presented in Figure 2-1). The methods used were Material
Flow Analysis (MFA) for the evaluation of environmental aspects and Quantitative
Microbial Risk Assessment (QMRA) for the estimation of health effects. The
pathogens chosen for evaluation were Campylobacter, rotavirus and Cryptosporidium.
Paper II
Incident reports dating several years were reviewed and interviews with key staff
were undertaken in order to assemble information regarding failures in treatment
and distribution. Based on a compilation of this material a QMRA was carried out
with the same pathogens as in Paper I. The QMRA was validated against
epidemiological data and included a sensitivity analysis.
Paper III
Norovirus concentrations in a surface water source in the Netherlands were
quantitatively described by a most-probable-number (MPN) approach. Statistical
distributions were fitted to these data for use in risk assessment. The viruses were
monitored monthly over a whole year and daily during part of the winter season.
Time series analysis was evaluated as a tool for the prediction of forthcoming
concentrations. In addition, samples were analysed for enteroviruses, rotaviruses, Fspecific bacteriophages and turbidity.
13
Therese Westrell
Conventional
Water works
Coag/Flocc.
Sediment.
Surface water
Rapid filtration
with GAC
Chlorination
Exposure
Distribution
system
Decentralised with ultrafiltration
Decentralised with microfiltration & RO
Surface water
Surface water
Exposure
Distribution
system
Distribution
system
All water use
Ultrafiltration
Tank
Exposure
Other uses
Microfiltration
Drinking &
cooking
RO
Tank
FIGURE 2-1. System structures analysed with MFA and QMRA in Paper I. GAC =
granular activated carbon. RO = reverse osmosis.
Paper IV
Three sources of information were used to estimate the daily water intake in the
Swedish population; a nationwide survey on health and environmental factors from
1999, data from a waterborne disease outbreak investigation in 2002, and a small
study on water consumption performed within the Urban Water program in 2003.
All three data sets were representative for the Swedish population regarding age and
sex distribution. The central issue was the daily consumption of cold tap water
however analyses were also made regarding the daily consumption of heated tap
water and bottled water.
Paper V
The QMRA dealt with the handling of faeces and possible reuse in the garden as
fertiliser. The study was based on a theoretical system where households would use
dry toilets in which faeces can be stored for a year before emptying, either for
additional storage indoors or for application in the garden. Scenarios were tested for
three different exposures; emptying and spreading of the material in the garden,
recreational activities in the garden and gardening; and for three different storage
times; zero, six or twelve months. Seven pathogens were included in the QMRA,
two bacteria, two viruses, two protozoa and one helminth. Two different
approaches were used: 1) calculation of the risks of exposure to faeces taking the
incidence of the pathogens in the population into consideration (‘unconditional’)
and 2) calculation of the risks when a person in the household was assumed to be
infected (‘conditional’).
14
Microbial risk assessment in urban water systems
Paper VI
This paper integrated QMRA with HACCP for the safe handling and reuse of
wastewater and sludge. As a case study a wastewater treatment plant of a medium
sized city in the south of Sweden was used. Places where people could be exposed
to wastewater or sludge were identified through site visits and in discussions with
staff. These included exposure to wastewater and sludge of staff at the treatment
plant, visitors to the wetland, people swimming at the bathing place, entrepreneurs
collecting and spreading sludge in agriculture and consumers of sludge-fertilised
crops. Six pathogens were included in the QMRA; two bacteria, two viruses and
two protozoa. The results were expressed as risk per exposure as well as the annual
number of infections per exposure route. The latter was also estimated in terms of
the resulting increase of endemic disease in the population
15
Microbial risk assessment in urban water systems
3 Quantitative microbial risk assessment (QMRA)
QMRA was the method chosen for assessing risks in water systems. A quantitative
approach was preferred since it enables a better comparison of different water and
sanitation systems. Substantial amounts of background data have been required for
the risk assessments, especially regarding the exposure assessment. Described here
are the methods of data collection, the data itself and the data handling in a QMRA
framework.
The rationale for the selection of pathogens is described under Hazard identification
in Section 3.1 followed by a short description of each pathogen. The Exposure
assessment in Section 3.2 deals with how exposures were identified and the
information needed to estimate doses of pathogens at exposure. Data availability
and quality is addressed, as is the line of thought on how the doses were derived. A
separation was made between exposure to ‘clean’ water, primarily drinking water but
also recreational water, and ‘dirty’ fractions, namely wastewater, sludge and faeces.
Groundwater, greywater, urine and stormwater were not dealt with in the thesis.
The Dose-response assessment Section 3.3 mainly deals with how the relationships
between dose and response have been derived in previous studies and limitations of
the data. Examples of dose-response parameters used in the risk assessments are
also given. Issues of data handling, in particular how to address variability and
uncertainty, are described in the Hazard characterisation Section 3.4.
3.1
Hazard identification
In hazard identification the microbial agents are identified as well as the spectrum of
human illness and diseases associated with each specific pathogen. This also
includes pathogenicity and virulence of the microorganism, aspects of acquired
immunity and multiple exposures (for example exposure on different days) of the
host etc.
There is a large range of known waterborne pathogens representing the different
groups; bacteria, viruses, protozoa (unicellular organisms) and helminths. Some, like
Salmonella typhi or Vibrio cholerae, have been known for a long time, while others, like
noroviruses and Escherichia coli O157, have been discovered quite recently
(LeChevallier et al., 1999a, 1999b). Since it is not feasible to assess the potential
impact of all waterborne pathogens in a risk assessment, a few are chosen as
reference pathogens (WHO, 2004). The choice of reference pathogens in this thesis
was based on the following criteria:
17
Therese Westrell
• The major types of organisms should be represented, i.e. bacteria, viruses and
parasites;
• The organisms should be occurring in the Swedish population;
• They should have a documented record of being involved in waterborne
disease outbreaks or constitute a hazard in sanitation;
• Some of the most persistent organisms should be included;
• Organisms with low infectious doses should be represented;
• Organisms with more serious symptoms should be included if relevant; and
• The organism and its occurrence should be sufficiently well described in the
literature.
The pathogens selected for use in QMRA (Papers I, II, V and VI) or studied in
Paper III are described in Sections 3.1.1-3.1.4. Most of the information is from the
microbial fact sheets of the recent WHO Guidelines for Drinking Water Quality
(WHO, 2004) if not stated otherwise.
The incidence rate of a disease is the yearly number of reported cases divided by the
total population, often expressed per 100 000 people (Table 3-1). The reported
number of cases is however often substantially underestimated. Several events have
to follow upon each other for a case to enter the statistics. The infected person,
who excretes pathogens in faeces, may not show any symptoms of disease. People
with symptoms must then feel so ill that they seek medical care. The doctor must
set the right diagnosis and then report the case further. Not all waterborne diseases
are notifiable (have to be reported to an authority) in Sweden and reporting of
diseases from laboratories is voluntary (Anonymous, 2004). Estimates of
underreporting (i.e. how many more cases exist in the community) are presented in
Table 3-1. The real incidence of rotavirus is for example estimated to be 21 x 35 =
754 per 100 000 (see Table 3-1). Generally, pathogens causing less severe symptoms
are less likely to be reported (Wheeler et al., 1999).
In order to become infected when exposed to a pathogen, pathogens must breach
the host’s defence mechanisms to reach the target cells where they multiply. ID50 is
the dose, or number of pathogens, at which 50% of a population will be infected.
Dose-response issues are further discussed in Section 3.3. An infected person
excretes pathogens, often in very high numbers and for many days (Table 3-1). Not
all infections are symptomatic, however. Morbidity is a measure of the percentage
of people that will acquire symptoms when infected. As can be seen in Table 3-1
this figure varies considerably.
18
Microbial risk assessment in urban water systems
TABLE 3-1. Different epidemiological statistics of the waterborne pathogens
represented in this thesis
Incidencea
(per 100 000)
Salmonella
Campylobacter
EHEC
Hepatitis A
Rotavirus
Norovirus
Adenovirus
Cryptosporidium
Giardia
Ascaris
42-58
78-97
0.8-1.4
0.8-7.8
21d
1.2d
300d
0.3-1.6
15-26
15-25v
Underreporting
3.2d
7.6d
4.5-8.3i
3k
35d
1562d
4-19t
20k
-
Morbidity
(%)
6-80e
25f
76-89i
70m
50n
70o
54r
39e
20-40u
15g
Excretionb
(g-1
faeces)
104-8
106-9 g
102-3
104-6
107-11
105-9 p
1011 s
107-8
105-8
104
Durationb
(days)
26-51
1-77h
5-12
13-30
1-39
5-22q
1-14r
2-30
28-284
107-557
ID50c
23 600
900
1 120
30
6
10?
1.7
165
35
0.7
a Based on reporting to Swedish national surveillance 1997-2003 (SMI, 2004) if not stated otherwise.
Cryptosporidium is from July 2004 a notifiable disease but was until then only included in the voluntary
laboratory reporting. b Adapted from Paper V, with 90% confidence interval, if not stated otherwise.
NB these are biased to those sick enough to be examined, so it may overestimate the excretion rates in
the general population infected with the pathogen. c Dose which will infect 50% of exposed
individuals. Based on dose-response models reported in Table 3-10. The dose-response for norovirus
is still unknown but as few as ten organisms may be sufficient to cause infection (Schaub and Oshiro,
2000). d Wheeler et al. (1999). The incidence for adenovirus was estimated from a community study
and the underreporting is therefore not required. e Haas et al. (1999). f Havelaar et al. (2000b).
g (Feachem et al., 1983). h Appendix 3 in Havelaar et al. (2000b). i Michel et al. (2000). k Mead et al.
(1999). m Lemon (1997). Children younger than two years rarely manifest symptoms when infected.
n Gerba et al. (1996b). o Graham et al. (1994). p Marshall et al. (2001) and K.O. Hedlund, pers. comm.
q Rockx et al. (2002). r Van et al. (1992). s Wadell (1984). t Calculated from Carrique-Mas (2001) based
on the incidence in the first column. u Tessier and Davies (1999). v Incidence in Denmark,
underreporting accounted for (Arnbjerg-Nielsen et al., 2004).
3.1.1
Bacteria
The chosen waterborne bacteria may be zoonotic, i.e. can be transmitted from
animals to humans and vice versa. Bacteria generally have a shorter survival in the
environment compared to enteric viruses, parasitic protozoa and helminth ova and
are often more easily killed by disinfection. On the other hand the bacteria are the
only group that may multiply in the environment (under favourable conditions). The
infectious dose is high for many bacteria, but low for some of the selected ones,
such as EHEC and Campylobacter (see Table 3-1).
Campylobacter
Campylobacter is the most commonly identified cause of waterborne disease
outbreaks in Sweden (Andersson and Bohan, 2001) (see Section 1.1.1) and the most
important cause of acute gastroenteritis both nationally (Table 3-1) and worldwide.
The most frequently isolated species from humans are C. jejuni and C. coli. The
disease outcome is mainly gastrointestinal symptoms (abdominal pains and
diarrhoea), but C. jejuni can also give rise to reactive arthritis, meningitis and the
19
Therese Westrell
severe Guillain-Barré syndrome (an acute immune-mediated disorder of the
peripheral nervous system) (Havelaar et al., 2000b; McCarthy and Giesecke, 2001).
The principal reservoir of pathogenic Campylobacter spp. is the alimentary tract of
mammals and birds, commonly found in broilers, cattle, pigs, sheep, wild animals
and birds, and domestic pets (Koenraad et al., 1997; Wahlström et al., 2003; WHO,
2004).
Salmonella
Salmonella is second only to Campylobacter in terms of the number of annual cases in
Sweden regarding reportable enteric diseases (Table 3-1) (SMI, 2004). Salmonella are
classified into a few species with numerous subspecies or serovars. Many of these
are zoonotic and occur frequently in poultry, cattle and swine stocks. Most species
cause self-limiting gastroenteritis however S. typhi and S. paratyphi can cause sepsis
and other serious symptoms. Only non-typhoid Salmonellae were considered in this
thesis. Salmonella has the highest infectious dose of the pathogens used in QMRA in
this thesis (Table 3-1). It is commonly found in wastewater (Table 3-5) and sewage
sludge (Table 3-8) and has been shown to be able to multiply in sludge and sludgeamended soil (Gibbs et al., 1997).
Enterohaemorrhagic E. coli (EHEC, VTEC or STEC)
Enterohaemorrhagic Escherichia coli (EHEC), verotoxin producing E. coli (VTEC) or
Shiga-toxin producing E. coli (STEC) are different names for the same group of
organism, of which the strain O157:H7 is the most commonly recognised. This
bacterium can give rise to bloody diarrhoea and 2-7% of cases develop haemolytic
uremic syndrome (HUS), which can cause severe kidney syndromes that can be
fatal. Cattle are thought to be a primary reservoir for EHEC organisms (AWWA,
1999). The infectious dose is thought to be low, however some discrepancies exist
between studies (Haas et al., 2000; Strachan et al., 2001; Teunis et al., 2004). The
disease is fairly rare among humans in Sweden (Table 3-1), and therefore the
prevalence of EHEC in municipal sewage is expected to be low.
3.1.2
Enteric viruses
Viruses excreted in faeces (enteric viruses) are much more host-specific than
bacteria and therefore a certain virus strain will normally only infect a certain host,
e.g. humans. Viruses need to infect host cells in order to replicate and can therefore
not multiply outside of the host. The surface structure of the capsid (the outer
“shell” surrounding the viral DNA or RNA) is more resistant than the cell wall and
cell membranes of bacteria and viruses therefore survive better in the environment.
Non-enveloped viruses are generally more persistent than enveloped viruses.
Infected individuals excrete high numbers of viruses (virions) in faeces. Since they
generally are very infectious only a few virions are sufficient to cause an infection.
20
Microbial risk assessment in urban water systems
Hepatitis A
The hepatitis A virus can cause infectious hepatitis, which affects the liver and gives
rise to the classical symptoms of jaundice. In many cases however, especially in
children, the infection is asymptomatic (Koff, 1998). An infection with hepatitis A
virus is thought to elicit life-long immunity. The disease is not so common in
Sweden (Table 3-1). The transmission of hepatitis via contaminated water and food
is well established (WHO, 2004).
Rotavirus
The symptoms of rotavirus infection are gastrointestinal illness with vomiting and
diarrhoea. Internationally, rotavirus accounts for nearly half of all cases of diarrhoea
in children younger than two years requiring admission to hospital, and may also
account for 5-10% of sporadic cases of diarrhoea in adults (Hrdy, 1987). It is also
the major cause of gastroenteritis in Swedish children (Uhnoo et al., 1986). The virus
has the highest infectivity of any known waterborne virus (Gerba et al., 1996b) and
asymptomatic infections occur frequently (Anderson and Weber, 2004). Only one
large waterborne outbreak has been reported in Sweden (Stenström et al., 1994).
Norovirus
Norovirus is one of two genera of the human caliciviruses (the other being Sapovirus).
They were formerly known as Norwalk-like viruses (NLV) or ‘small round
structured viruses’ (SRSV). The symptoms of norovirus infection are acute viral
gastroenteritis and vomiting, which generally ceases within a few days. Noroviruses
affect all age groups and are today considered to be the most common cause of
gastroenteritis in the western world regarding the number of outbreaks and people
affected (Koopmans and Duizer, 2004). These viruses can be passed from personto-person but are also transmitted via contaminated water, foods and fomites (solid
surfaces). The number of reported waterborne outbreaks with noroviruses is
steadily increasing both in Sweden and internationally (Andersson and Bohan,
2001). In Paper III it was shown that the concentration of noroviruses in surface
water can be substantial.
Adenoviruses
Human adenoviruses consist of 51 antigenic types associated with a wide range of
infections including gastrointestinal, respiratory, urinary tract and eye infections.
The types of particular concern for waterborne gastrointestinal illness are
adenovirus types 40 and 41, which are excreted in faeces along with other serotypes.
Culturable adenoviruses are frequently found in surface waters (Pina et al., 1998),
although often in low concentrations (Tani et al., 1995), yet some may come from
other animals (de Motes et al., 2004) and therefore will not be infectious in humans.
Furthermore, in the control of Cryptosporidium, changes in disinfection from chlorine
to UV may result in increased adenovirus risk, as they are extremely resistant to UV
21
Therese Westrell
disinfection (Meng and Gerba, 1996) and can survive for long periods in water
environments (Enriquez et al., 1995).
3.1.3
Parasitic protozoa
Protozoa are unicellular parasitic organisms with complex life cycles. After passing
several stages within the host (including sexual reproduction) a transmission stage is
formed which is excreted in the faeces. These so called cysts or oocysts are very
resistant to different environmental conditions. Many of the protozoa are zoonotic.
Cryptosporidium
The host ranges of different types of Cryptosporidium vary. Infections of
Cryptosporidium in humans are caused by C. hominis, previously classified as C. parvum
genotype 1, or by the animal genotype 2, C. parvum (Carey et al., 2004; Xiao et al.,
2004). The protozoa cause self-limiting diarrhoea, however cryptosporidiosis can be
life threatening in immunocompromised people. C. parvum is very common among
newborn calves that can excrete oocysts in high numbers, but is also frequently
found in adult livestock and other ruminants. The oocysts are extremely resistant to
chlorination and have been involved in many waterborne outbreaks, e.g. the
Milwaukee outbreak (see Section 1.1.1).
Giardia
The flagellated protozoa Giardia has been found in a variety of animals. The species
infecting humans is G. intestinalis (syn. G. lamblia, G. duodenum), which can also infect
numerous mammals (AWWA, 1999). Symptoms generally include diarrhoea and
abdominal cramps, however many infections may be asymptomatic. Giardiasis may
be chronic in some patients, lasting for more than one year. Giardia is the second
most common identified etiological agent in waterborne outbreaks of disease in
Sweden (Andersson and Bohan, 2001).
3.1.4
Helminths
Parasitic helminths have complex life cycles where the survival stages, ova, are
excreted from the host to later be ingested by a new host or a middle host. The ova
are extremely resistant to different environmental conditions although they generally
cannot withstand higher temperatures (Feachem et al., 1983). In Sweden, helminths
are mainly a problem among animal herds.
Ascaris
Most infections with the human roundworm Ascaris lumbricoides are asymptomatic
although very severe symptoms can occur due to migration of adult worms to the
liver, gall bladder or appendix (Feachem et al., 1983). The worms can lead to
impaired nutritional status in their host (AWWA, 1999). Ascariasis occurs
22
Microbial risk assessment in urban water systems
worldwide, especially in warm climates and is often associated with poor sanitary
conditions. An estimate of the worldwide prevalence is 1 273 million infections
(AWWA, 1999). The Ascaris ova can survive for several years in moist soils,
however they are sensitive to desiccation and are easily killed by high temperature,
e.g. 100% destruction after one hour at 55 ºC (Feachem et al., 1983).
3.2
Exposure assessment
Exposure assessment is an attempt to determine the frequency, duration and
magnitude of pathogen exposure by one or more pathways. The assessment is
dependent on adequate methods for recovery, detection, quantification, sensitivity,
specificity, virulence and viability of the microorganisms in question and is often
dependent on studies and models of transport and fate in the environment.
Exposure assessment uses a wide array of information sources and techniques. Most
likely, data will not be available for all aspects of the exposure assessment and those
data that are available may sometimes be of questionable or unknown quality. In
these situations qualified assumptions must be made. These are based on
professional judgments and inferences based on analogy with similar
microorganisms or processes etc. In the end the exposure assessment will be based
on a number of variables with varying degrees of uncertainty. Ideally it is important
to capture data uncertainty versus variability in full QMRA models, so these two
‘dimensions’ can be provided in the final reporting of infection risk.
Exposure to pathogens via the environment can occur through different pathways
(partly described in Section 1.1). Pathogenic microorganisms can also enter the body
in several ways. The most common is via ingestion but other routes can also be of
importance for some microorganisms, like exposure via inhalation, eye or dermal
contact (Haas et al., 1999). Standard values for use in risk assessments have been
published by the USEPA (USEPA, 1997) and others (McKone and Daniels, 1991;
Finley et al., 1994).
In Paper I exposure to potential pathogens in drinking water and water ingested
during showering was assessed. The former exposure, also used in Paper II, was
assumed to occur daily with median intakes of 0.96 L as reported by Roseberry and
Burmaster (1992), while the latter was assumed to be 10 mL ingested once a week.
In Paper IV the water consumption in Sweden was assessed and quantified for use
in future risk assessments. Paper V involved the accidental ingestion of faeces
during handling and reuse in the garden where the exposure was based on a survey
of studies of daily soil ingestion rates for adults and children. In Paper IV exposure
to wastewater and sludge during handling and reuse was considered. The exposure
volumes, frequencies and number of persons affected were based on a combination
23
Therese Westrell
of exposure data described in the literature and assumptions based on site visits and
in discussion with people at the site. These are presented in Table 3-2.
TABLE 3-2. Exposure points identified in Paper VI with assumptions on volume ingested,
frequency and the number of persons affected in a population of 29 000 people.
Volume ingested
(mL or g)
1. WWTP worker at pre-aeration
1
2. WWTP worker at belt press
5
3. (Un)intentional immersion at wetland inlet
30
4. Child playing at wetland inlet
1
5. Recreational swimming
50
6. Child playing at sludge storage
5
7. Entrepreneur spreading sludge
2
8. Consumption of raw vegetables
1
Type of exposure
Frequency
(times * year-1)
52
208
1
2
10
1
30
2
Number of
persons affected
2
1
2
30
300
2
2
500
In order to derive the dose of pathogens at each exposure, different points of
departure are made. When assessing the risk of infection from drinking water there
is the problem that most pathogens cannot be detected in potable waters due to the
lack of suitable analytical methods, low concentrations of the pathogens regarded as
a cause of concern and potentially large variations with time. A way to circumvent
this is to start from the occurrence of the pathogen in the raw water and calculate
the concentration in drinking water considering the removal or inactivation during
treatment (Teunis et al., 1997). Filtering of large volumes of water is often necessary
in order to detect pathogens in surface water (Figure 3-1).
FIGURE 3-1. In Paper III large
volume water samples of 200-500 L
were filtered at the point of sampling
while small volume water samples, 10
L, were filtered in the laboratory as
shown in the picture.
Exposure to raw or treated wastewater is more straightforward since several studies
report the concentration of different pathogens in wastewater. Another approach is
to calculate the concentration of pathogens in wastewater from epidemiological data
regarding the incidence of the illnesses in the society, adjusted for the rate of
24
Microbial risk assessment in urban water systems
underreporting (i.e. the proportion of cases that will not be reported to
epidemiological statistics) and morbidity (i.e. the proportion of symptomatic cases to
infections), the excretion of the pathogens from an infected host and the dilution in
wastewater. Such an approach was used in Paper V to estimate the potential
concentrations of pathogens in faeces.
Due to the problem of detection of many pathogens in water, indicator organisms
have been used to imply potential occurrence of pathogens in different water
streams. Their concentration and reduction are often used as surrogate values for
pathogens. The traditional indicators are bacteria of faecal origin; total coliforms,
faecal coliforms (a subset of total coliforms), Escherichia coli and Enterococci. During
recent years the suitability of indicators has been largely criticised (Ashbolt et al.,
2001). The bacterial indicators have been found to be more sensitive to decay than
many of the other pathogenic microorganisms and thus do not adequately reflect
the survival of more sturdy pathogens such as viruses and protozoa. If risk
assessment is based solely on the traditional indicators it will probably
underestimate the risks.
There is also no direct correlation between the indicators and the occurrence of a
certain pathogen in water (Jiang and Chu, 2004, Paper III). The occurrence of
pathogens in a water source should therefore be based as much as possible on real
pathogen measurements, either from the water of interest or from similar sources
described in scientific papers. Indicator organisms can nonetheless be useful as
process indicators, i.e. to assess the removal of microorganisms in different
treatments or model pathogen survival in the environment. Additional indicators,
such as bacteriophages and spore-forming bacteria, can be included to better model
the behaviour and survival of viruses and protozoa respectively.
Many studies report the reduction of microorganisms by different treatment
processes and their die-off or potential re-growth in the environment. Most of these
are based on indicator bacteria but some were undertaken with pathogens. There
are also differences in the scale of studies, ranging from full-scale to pilot to benchscale studies. The time and cost of performing microbiological investigations makes
it unfeasible to carry out for every system under evaluation. Therefore previous
studies reported in the literature are very useful. Many variables included in a
microbial risk assessment are however subjected to regional differences, as
mentioned in Section 1-3. There is therefore a great need for locally-produced data
regarding critical input variables (further dealt with in Section 6.2 and 7.1).
3.2.1
Source water to drinking water
The main pathogen sources of surface water contamination are sewage effluents and
agricultural run-off. Concentrations of pathogens in raw water vary substantially
25
Therese Westrell
depending on the degree of anthropogenic activities. Wastewater discharge may
constitute a significant proportion of the water flow in rivers. Any enteric pathogen
that occurs in the population can potentially be found in surface waters impacted by
wastewater discharges, though will fluctuate due to seasonality in the incidence of
the disease and variability in treatment efficiency. Run-off from agricultural land can
contain pathogens from livestock, such as Cryptosporidium and EHEC. Campylobacter
can be frequent in the bird population (Waldenström et al., 2002; Wahlström et al.,
2003).
In a survey of Swedish surface water sources, Giardia was detected in 26% and
Cryptosporidium in 32% of the investigated waters (Hansen and Stenström, 1998)
(Table 3-3). When screened for Campylobacter, 7% of all samples from Swedish
surface waters were found positive (SLV, 2002). With repeated sampling in some
water sources the bacterium was found in 38% of the samples, often in the absence
of faecal indicator bacteria. Campylobacter was also detected in groundwater with
clear faecal contamination. In Norway Campylobacter was isolated from 53% of the
samples from a river (Rosef et al., 2001) and from 17% of Finnish lakes and rivers
(Hörman et al., 2004).
In Paper I and II previously unpublished data from Hansen was used where
Cryptosporidium had been detected in 45% of the samples (Table 3-3). Quantitative
data were lacking regarding Campylobacter and rotavirus, which instead were taken
from Obiri-Danso and Jones (1999) and Gerba et al. (1996b), respectively. In Paper
III the concentrations of noroviruses were estimated from a dilution series of
nucleic acid extraction detected with reverse transcriptase polymerase chain reaction
(RT-PCR).
Environmental factors such as temperature, UV light, water currents and rainfall
affect the concentration of viable pathogens reaching the waterworks, as does
pathogen adhesion to particles and sediments (Rao et al., 1984; LeChevallier et al.,
1991; Atherholt et al., 1998; Obiri-Danso and Jones, 1999; Payment et al., 2000).
Interaction with autochthonous populations of microorganisms will also have a
substantial effect on the survival of pathogens in aquatic environments (Medema et
al., 1997). The time (in days) for 90% die-off of some pathogens in fresh water can
be found in Table 3-7.
Drinking water treatment units act as microbiological barriers. A microbiological
barrier is “an appliance or action that counteracts the occurrence of disease-causing
viruses, bacteria and parasitic protozoa in the drinking water” (from the Swedish
drinking water regulations, SLV, 2003). The number of barriers needed is dependent
on the quality of the raw water (SLV, 2003). The barriers could be based either on
mechanical removal or inactivation via disinfection. Examples of mechanical
barriers are sedimentation, flotation and filtration, which physically (and sometimes
26
Microbial risk assessment in urban water systems
biologically) remove the microorganisms. Through prior addition of chemical
coagulants the adsorption of microorganisms to particles is increased and the
aggregation of these to form larger particles enhances their removal. In the new
national guidelines membrane filtration with pore sizes of 100 nm or less is also
accepted as a barrier. These membranes include nanofilters, reverse osmosis and
most ultrafilters, however not microfilters with regard to viruses. Disinfection acts
by damaging the surface structures or the nucleic acid of the organisms, i.e. DNA or
RNA. Chlorine for example acts by disrupting cell permeability but also damages
nucleic acids and enzymes while UV light mainly causes damages to the nucleic acid
(Bitton, 1994).
TABLE 3-3. Concentration (L-1) of different pathogens in surface water
Campylobacter
EHEC
Enterovirus
Hepatitis A
Rotavirus
Norovirus
Adenovirus
Cryptosporidium
Giardia
Mean
Det. in
100 mL
40-70b
Det. in
90 mL
0.3
41
Det. in
1L
0.2-29
Range
n.a.
Posa
7%
Place
Sweden
<10-600
94%
UK
n.r.
1.7
Canada
0.003-0.90
0-190
100%
n.r.
The Netherlands
Japan
Paper III
Tani et al., 1995
n.a.
43%
Spain
Pina et al., 1998
0-41
8-100%
Gerba et al., 1996b
n.a.
9.4%
North and South
America
Finland
Paper III
Paper III
Pina et al., 1998
Det. in
1L
40
390
Det. in
2L
n.r.
1.0b
<1-240
12- 1 700
n.a.
38%
100%c
65%
The Netherlands
The Netherlands
Spain
<1-25
<0.01-4.6
24%
32%
Japan
Sweden
0.2
<0.01-16
45%
Sweden
0.1
<0.1-0.4
21%
Norway
0.5
<0.01-4.6
26%
Sweden
0.1
<0.1-4
16%
Norway
Percent positive samples. b Geometric mean. c Winter months.
Det. = detected, n.r. = not reported, n.a. = not applicable
a
27
Reference
SLV, 2002
Obiri-Danso and
Jones, 1999
Gannon et al., 2004
Hörman et al., 2004
Tani et al., 1995
Hansen and
Stenström, 1998
Hansen, unpublished.
Used in Papers I and
II
Robertson and
Gjerde, 2001
Hansen and
Stenström, 1998
Robertson and
Gjerde, 2001
Therese Westrell
A compilation of studies regarding the removal and inactivation of microorganisms
in drinking water treatment processes was undertaken for Papers I and II. The
removal was described as triangular distributions of decimal reduction (i.e. expressed
as log10 removal) for use in the risk assessment (see further Section 3.4.3). The
decimal reduction or log10 removal is:
Log 10 (C in ) − Log 10 (C out )
where Cin is the incoming and Cout the outgoing concentration.
TABLE 3-4. Median and range of removal in log10 of microorganisms in
drinking water treatment processes. Modified from Hijnen et al. (2004),
Paper II and Kärrman et al. (2004).
Process
Coagulation/flocculation
Rapid sand or GACb filtrationc
Slow sand filtration
Chlorinationc
UV inactivationc, e
Microfiltrationc
Ultrafiltrationc
Bacteria
1.7 (0.5-3.9)
1.0 (0.3-1.5)
2.2 (1.3-3.4)
3.5 (2.5-5.0)
(6.5-8)
7 (3-8)
8 (7->8)
Viruses
1.9 (0.2-4.3)
(0.7-1.2)
2.1 (0.9-3.5)
2.0 (1.5-3.0)
(3-4)
2 (<1-3)
6 (1->8)
Protozoa
2.0 (0.4-3.7)a
0.6 (0-1.4)
n.d.d (0.3->6.5)a
0.4 (0-1.0)
(3->4.5)
6.5 (>4-7)
8 (3-8)
Cryptosporidium. Lower removal reported for Giardia. b GAC = granular activated carbon.
Compilation adjusted for the process conditions described in Paper III and Kärrman et
al. (2004) regarding the pathogens Campylobacter, rotavirus and Cryptosporidium. d n.d. = not
determined. e Estimated at a UV dose of 30-40 mWs/cm2
a
c
As shown in Table 3-4 the removal of microorganisms varies between treatment
process and organism groups. This is due to different properties of the
microorganisms. Viruses are for example more likely to pass barriers that are solely
based on mechanical removal (e.g. microfiltration) due to their small size, while
protozoa and especially Cryptosporidium, are very resistant to chlorination. The
implications for this regarding doses of pathogens reaching consumers is further
discussed in Paper II. There can also be large differences between organisms within
a certain group. The resistance to UV disinfection for example, varies substantially
among the enteric viruses with hepatitis A being the least and adenoviruses the most
resistant (Hijnen et al., 2004). Such differences may depend on structural differences
between the organisms, for example whether the organisms have DNA or RNA and
whether this is single or double stranded (Meng and Gerba, 1996).
After treatment the water is supplied to the consumers via the distribution network.
If a chlorine residual is present additional reduction of pathogens is possible.
Payment et al. (1999) however found that poliovirus and indicators, except E. coli
were relatively unaffected by the chlorine residuals in the tested systems.
28
Microbial risk assessment in urban water systems
Contamination of treated drinking water with wastewater during distribution is one
of the most common causes of waterborne outbreaks (Stenström et al., 1994).
Effects of microbial contamination of the distribution network on the health of the
consumers were assessed in Paper II.
3.2.2
Wastewater and sewage sludge
The occurrence of pathogens in sewage is dependent on the infection levels in the
population, which may vary with season (Mounts et al., 2000; Nylen et al., 2002). If
slaughterhouses are connected to the municipal sewage they can also contribute
substantial amounts of zoonotic pathogens, e.g. Campylobacter (Höller, 1988;
Koenraad et al., 1996). Some pathogens such as Salmonella and adenoviruses are
common in the population and are likely to be detected in wastewater (Table 3-5).
Others such as noroviruses have a more seasonal occurrence, and are detected in
higher concentrations in wintertime (Lodder et al., 1999; Ottoson et al., submitted)
(Table 3-5). Since EHEC have only been detected with PCR in sewage, the
concentration of this organism for use in Paper VI was based on the incidence rate,
excretion rate and duration, and adjusted with the ratio of calculated and detected
numbers of E. coli in sewage.
TABLE 3-5. Concentration (L-1) of different pathogens found in wastewater
Campylobacter
Salmonella
EHEC
Enterovirus
Hepatitis A
Rotavirus
Norovirus
Adenovirus
Giardia
Cryptosporidium
Ascaris
Mean
160 000
Det. in
1 mL
22 000
Det.
Det. in
25 g
28 000
Det in
50 µL
215
1 900
7 600d
2 000
20
30
Range
500-4 400 000
n.a.
Pos.a
100%
41%
930-110 000
n.a.
n.a.
100%
93%b
53%b
4 200-720 000
n.a.
100%
23%
40-510
<800-4 500
250-25 000
260-13 000
<8-160
n.r.
100%
78%c
100%
100%
28%
n.r.
Place
Reference
Germany Höller, 1988
Sweden Carlander and
Stenström, 2001
Finland Koivunen et al., 2003
Germany Höller et al., 1999
France Vernozy-Rozand et al.,
2002
Sweden Ottoson et al., submitted
France Schvoerer et al., 2000
USA
Sweden
Sweden
Sweden
Sweden
USA
Rao et al., 1987
Ottoson et al., submitted
Bofill-Mas et al., 2000
Ottoson et al., submitted
Ottoson et al., submitted
AWWA, 1999
Percent positive samples. b Positive for stx genes by PCR, E. coli O157:H7 one of the serotypes
detected. c 78% positive in winter time, otherwise <100 L-1. d Mean calculated from only four
samples. Det. = detected, n.a. = not applicable, n.r. = not reported
a
The mathematical procedure used in Paper III to estimate concentrations of
noroviruses from dilution series of RT-PCR was used also in Ottoson et al.
(submitted) for the estimation of enterovirus and norovirus concentrations in sewage.
29
Therese Westrell
The treatment in Swedish wastewater treatment plants normally consists of
sedimentation, chemical precipitation and biological treatment (mostly activated
sludge, but also trickling filters). During the 1990’s, tertiary treatment for nitrogen
reduction was also implemented in many municipalities (MISTRA, 1999), resulting
in the introduction of wetland treatment and similar solutions. In an investigation
by Stenström (1987) the reduction of indicator organisms in the water phase at
different wastewater treatment plants in Sweden was about 99%. A higher reduction
would be desirable since the concentration of some organisms is still in the range of
hundreds to several thousands per millilitre. The log10 removal of microorganisms in
the wastewater treatments used in Paper VI is listed in Table 3-6.
TABLE 3-6. Mean and range of removal in log10 of microorganisms in the
wastewater treatment processes used in Paper VI. Based on previous
measurements of faecal coliforms, coliphages and clostridia spores in this
system (Stenström et al., 1985) and in a constructed wetland (Stenström
and Carlander, 2001).
Process
Pre-aeration and sedimentation
(chemically aided)
Activated sludge
Chemical precipitation
Sand filter
Wetland
Bacteria
Viruses
Protozoa
0.2 (0-0.7)
0.2 (0-0.6)
0.2 (0-1.1)
1.3 (1.1-1.8)
0.4 (0.1-1.0)
0.5 (0.1-1.0)
2
0.6 (0.2-0.7)
0.7 (0.5-1.0)
0.2 (0-0.5)
1
0.7 (0.4-0.9)
0.3 (0-1.2)
0.4 (0.3-0.8)
1
The treated wastewater is discharged into the receiving waters. Pathogens can
survive for substantial periods in water environments (Table 3-7) and can
subsequently contaminate surface waters used for drinking water production and
recreational waters (see Section 3.2.1). Numerous epidemiological studies have
shown that swimming in wastewater contaminated water results in a greater risk of
gastroenteritis (Gerba et al., 1996b).
The die-off of pathogens in water and other materials often follows a first-order
decay rate (Tchobanoglous and Burton, 1991):
r = C×k
where r is the rate of die-off per unit time per unit volume of water, C is the
concentration of the pathogen and k the decay constant. In many studies the time
for 90% reduction is instead reported as T90. The relationship between k and T90 is
T90 =
− ln(0.1) 2.3
≈
k
k
30
Microbial risk assessment in urban water systems
The T90-values used for the survival of pathogens in lake water in Paper VI are
presented in Table 3-7. The temperatures recorded in the studies ranged from 4 ºC
to 25 ºC and the higher temperature the shorter the survival. Although it is unlikely
that people will go swimming in water temperatures below 15 ºC, the lower
temperatures were included since the survival can be very long, especially for
Cryptosporidium (up to three months for 1 log10 reduction).
TABLE 3-7 Die-off of pathogens in fresh water expressed
as days for 90% reduction, T90. Values used in Paper VI
Organism
EHEC
Salmonella
Rotavirus
Adenovirus
Giardia
Cryptosporidium
T90
7-19
same as for EHEC
5-16
21-29
23-30
40-100
Reference
Wang and Doyle, 1998
Assumption
Raphael et al., 1985
Enriquez et al., 1995
DeRegnier et al., 1989
Medema et al., 1997
During wastewater treatment, many of the microorganisms are trapped in, or
adsorbed to, particulates and concentrated in the sludge (Chauret et al., 1999).
Concentrations of pathogens in untreated sludge are presented in Table 3-8. In
Paper VI a different approach was however used, in order to be able to utilise
national data on the occurrence of some pathogens in sewage. The ratio between
concentrations in sewage and in sludge was calculated from the data by Chauret et
al. (1999) for indicator bacteria, protozoa and bacteriophages, respectively. These
ratios were then used to estimate the concentration of pathogens in sludge based on
their concentration in sewage.
TABLE 3-8. Concentration (g-1 wet weight) of different pathogens found in sewage sludge
Mean
Campylobacter
Det. in 25 g
Salmonella
Det. in 25 g
EHEC
Det. in 25 g
Det in 25 g
Giardia
441
Cryptosporidium
529
Nematodesb
n.r.
Range
n.a.
n.a.
n.a.
n.a.
<25-1180
<25-381
0.2-4.5
Posa
30%
64%
16%
2%
n.r.
n.r.
n.r.
Place
Sweden
Sweden
France
Sweden
Canada
Canada
France
Reference
Sahlström et al., 2004
Sahlström et al., 2004
Vernozy-Rozand et al., 2002
Sahlström et al., 2004
Chauret et al., 1999
Chauret et al., 1999
Gantzer et al., 2001
a Percentage positive samples. b Ascaris constituted 35% of viable nematode eggs. Figures are reported
per gram dry matter. Det. = detected, n.a. = not applicable, n.r. = not reported
The most common sludge treatment in Sweden is mesophilic anaerobic digestion.
Chauret et al. (1999) found that the reduction or inactivation of microorganisms in
mesophilic anaerobic digestions of sludge was low, especially for protozoa.
Sahlström et al. (2004) also showed that both Salmonella and Campylobacter were
detected in digested sludge. Thermophilic digestion has however shown to be much
more efficient. While the removal of total coliforms in mesophilic digestion was 1.5
31
Therese Westrell
log10 the corresponding figure for thermophilic digestion was 5.9 log10 (Sahlström et
al., 2004). Aerobic digestion has been shown to completely eliminate Campylobacter
(Koenraad et al., 1997), which may be attributed to the bacteria’s inability to cope
with aerobic conditions (these bacteria are microaerophilic).
In Sweden, reuse of waste products from wastewater treatment has mainly included
sludge. Approximately 35% of all of the sludge from municipal plants was earlier
used in agriculture, however the percentage used today has decreased after the
directive issued by the Federation of Swedish Farmers (started after discussions on
the presence of bromated flame retardants in the sludge) (MISTRA, 1999). The
wastewater treatment plant used as a case study in Paper VI was still using a part of
its sludge in agriculture, however the assumption made in the paper was that all
sludge was applied.
3.2.3
Faeces
The faeces fraction is normally disposed of by flushing to municipal sewers. Other
systems where the faeces are collected dry however also occur. This type of system
was theoretically investigated in Paper V. A separate risk assessment has also been
performed on a real system in Sweden, which is discussed in Section 7.5.
The survival of pathogens in faeces is poorly investigated. Carlander and Westrell
(1998) studied the survival of bacteriophages and Ascaris ova in faeces in dry urinediverting toilets in Vietnam. In Paper V an extensive literature survey was
undertaken where studies on the survival of pathogens in animal manure, animal
slurry and sewage sludge were assumed to be applicable. A similar procedure was
used for estimating the survival of pathogens after incorporation into soil. These
resulting T90-values are presented in Table 3-9.
TABLE 3-9. Inactivation rates of pathogens
in faeces and soil expressed as days for 90%
inactivation, compiled for Paper V.
T90 faeces
T90 soil
(mean ± stdv)
(mean ± stdv)
Salmonella
30 ± 8
35 ± 6
EHEC
20 ± 4
25 ± 6
Rotavirus
60 ± 16
30 ± 8
Hepatitis A
55 ± 18
75 ± 10
Giardia
27.5 ± 9
30 ± 4
Cryptosporidium
70 ± 20
495 ± 182
Ascaris
125 ± 30
625 ± 150
32
Microbial risk assessment in urban water systems
A potential method for treatment of faeces in small-scale systems of the kind
described in Paper V is through the addition of urea (Vinnerås et al., 2003b) or by
co-composting with organic household waste (Vinnerås et al., 2003a). The latter
however requires thorough mixing of the material, which will increase the exposure
to potential pathogens in the faeces.
3.3
Dose-response assessment
Dose-response assessment aims at presenting a mathematical relationship between
the dose and the probability of infection or illness in exposed persons. Most doseresponse studies have been based on human feeding trials, i.e. volunteers have been
fed with pathogens in different doses and the percentage of subjects
seroconverting/excreting the pathogen (or other outcome such as illness) at a
certain dose is calculated. Feeding trials using human subjects can provide useful
dose-response analysis data, however the doses applied in these studies are usually
high, and the subjects are predominantly healthy individuals. Furthermore, these
studies often use one or a limited number of strains that may not represent all the
virulence characteristics of a species.
Other possible data sets that can be used are those from epidemiological outbreak
studies. Epidemiological data, if collected well and if information such as attack rate
and ingested dose are provided, can be an ideal data set. These data would
essentially provide “real-world’ response using subjects that are representative of the
population at large. Although epidemiological studies are abundant the enumeration
of the disease-causing agent in foods is difficult and often only provide a
presence/absence result.
By different mathematical methods dose-response models can be fitted to
experimental data (Crockett et al., 1996; Teunis et al., 1996). The risk of becoming
infected is dependent on the occurrence of two conditional probabilities: the
probability that the organism is ingested and the probability that the organism is
able to survive and infect the host once it is ingested. The environment, the
pathogen and the host are all variables that play an important role in the probability
of infection. Environmental influences include the food vehicle and the stability of
the microbiota of the gastrointestinal tract. Pathogen influences include the dose,
virulence, and the colonisation potential in the host gastrointestinal tract. Host
influences include immune status, age and stomach contents (Coleman and Marks,
1999), which also influence the conditional probability of illness given infection.
It was previously believed that a threshold number of organisms, or minimum
infectious dose, had to be ingested before any infection or adverse effects could
occur. Today the currently accepted theory is that of the single hit model, i.e. that a
single pathogen particle has the ability to initiate an infection or illness (Haas, 1983).
33
Therese Westrell
The probability of causing infection increases with the dose of the pathogen. The
primary non-threshold, single-hit models currently used in microbial risk assessment
are the exponential and beta-Poisson dose-response models.
In the exponential model it is assumed that all of the ingested organisms have the
same probability, r, of causing an infection. The dose ingested is assumed to be
Poisson distributed with a mean of D organisms per portion (Haas, 1983; Haas et
al., 1999).
Pinf = 1 − e − rD
where Pinf is the probability of infection, r is the probability of one organism
initiating an infection and D is the dose.
In the beta-Poisson model, heterogeneity in the organism/host interaction is
introduced and r is assumed to follow a beta-distribution (Haas, 1983; Haas et al.,
1999). The resulting model is rather complex but can be approximated under the
assumption than β is much larger than both α and 1 to:
 D
Pinf = 1 − 1 + 
β

−α
where Pinf is the probability of infection, D is the dose ingested and α and β are the
dose response parameters.
Several publications are now available on dose-response relations fitted to human
feeding trials or outbreak data. The dose-response relations used in the risk
assessments of this thesis are presented in Table 3-10. Two different models for
EHEC have been used, which vary substantially. In a recent study by Teunis et al.
(2004) the data from a foodborne outbreak with E. coli O157:H7 was shown to
agree better with a dose-response model based on Shigella than the rabbit model
used by Haas et al. (2000). For a worst-case evaluation the exact single-hit model
(r = 1), which represents the maximum risk curve (Teunis and Havelaar, 2000), can
be used. This was applied for the parasite Ascaris where no dose-response studies
are available. Norovirus dose-response studies are currently under development,
however not yet published. Preliminary results however indicate that as few as 10
PCR-detectable units may cause infection in human adult volunteers (Schaub and
Oshiro, 2000). Additional information, e.g. regarding the uncertainty of the
parameter estimates in dose-response relations, can be found in Teunis et al. (1996).
Differences have been discovered regarding pathogenicity and virulence between
strains or isolates (Teunis et al., 2002a; Coleman et al., 2004). In the future this
34
Microbial risk assessment in urban water systems
heterogeneity in infectivity should be addressed, however as long as most detection
methods used for monitoring pathogens in water do not discriminate between
strains or sometimes not even species, such distinctions cannot be made. The
incorporation into dose-response relations of differences among hosts in
susceptibility to infection (due to immunity) has been proposed by Teunis et al.
(2002b).
TABLE 3-10. Dose-response parameters for the pathogens used in the risk assessments
of Papers I, II, V and VI
Comments
H.f.t.a by Black et al. 1988
H.f.t. by McCullough and
Wesley Eisele 1951a,
1951b, 1951c. Several
strains
E. coli O157:H7
α = 0.2099,
Based on h.f.t. on Shigella
N50 b = 1 120
by DuPont et al. (1969,
1972) and Levine et al.
(1973)
α = 0.49,
Rabbit study by Pai et al.
N50 b = 5.96x105
1986
Hepatitis A
α = 0.2, N50b = 30
Assumption by Shuval et
al., 1997
Rotavirus
α = 0.253, β = 0.422 H.f.t. by Ward et al. 1986
Adenovirus 4
kc = 2.397
H.f.t. by Couch et al. 1966
Giardia lamblia
r = 0.0199
H.f.t. by Rendtorff 1954
Cryptosporidium parvum kc = 238.6
H.f.t. by DuPont et al.
1995
Ascaris
r=1
None available
a H.f.t. = human feeding trials. b β = N (21/α-1). c r = 1/k.
50
Campylobacter jejuni
Salmonella spp.
(non-typhi)
Parameters
α = 0.145, β = 7.59
α = 0.3126,
N50b = 23 600
Reference
Medema et al., 1996
Haas et al., 1999
Crockett et al., 1996
Haas et al., 2000
Shuval et al., 1997
Teunis et al. 1996
Haas et al., 1999
Teunis et al. 1996
Haas et al., 1996
Assumption
Much is however still unknown regarding the interaction of pathogens in a host.
Some of the dose-response relations are based on animal studies and the
extrapolation to humans is uncertain. Since low doses are normally encountered in
environmental transmission of pathogens there is a necessity to extrapolate the
dose-response curve to the low doses in question. Due to ethical reasons doseresponse studies have not been made for pathogens that give more severe
symptoms. Here, outbreak data can provide the needed information. Other
drawbacks are limitations in cultivation technique that can result in either
overestimates or underestimates of the correct number of viable organisms. As
mentioned earlier, the host defences (immune system) have a large impact on which
people get infected and particularly which develop more severe diseases from a
certain dose of microorganisms.
35
Therese Westrell
3.4
Risk characterisation
The final step of the risk assessment combines the information from the previous
steps to estimate the likelihood of an adverse consequence. It should include
descriptions of the variability and uncertainty of the hazards and preferably a
discussion on the magnitude of the public health problem (Haas et al., 1999).
3.4.1
Stochastic modelling versus point estimates
Point estimates can be used to calculate risks, i.e. one value is chosen to represent
each variable and the risk is calculated. Mean values of variables are chosen to
calculate the average risk while extreme values, such as the 99-percentile, can give an
idea of the worst-case situation. Such an approach however does not give a
comprehensive picture nor appropriate weight of all combinations. Stochastic
modelling where each variable is described as a distribution can instead be used. By
random sampling of each distribution with Monte Carlo methods the output risk
distribution can be obtained (Figure 3-2).
,
0,018
0,016
0.0012
0,013
0.001
0,011
0,009
0.0008
0,007
0.0006
0,004
0.0004
0,002
0,000
0.0002
0
7,375
14,75
22
Source water
x
500
Reduction during
treatment
x
1000
1500
2000
Exposure
2500
3000
→
Dose
1
0,8
0,6
0,4
0,2
0
1,00E-08
Dose
x
Dose-response equation
→
1,00E-06
1,00E-04
1,00E-02
Risk of infection
FIGURE 3-2. Schematic picture on the logical sequence of events when calculating the
risks of infection, for example from Cryptosporidium in drinking water. Arrows indicate
random sampling from distributions with Monte Carlo methods. The distributions can
be presented either as probability density curves, as shown for the exposure, or as
cumulative density curves, as shown for the risk of infection.
36
Microbial risk assessment in urban water systems
3.4.2
Variability and uncertainty
Variability is the inherent variation in the data, which cannot be reduced. The
uncertainty reflects flaws in the data collection and can accordingly be reduced by
increased investigations. One simple example is the body height of individuals in a
population. After measuring a few individuals, statistical measures about the
population as a whole can be derived, such as mean, standard deviation etc. If more
individuals are measured the certainty in the statistical measures will increase,
however the variability, for example expressed as the range of heights, will not be
reduced.
The USEPA (1997) point out three different types of variability, which can be
useful for classifying variability in different variables, e.g. concentrations, removals
etc. These three are:
• spatial variability – variability across locations. Can occur at different levels,
e.g. regional, local etc.
• temporal variability – variability over time whether long or short term
(studied in Paper III)
• inter-individual variability – variability among individuals (studied in Paper
IV)
There are also different types of uncertainty, also described by USEPA (1997):
• scenario uncertainty (e.g. incorrect or insufficient information, overlooking an
important pathway)
• parameter uncertainty (e.g. small or unrepresentative samples)
• model uncertainty (e.g. excluding relevant variables)
Often the second order uncertainty, i.e. the uncertainty in the parameter estimates of
the distribution functions, is intended when mentioned in risk assessment papers.
Second order uncertainty bounds are presented in the figures of Paper III and IV.
An example from Paper III is shown in Figure 3-3.
FIGURE 3-3. Variability and parameter
uncertainty exemplified by the gamma
distribution of the concentrations of
noroviruses in surface water during the
winter season 2002/2003 from Paper III.
uncertainty
variability
37
Therese Westrell
3.4.3
Fitting of stochastic distributions
Variability can be taken into consideration in a distribution function. The
concentrations of pathogens in source waters are for example usually low but with
occasional peaks. Lognormal distributions were therefore chosen to describe this
variability, used for example in Paper II. In Paper III the counts of culturable
viruses in a sample were described with a Poisson distribution while the variable
concentrations between samples, due to spatial and temporal variations, were
described by a Gamma distribution.
Different methods can be used to fit a distribution to data. Frey and Burmaster
(1999) describe two equally well-suited methods, maximum likelihood estimation
and bootstrap simulation, and also describe how to characterise the uncertainty in
the estimates. Maximum likelihood estimation was used in Papers III and IV and is
also one of the methods used in the mathematical package @Risk™ (Palisade
Corporation, Newfield, NY) for fitting distributions.
Maximum likelihood estimation is used for finding the parameter values of a
distribution that maximise the probability of obtaining a particular set of data. If
several distributions are tested the one with the highest likelihood will accordingly
have the best fit. The likelihood function is
n
L( x1 , x2 ,..., xn | θ 1 ,θ 2 ,...,θ k ) = L(θ ) = ∏ f ( xi ;θ 1 ,θ 2 ,...,θ k )
i =1
where L is the likelihood, θ the parameter estimates and x the observed sample
values. Often the natural logarithm (ln) of the likelihood function is used, or even
the negative log-likelihood, for computational reasons.
Often though, we do not have access to raw data for all variables in the risk
assessment. Instead, the most suitable data can be identified from literature studies.
This was the method primarily used for estimating removal efficiencies in Papers I,
II, V and VI. The values chosen for use in the QMRA were based on a judgement
on the quality and applicability of the studies of the specific case. Studies were
ranked according to quality of the study (in terms of number of samples, choice of
methods, description of set-up etc.), scale (lab, pilot, full-scale etc.), organisms (real
pathogens, indicator organisms, other indicator parameters such as particles etc.) and
process conditions (processes characteristics should be as similar as possible to
those under evaluation in the QMRA). Results from investigations on the removal
of human viruses in full-scale operation were for example weighted higher than labbench experiments with bacteriophages. Several studies were combined to get an
estimation of the removal. Often triangular distributions were used with the most
likely value and minimum and maximum values. When even less data were available
38
Microbial risk assessment in urban water systems
uniform distributions were used where all values between minimum and maximum
have the same probability of occurring. A good description on how to make this
assessment more semi-quantitative can be found in Hijnen et al. (2004) where
weights for different quality aspects of studies are applied when comparing the
results.
3.4.4
Sensitivity analysis
As seen in previous sections many of the input variables in a QMRA have large
statistical variability and uncertainties, while the quantitative effect of various
phenomena is unknown. With sensitivity analysis the effects of the input variables
on the output risk can be assessed. It can therefore be a valuable tool in quantitative
risk assessment for examining the main risk-determining phenomena, as well as the
variables that mainly determine the inaccuracy (or spread) in the risk estimate, thus
identifying where most effort should be placed in reducing uncertainties (Frey and
Patil, 2002).
Many mathematical packages used in QMRA allow for sensitivity analysis. One
example is @Risk™ where the sensitivity analysis is made with multivariate stepwise regression and/or Spearman rank order correlation. The sensitivity analysis in
Paper II was adapted from Zwietering and van Gerwen (2000). The impact of the
lowest and highest values, respectively, of a distribution (minimum and maximum
for closed distributions and 5- and 95-percentiles of open distributions) on the
output risk was assessed for each variable with all other variables left unaltered. The
resulting ‘best-case’ and ‘worst-case’, respectively were compared to the median risk
from when all variables were used.
3.4.5
Different ways of expressing risks
The risk of infection can be described or characterised in a number of different
ways depending on the purpose, scope, and level of detail of the assessment. One
common measure is the risk per person and exposure (used e.g. in Paper I). This
can in turn be described as the mean, the median, or any percentile or confidence
interval of the risk. Mean risks are sometimes presented in risk assessments,
however since extreme values from the upper or lower percentiles of an output risk
distribution can affect the average dramatically (especially when using skewed input
distributions) the median risk is often preferable. Both the median and the 95percentile are presented in all risk assessments in this thesis. In Papers I and II the
95% confidence interval is depicted in the figures.
By multiplying this figure with the persons affected, the number of infections in this
population per exposure is achieved (see e.g. Papers II and VI). This transformation
is possible since the dose-response models are ‘population-based’, i.e. based on the
39
Therese Westrell
proportion of subjects that will be infected not the probability that each single
person is infected. This measure may be more useful in risk communication with
the public or specific stakeholders. The risk can also be described as the annual risk
of infection per person or population. This is calculated by taking the number of
yearly exposures into consideration:
Pyearly = 1 − (1 − Pinf )
n
where Pinf is the risk per exposure and n the number of exposures per year. The
annual risk can be valuable when evaluating risk impacts over longer time periods
rather than single exposures, which was the reason for its use in Paper II.
40
Microbial risk assessment in urban water systems
4 Epidemiological concerns
4.1
Susceptibility and immunity
Not all individuals in a population are susceptible to infection. Exposure to a certain
agent may have resulted in whole or partial immunity against that agent. Many
people in Sweden are for example vaccinated against hepatitis A, which may confer
lifelong immunity. Prior exposure to Cryptosporidium appears to increase the
resistance to illness after new exposure (Balbus and Embrey, 2002). Rotaviruses
have previously been believed to provide lifelong immunity. Since in principle all
children have been infected before the age of two (Hrdy, 1987), no adults would
ever be infected. This is however not true (Hrdy, 1987; Svenungsson et al., 2000;
Nakajima et al., 2001) and the role of rotaviruses in adult gastroenteritis is now being
reassessed (Anderson and Weber, 2004).
Recent evidence suggests that not all persons can be infected by certain pathogens.
The ‘ability’ to become infected with noroviruses for example appears to be
dependant on the presence of certain histo-blood types (Huang et al., 2003;
Lindesmith et al., 2003) and thus controlled by genetic factors. While some people,
estimated at approximately 20% of a population, seem to be resistant to infection
(Lindesmith et al., 2003) others have been shown to be re-infected despite high
levels of antibodies (Parrino et al., 1977). The latter effect could also be attributed to
the numerous strains that are in circulation, i.e. immunity only protects for a certain
strain. The ability of a pathogen to evolve new surface structures and become more
infectious or less infectious must also be considered, i.e. a pathogen can change into
new forms that the humoral response does not recognise.
4.2
Sensitive subpopulations
Young children, the elderly, pregnant women and the immunocompromised (organ
transplants, cancer patients, AIDS patients) are more sensitive to infections than
others. The elderly may be less able to build up an effective defence against
microbial or chemical contaminants because of a weakened immune system or preexisting disease. Infants and children are sensitive to microbial and chemical
contaminants because their defence mechanisms may not be fully developed. These
sensitive populations can represent a rather large part of the total population in a
country. In the US for example, this group constitutes almost 20% of the
population and is expected to increase significantly, because of increases in life-span
and the number of immunocompromised individuals (Gerba et al., 1996a).
If risk assessments are based on dose-response models from studies on healthy
adults, they may underestimate the risk for these vulnerable groups. Pouillot et al.
(2004) however made an attempt to include the immunodeficient population when
41
Therese Westrell
assessing the risk from oocysts in drinking water by using dose-response data from
a study on immunosuppressed mice. Parkin et al. (2003) made a qualitative risk
assessment for children exposed to enteroviruses in river water. In Paper IV the
drinking water intake of sensitive subgroups in the population was assessed.
4.3
Secondary transmission
A person infected with a gastrointestinal pathogen excretes the pathogen in her
faeces, regardless of any symptoms of disease. The pathogen can then be
transmitted further to other persons, either through direct contact, aerosols or
through objects contaminated with faeces. Since some pathogens are zoonotic they
can also be further transmitted to animals. Secondary transmission is often specific
to the setting (e.g. different in homes than in hospitals or day-care centres) and also
highly dependent on the infectivity of the pathogen.
4.4
Dynamic modelling
The above-mentioned issues could be taken into consideration in QMRA by using
dynamic models. Dynamic models can incorporate the proportion of individuals in
a population that are either susceptible to infection, infectious and asymptomatic,
infectious and diseased, or immune, and the transition between these states
(Eisenberg et al., 1996) (Figure 4-1). In this way the impact of pathogen exposure on
a whole population can be assessed, including the role of secondary transmission in
sustaining the endemic rate in the population (Chick et al., 2001). Examples on how
to use dynamic models have been presented by Chick et al. (2001) and Eisenberg et
al. (2002) regarding the assessment of the impact of waterborne pathogens from
drinking water and by Eisenberg et al. (2004) regarding the beneficial uses of
biosolids. Such models however require considerable amounts of supplementary
data.
Exposure
No infection
No
symptoms
Infection
Symptoms
Susceptible
Recovery
Death
Immune
FIGURE 4-1. Possible epidemiological states of a person
after the exposure to pathogens
42
Microbial risk assessment in urban water systems
4.5
Health indices
The hazards that may be present in water are associated with a diversity of adverse
health outcomes. While most waterborne infections cause acute symptoms such as
diarrhoea, some such as hepatitis virus have delayed symptoms. There are also
differences in the severity of the symptoms, ranging from mild gastroenteritis to, for
example, the severe haemolytic uremic syndrome (HUS) caused by EHEC or
Gullain-Barré syndrome caused by Campylobacter. In many QMRA the risk of
becoming infected with a specific pathogen after a certain exposure is presented,
however the outcome of these infections are seldom addressed in terms of illness or
fatalities. One way to progress further in the evaluation of health impacts is via the
use of a health index. Several health indices exist (McAlearny et al., 1999), among
them the Disability Adjusted Life Years (DALY).
4.5.1
Disability adjusted life years (DALY)
In 1992 (with the first stages proposed already in 1988), the Global Burden of
Disease (GBD) Study was initiated at the request of the World Bank. The study
represents a unique achievement based on the collaboration of over 100 scientists
from more than 20 countries and describes the world’s disease burden status and
trends in the health of populations. The burden from 107 diseases and injuries and
10 major risk factors or risk groups for various age groups and geographical regions
were described (Murray and Lopez, 1996; Murray and Lopez, 1997).
To allow comparisons to be made between different health outcomes and allowing
quantification of non-fatal outcomes a new unit was introduced: DALY or
Disability Adjusted Life Years. DALYs are the sum of life years lost to premature
mortality and years lived with disability adjusted for severity (Murray and Lopez,
1997). The basic principle of the DALY is to weight each health effect for its
severity from 0 (normal good health) to 1 (death). By multiplying this weight with
the duration of the effect and by the number of people affected by a particular
outcome, it is then possible to sum the effects of all different outcomes due to a
particular agent. In consequence, the DALY is the sum of years of life lost by
premature mortality (YLL) and years of healthy life lost in states of less than full
health, i.e. years lived with a disability (YLD), which are standardised by means of
severity weights. In other words, DALY = YLL + YLD.
Although the choice of values incorporated into DALYs have been debated and
discussed (Arnesen and Nord, 1999) its key advantages lie in its “aggregation” of
different effects and the combining of quality and quantity of life. This metric can
be used to promote and enable the setting of rational public health priorities. As
such it has been proposed by the WHO for comparing the health impact of
different agents in water (WHO, 2004). It can also be used for comparing the health
43
Therese Westrell
effects of microbial and chemical risks, for example in order to compare the risks
from disinfection by-products in ozonation, with the risk from Cryptosporidium
parvum in drinking water (Havelaar et al., 2000a). Furthermore, the use of DALY
may enhance the understanding of risk in risk communications. Havelaar et al.
(2000b) used DALY for assessing the total health burden in the Netherlands due to
infection with Campylobacter species. It could also be useful in a holistic assessment
of water systems and in comparisons of water systems. WHO now promote the use
of 1 µDALY/y as the benchmark health burden (see Section 5.3).
44
Microbial risk assessment in urban water systems
5 Risk management of water systems
5.1
Guidelines and regulations
Guidelines and regulations exist at different levels for protecting the health of the
public when exposed to drinking water, recreational waters, crops amended with
sludge etc. Many regulations are based on critical levels of indicator organisms that
should not be exceeded in order to be classified as acceptable. The frequency of
sampling is usually also regulated. Both the drinking water and the recreational
water regulations in Sweden are set up in this way (SNV, 1996; SLV, 2001). There
are currently no regulations regarding levels of microorganisms in treated
wastewater. Upcoming Swedish regulations on the use of sludge for crop
fertilisation will include treatment recommendations for use on different types of
crop (Schönning, 2003). The current regulations however do not require any sludge
treatment, only mixing of sludge into soil within 24 hours.
The revised WHO Guidelines for Drinking Water Quality (WHO, 2004) promote
protection of drinking water through use of Water Safety Plans, which incorporates
water quality management from the 'source to tap'. A similar approach has also been
proposed in the Bonn Charter for Safe Drinking Water by the IWA (2004). The
approach in the 3rd edition of the WHO guidelines is based on developing healthbased targets. These are based on a consistent framework applicable to all types of
hazards and for all types of water supplies. This approach is thought to be more
flexible than the previous guideline values since it can account for national priorities
and supports a risk–benefit approach. The framework includes different types of
health-based targets that differ considerably with respect to the amount of resources
needed to develop and implement the targets and in relation to the precision with
which the public health benefits of risk management actions can be defined. The
different types of targets are: health outcomes, either based on epidemiology or
quantitative risk assessment; water quality; performance or specified technology.
QMRA has also been incorporated into the WHO Guidelines for Safe Recreational
Waters (WHO, 2003) and will be included in their upcoming guidelines for the safe
use of wastewater and excreta in agriculture.
5.2
Hazard analysis and critical control points (HACCP)
Within food and drinking water production the risk management system Hazard
Analysis and Critical Control Points (HACCP) has been applied (FAO, 1997;
WHO, 2004). HACCP offers a preventative management and quality assurance
approach rather than random monitoring of the end product. The system involves
identification of critical points to control hazards and maintain best management
practices throughout production and distribution. Criteria are established for each
45
Therese Westrell
control point, which are monitored, and corrective actions are established that
should be carried out when critical limits are not met (FAO, 1997).
The working process in the HACCP system consists of seven consecutive steps:
1 Conduct a hazard analysis.
2 Determine the Critical Control Points (location in the process that a certain
hazard can be controlled, either through total prevention, elimination or
reduction).
3 Establish critical limits (a criterion that separates acceptability from
unacceptability, for example a certain temperature, time, moisture level, pH
etc.).
4 Establish a system to monitor control of the CCP.
5 Establish the corrective action to be taken when monitoring indicates that a
particular CCP is not under control.
6 Establish procedures for verification to confirm that the HACCP system is
working effectively.
7 Establish documentation concerning all procedures and records appropriate to
these principles and their application.
Originally developed for control of microbial hazards during space flights, it is now
mandatory within the production and distribution of food. The system facilitates
managing and is compatible with other quality management systems, such as the
ISO 9000-series (FAO, 1997). Current EU legislation decrees the incorporation of
HACCP within drinking water production and it is under implementation in
Swedish waterworks (SLV, 2003). The WHO also incorporates HACCP as part of
their Water Safety Plans (WHO, 2004). Examples on how to use HACCP in
drinking water have been given by Havelaar (1994) and Jagals and Jagals (2004).
As well as controlling microbial risks within drinking water production, HACCP
might also be used for safeguarding the quality of different fractions from
wastewater and sludge reuse. The approach is in many ways similar to that already
utilised in the reuse of wastes within agriculture for reducing the risks for
transmission of diseases, namely use of control measures on different levels: waste
treatment, crop restrictions, localised application methods, control of human
exposure and a combination of the different methods (Blumenthal et al., 1989).
Other researchers have suggested HACCP for water reuse in the food industry
(Casani and Knøchel, 2002) and in a wastewater reuse system including
groundwater recharge and production of drinking water (Dewettinck et al., 2001). It
has recently also been recommended by Water UK for biosolids treatment and use
in agriculture (Water UK, 2004).
The HACCP system contains some of the elements of QMRA, including parts of
hazard identification and exposure assessment, but could benefit from more
46
Microbial risk assessment in urban water systems
contributions from quantitative microbial risk assessment (Notermans et al., 1994;
Haas et al., 1999). In Paper VI this idea was developed by trying to adopt HACCP
to wastewater and sludge handling and reuse. A major difference in producing safe
wastewater or biosolids compared to safe food is that the wastewater already
contains the hazards while in the latter the goal is to prevent contamination. The
focus must therefore be placed on controlling the exposure to wastewater or sludge
and eliminate or reduce the hazards through effective treatment.
The hazards or hazardous exposures must be ranked in order to distinguish between
important and less important hazards for setting priorities in risk management. The
risk associated with each hazard or hazardous event is described in a matrix in
HACCP with the likelihood of occurrence (e.g., certain, possible, rare) and the
severity of consequences if the hazard occurred (e.g., insignificant, major,
catastrophic). In Paper VI the occurrence of a hazardous exposure was instead
included in the risk estimates and a suggestion of severity of consequences based on
the endemic disease in the population was developed (Table 5-1).
TABLE 5-1. Suggested definitions of severity of consequences of hazards based on
increase of endemic disease in the community. Developed for use in the ranking of
hazardous exposures in Paper VI
Item
Definition
Catastrophic
Moderate
Major increase in diarrhoeal disease >25% or >5% increase in more
severe disease or large community outbreak (100 cases) or death
Increase in more severe diseasesa (0.1-5%) or large increase in diarrhoeal
disease (5-<25%)
Increase in diarrhoeal disease (1-<5%)
Minor
Slight increase in diarrhoeal diseases (0.1-<1%)
Insignificant
No increase in disease incidence (<0.1%)
Major
a
Here represented by EHEC
5.3
Tolerable risk
In order to decide a baseline for the infection risks that can be tolerated from a
societal point of view there is a need to set up an ‘acceptable’ or ‘tolerable risk’ level.
The whole concept of acceptable risk can seem controversial. How can we decide
which risks are acceptable? Is it acceptable that people get ill or die due to
pathogens in drinking water? At the same time we have to balance costs against
benefit. Is it defendable or even possible to treat drinking water to the extent that
no microorganisms remain?
In the U.S. the Environmental Protection Agency (USEPA) have accepted a yearly
risk of 1 infected person in 10 000 from drinking water, which is often expressed as
47
Therese Westrell
a risk of 10-4 per person. This level has been used by several researchers in risk
assessments of drinking water and also to compare risks from other types of
exposure (Gerba et al., 1996b; Rose et al., 1996; Tanaka et al., 1998; Jolis et al., 1999).
In the 3rd WHO Drinking Water Guidelines a “reference level” of risk of 10-6
DALYs per person and year is suggested (WHO, 2004). This is approximately
equivalent to a lifetime excess cancer risk of 10-5 (i.e. one excess case of cancer per
100 000 of the population ingesting drinking water containing the substance at the
guideline value over a lifespan). For a pathogen causing watery diarrhoea with a low
case fatality rate (i.e. the proportion of cases for which the disease would be lethal,
e.g. 1 in 100 000), this reference level of risk would be equivalent to 1/1000 annual
risk of disease to an individual (approximately 1/10 over a lifetime).
The acceptable risk level of 1 in 10 000 has been used for the discussion of risks in
Paper I, II, and V and is in Paper III referred to in terms of the virus
concentration in finished drinking water resulting in this risk level. In Paper VI the
approach described in the previous section was used for risk classification and
prioritisation.
48
Microbial risk assessment in urban water systems
6 Short summary of results
6.1 Paper I – Centralised versus decentralised drinking water
treatment
The systems analysis included the evaluation of environmental aspects and
microbial health effects. The system structures are shown in Figure 2-1.
Both air and water emissions from the production and transport of chemicals were
relatively small. The energy use was dominated by the energy needed for
distribution, which was similar for the three systems. The system with two
membranes had approximately the same energy use as the conventional system. The
decentralised system with ultrafiltration could however reduce total energy use with
25% compared to the other systems.
The conventional treatment exceeded the 10-4 median yearly risk of infection
primarily regarding viruses but also regarding protozoa and slightly for bacteria
(Figure 6-1). Both membrane systems reduced the risks of infection among the
consumers substantially, even during partial damages of the membranes (a few
broken capillaries, represented by the upper range on the confidence interval in
Figure 6-1 and 6-2). Possible growth of bacteria in storage tanks however increases
the risk of infection with bacteria.
1
0,01
0,0001
1E-06
1E-08
1E-10
1E-12
Cryptosporidium
Rotavirus
Two
membranes
One
membrane
Conventional
Untreated
Two
membranes
One
membrane
Conventional
Untreated
Two
membranes
One
membrane
Conventional
Untreated
1E-14
FIGURE 6-1. Yearly
risk of infection from
drinking water per
person, showing median
values
and
95%
confidence intervals.
Median values for the
two-step
membrane
alternative
is
zero
regarding Cryptosporidium
and Campylobacter.
Campylobacter
Two consecutive membranes gave a good risk reduction regarding the exposure to
drinking water since the probability of simultaneous failures was low. Regarding
shower exposures, the risks were higher in the second membrane scenario (for
rotaviruses also higher than in the conventional system) than the first due to the fact
that only microfiltration, which has lower removal capacity (pore size 200 nm)
compared to nanofilters (pore size 10 nm), was used for treatment of this water
(figure 6-2).
49
Therese Westrell
1
FIGURE 6-2. Yearly risk
of infection per person
through ingestion of 10
mL water from the shower
once a week. Diagram
shows median values with
95% confidence intervals.
0,01
0,0001
1E-06
1E-08
1E-10
1E-12
Cryptosporidium
6.2
Rotavirus
Two
membranes
One
membrane
Conventional
Untreated
Two
membranes
One
membrane
Conventional
Untreated
Two
membranes
One
membrane
Conventional
Untreated
1E-14
Campylobacter
Paper II – Failures in drinking water treatment and distribution
The major finding in Paper II was that a larger number of people were estimated to
be at risk of becoming infected from pathogens passing the treatment under normal
operating conditions rather than at failure incidents when calculated on a yearly
basis. The incidents in the waterworks were short and did not have a substantial
effect on the yearly risk of infection, although the potential for plug-flows was not
fully considered. Incidents on the distribution network were less common and only
affected parts of the population, however these were of high risk of becoming
infected.
Rotavirus, used as a model for viral transmission, caused the largest number of
potential infections. The most infections with Campylobacter would originate from
contamination of reservoirs in the distribution network. The calculated risk of
Cryptosporidium infections was significantly lower than for the other two pathogens.
The sensitivity analysis pointed out the concentration of pathogens in the raw water,
the chlorination regarding Campylobacter, and the dose-response functions to be the
most critical factors for the risk of infection.
The simulated total number of annual infections in the system was within the range
of the equivalent figures estimated from epidemiological data and fraction of
gastroenteritis attributable to tap water.
6.3
Paper III – Norovirus fluctuations in surface water
Noroviruses were almost exclusively found during the winter season, while
enteroviruses were detected throughout the year with minor peaks in March-April
and October-December. Rotaviruses were only detected once in December the
subsequent year. The levels of noroviruses were several orders of magnitude higher
than those of enteroviruses. All viruses (also bacteriophages) were in low numbers
50
Microbial risk assessment in urban water systems
or non-detectable during the summertime. There was no coincidence in peaks with
F-specific bacteriophages or turbidity and human viruses, although some similarities
in patterns did occur.
The intensified sampling of noroviruses during the winter season revealed that the
winter peak consisted of several shorter peaks of varying duration and magnitude.
The peak concentrations were also several orders of magnitude higher than the
previous year.
Small volume water samples of 10 litres were as efficient in detecting viruses as large
volume water samples of 200-500 litres. They also gave a clearer signal in the gel
and blotting in the undiluted RNA extracts due to less concentration of inhibitors.
Norovirus concentrations could be estimated from the results of dilution series in
the PCR with an MPN approach. The time series analysis was shown to predict
forthcoming sample concentrations accurately and gave a narrower 95% confidence
interval than the MPN estimate. Probability distributions were fitted to the
norovirus concentrations of the whole year and separately for the winter season.
6.4
Paper IV – Drinking water consumption patterns
Water consumption was shown to differ with demographic and socio-economic
factors. Women were for example found to consume higher quantities of cold tap
water than did men (on average 0.95 litres compared to 0.79 litres). Tap water
consumption was higher in the countryside than in the large cities and consumption
was shown to decrease with increasing income. A lognormal distribution was fitted
to the quantitative data from the outbreak investigation representing the whole
population.
Sensitive subgroups were shown to have higher cold tap water consumption than
other groups. The oldest age group, 70 years and above, had the highest daily intake.
Although young children had the lowest intake in volume they will have the highest
intake in relation to their body weight (Forhammar et al., 1986; USEPA, 2000). In
the stratification based on health, people of very poor health had the highest intake.
The heated tap water consumption was somewhat lower than the cold tap water
consumption. Men appeared to consume more water in hot beverages than did
women (some inconsistency in results between data sets). No differences were
detected between age groups, representing 20 years and upward.
Bottled water consumption was generally low with a calculated average of 60 mL
per person and day. A slightly higher proportion of women consumed bottled
water, however no significant differences were found in consumed volumes
51
Therese Westrell
between sexes. The bottled water intake was significantly higher in cities compared
to the countryside and increased with increasing income.
6.5
Paper V – Risks associated with local handling of faeces
In approximately 9 out of 10 gardens, the use of stored faeces as a fertiliser would
not result in a risk of infection because no pathogens were excreted and collected in
the container. One out of 200 containers would contain two pathogens or more.
Rotavirus and Giardia would be the most frequently-occurring pathogens based on
the incidence in the population.
The die-off during storage would be substantial for some of the pathogens, e.g.
Salmonella, while other pathogens, especially Ascaris, have a much higher persistence
in faeces. The pathogen with the most severe symptoms, EHEC, was reduced to
very low levels already during the storage in the toilet and did not constitute any
significant risk in any of the scenarios.
Use of material directly after emptying the toilet container resulted in median risks
exceeding 10-4 for the unconditional scenario regarding rotavirus and the parasites.
After one year of storage however the median risks were below this level for all
pathogens, also in the conditional scenario (i.e. a family member excreting the
pathogen) with the exception of Ascaris. The worst-case risks however exceeded the
level regarding the viruses and parasites.
The exposure to faeces in terms of ingested amounts was lower during recreational
activities or gardening than when emptying the container due to the mixing with
soil, however since the frequency of exposure was higher in the former exposure,
the annual risks were almost as great.
6.6 Paper VI –HACCP for safe handling and reuse of wastewater and
sludge
The highest individual health risk per single exposure was achieved through
exposure to droplets and aerosols for workers at the treatment plant, particularly at
the belt press for sludge dewatering, and through contact with digested sludge for
children or for entrepreneurs when spreading sludge, with a risk of viral infection
nearly or equal to 1. The lowest risk was from recreational swimming with infection
risks of 10-11-10-5 for the range of pathogens evaluated.
The risk entity ‘yearly number of infections’ took the number of people exposed at
each exposure point during a year, into consideration. These values were generally
very low for non-viral pathogens (<<1) although it reached or nearly reached the
52
Microbial risk assessment in urban water systems
maximum number of infections for both adenovirus and rotavirus in the high-risk
exposures mentioned above (which each included two exposed persons). Although
several of the exposures only resulted in fractions of infections they had a large
impact on the community as a whole in terms of increase in endemic disease level in
the community, due to the already low incidence of these pathogens.
Control measures to reduce the hazardous exposure were identified as optimisation
of treatment processes for wastewater, use of personal protective equipment for
staff, change from mesophilic to thermophilic digestion or prolonged sludge storage
times, fencing of sludge storage, and crop restrictions.
53
Microbial risk assessment in urban water systems
7 Discussion
7.1
Use of appropriate data and models
Indicator organisms are used for assessing the faecal contamination of waters. In
Paper III peaks of F-specific bacteriophages and turbidity did not coincide with
those of noroviruses. Since indicator organisms are constantly excreted from
humans and animals they would not reflect peaks in pathogen occurrence that were
due to varying incidence in the human or animal population.
The sensitivity analysis in Paper II pointed out the concentrations of pathogens in
the surface water as one of the most critical factors for the risk of infection. The
concentrations of Campylobacter and rotavirus were based on international data.
During recent sampling in the water source in question no Campylobacter were
detected and the viral contamination seems to be low based on negative findings of
noroviruses and enteroviruses. Cryptosporidium was however found in concentrations
similar to those used in the QMRA. This suggests that the risks of infection with
Campylobacter and rotavirus may have been overestimated in Paper II however they
may be realistic at certain events when the raw water concentrations are higher than
usual. This highlights the importance of availability to site-specific data regarding
the occurrence of pathogens from different sources.
In order to be able to use data compiled from many different studies, the treatment
in the waterworks or wastewater treatment plant in Papers I, II and VI was
modelled as separate distributions for each process. By using separate distributions
it is however assumed that the processes act independently of each other. Hijnen et
al. (2002) concluded that variability in treatment efficiency in the overall treatment
was smaller than expected from the sum of the variation of the processes. This
procedure may therefore result in an overestimation of the treatment variability.
In Paper IV the drinking water consumption in Sweden was assessed. A log-normal
distribution was fitted to one of the data sets for use in future QMRA on drinking
water associated health risks. The fitted distribution was shown to be similar to that
reported by Roseberry and Burmaster (1992) (Figure 7-1) which justified the use of
their model in Papers I and II. The distribution fitted to the Swedish data was
however consistently lower. Besides country-specific differences this effect could be
explained by the fact that the data used by Roseberry and Burmaster included all tap
water consumption and not only the direct tap water consumption. Our model
would therefore be more suitable for the use in QMRA since only the cold tap
water constitutes a risk.
55
Therese Westrell
Cumulative frequency (%)
100
Raw data
80
60
Fitted
lognormal
40
Roseberry &
Burmaster
20
0
0
7.2
FIGURE 7-1. Cumulative
frequencies of daily drinking
water consumption. Results
from Paper IV compared to
the lognormal distribution
of Roseberry and Burmaster
(1992).
500
1000 1500 2000
Volume (mL)
2500
3000
The importance of different agents
Infections with viruses generally constituted the highest risk in Papers I, II and VI.
Only in Paper V where the helminth Ascaris was included did viruses not present
the highest risk, which was due to the maximum-risk dose-response model used for
Ascaris in the absence of other information. While Ascaris is very uncommon in
Sweden, viruses are frequently found in both surface water (Table 3-3) and
wastewater (Table 3-5) and apart from being very infectious, they are also excreted
in high titers from infected persons (Table 3-1). In Paper II it was shown that high
numbers of people could be infected with rotavirus from drinking water. As
mentioned above, the risks may have been overestimated in this case due to the use
of non-specific literature data regarding the virus concentrations in raw waters.
There is also a chance that a large part of the population is already immune to this
pathogen due to early infections in life, however the role of rotavirus in adult
gastroenteritis may be largely underestimated (Anderson and Weber, 2004).
When the dose-response models for norovirus are published, norovirus may be the
virus of choice for assessing risks associated with drinking water due to the high
concentrations detected in sewage and the numerous waterborne outbreaks.
Adenoviruses may be useful as a conservative virus index organism due to its
persistence in the environment and resistance to some treatment methods. Hepatitis
A is uncommon in Sweden but could be appropriate for assessing the more severe
symptoms of diseases associated with reuse of wastewater and sludge, especially in
high endemic areas.
As mentioned previously, Campylobacter causes most waterborne outbreaks of disease
in Sweden where a causative agent has been identified. In Paper II most infections
with Campylobacter jejuni would originate from contamination in the distribution
network, assumed to originate from sewage. Since most bacteria are sensitive to
56
Microbial risk assessment in urban water systems
chlorine, failures in chlorination or contamination in the distribution network with
low chlorine residuals may be the main hazardous events regarding infections with
Campylobacter via drinking water. This is supported by the study of Nygård et al.
(2004a) who, in their analysis of the relationship between infections with
Campylobacter in Sweden and parameters concerning tap water and livestock,
concluded that contamination occurring in the water distribution system may be
more important than previously considered. Regarding the associated health risks
with faeces, wastewater or sludge, Salmonella and EHEC were considered to be
appropriate. Salmonella has a high incidence in the population and are found in high
concentrations in wastewater. EHEC is more rare but causes much more severe
symptoms. Both Salmonella and E. coli have been shown to be able to re-grow in
sludge (Gibbs et al., 1997; Gantzer et al., 2001).
The calculated risk of Cryptosporidium infections was significantly lower than for
rotavirus and Campylobacter in Paper II. This contrasts to the current attention
within the water industry where major efforts are placed on minimising the risk of
Cryptosporidium. One explanation provided above was that the risks with the other
pathogens might have been overestimated. While the concentrations of
Cryptosporidium and Giardia (oo)cysts in Swedish surface waters are similar (Table 33), Giardia is found in much higher concentrations in wastewater (Table 3-5). This
suggests that Cryptosporidium oocysts in surface waters are more likely to originate
from animals than from municipal sewage. It also implies that Giardia may be a
more suitable agent for assessing the risk of protozoa from wastewater fractions,
including faeces.
7.3
Fluctuating pathogen concentrations in surface waters
In Paper III it was shown that the occurrence of human viruses had a seasonal
variation. Noroviruses had a distinct winter peak, which was even higher in the
intensified sampling during the second winter. The latter may be due to the
exceptionally high number of outbreaks in the population during the season of
2002/2003 (Lopman et al., 2004). Turbidity was also markedly increased in January
2003, which may be related to the occurrence of an unusual event, e.g. a heavy
rainfall.
The occurrence of other waterborne pathogens can also be seasonal. Newborn
calves can excrete large numbers of oocysts and spring is therefore a time for large
potential Cryptosporidium parvum contamination. Peak concentrations of both
Cryptosporidium and Giardia (oo)cysts in a watershed have been shown to coincide
with calving activity (Ong et al., 1996). During heavy rainfall and snowmelt, run-off
from pasture or agricultural land can cause substantial contamination of water
bodies (Mawdsley et al., 1995; Kistemann et al., 2002). Such events may also cause
combined-sewer overflows and re-suspension of pathogens from bottom sediments
57
Therese Westrell
(Atherholt et al., 1998). The pathogen load in an Australian river after one heavy
rainfall was estimated to be equivalent to two years of load during ‘baseflow’
conditions (N. Ashbolt, pers. comm.). The need for the development of effective
management strategies at urban run-off during storm events was highlighted by
Jiang and Chu (2004) due to the increased levels of viruses in rivers following such
events.
A question that therefore remains is whether there is a need for additional drinking
water treatment during certain seasons or events to avoid waterborne transmission.
The norovirus concentrations detected in the source water in wintertime was high
(Paper III) and even though some proportion of the viruses may not have been
infectious, there could be a substantial risk for waterborne transmission in case of
insufficient water treatment. In order to reduce the maximum concentration of
noroviruses detected in Paper III to the acceptable virus level in drinking water
suggested by Regli et al. (1991) (<10-7 viruses per litre) subsequent drinking water
treatment must reduce the levels of viruses by 10 log10. The storage reservoirs
following on the point from where the water was sampled have been shown to
reduce the levels of F-specific bacteriophages by 3 log10 during five months (van
Breemen et al., 1998). Additional removal of up to 7 log10 is then still required in the
waterworks. While some discrepancies exist between studies on the efficacy of
chlorine for disinfection of noroviruses (Keswick et al., 1985; Thurston-Enriquez et
al., 2003), these viruses seem to be effectively inactivated with ozone (Shin and
Sobsey, 2003). The ordinary treatment conditions may not however be adjusted to
cope with such fluctuating concentrations in the raw water.
If discharged wastewater is the most likely source of norovirus peaks, their
concentrations in surfaces waters could be reduced by optimising or introducing
additional treatment in the wastewater treatment plant. Soller et al. (2003) used
dynamic disease transmission models to assess the public health benefits of
additional wastewater treatment also in winter time, on the risk of acquiring viral
gastroenteritis from recreational waters. Although the risk already under the existing
treatment scheme was several orders of magnitude below the tolerable illness level
set up by the USEPA, the additional winter treatment would further reduce the risk
by 15-50%.
Another way to proceed in the management of peaks is by introducing an early
warning system. The time series analysis used in Paper III was shown to be useful
in these respects since it can discriminate an increase or decrease in concentration
from random fluctuations. This however requires frequent sampling and rapid
processing of samples to allow for the detection of short-term fluctuations. Time
series analysis also gives the opportunity to control for events, for example by
implementing an action plan to follow when a critical limit is exceeded. This fits
58
Microbial risk assessment in urban water systems
very well into the HACCP system, which nowadays is mandatory for the
management of drinking water production within the EU.
The distribution functions developed for noroviruses in Paper III could also be
valuable in HACCP in order to calculate how probable different virus levels are.
This information could then be used to assess the treatment efficiency needed to
ensure safe drinking water.
7.4
Use of epidemiology in QMRA
Epidemiology can be integrated into QMRA in different ways and serve different
purposes. One is by the use of dynamic models described in Section 4.4 where it
can be used to assess the impact of pathogen exposure on a whole population,
incorporating susceptibility, immunity and secondary transmission.
The ranking of a hazard or hazardous event in HACCP is usually done by
combining the severity of its consequences and the likelihood of its occurrence
(Deere et al., 2001). This was in Paper VI modified for the application in HACCP
of wastewater and sludge treatment and reuse. The ranking was based on the
increase of endemic disease expected in the community after a certain hazardous
exposure and was estimated on a yearly basis (see Table 5-1). As such it could be
applied in different parts of the world with differing endemic rates. It may however
be inappropriate to use in areas where the reuse of untreated wastewater and sludge
already contributes substantially to the endemic rate. The level at which the
exposures were considered to become severe, 0.1% increase in endemic disease, is
also the reference level of disease suggested by the WHO for pathogens causing
watery diarrhoea (see Section 5.3).
As a means to evaluate the plausibility of the results of the QMRA, a comparison
was made in Paper II between the simulated number of annual infections and the
number of infections estimated from epidemiological data. The proportion of
gastrointestinal illness attributable to drinking water used in the calculations, 1440% from the study by Payment et al. (1997), is however debated in the scientific
community. In the first Water Evaluation Trial, out of five initiated by the USEPA
and the Centers for Disease Control and Prevention (CDC), no excess
gastrointestinal illness was detected in households without additional water
treatment at the kitchen tap (Sinclair, 2003b). There were however large differences
in the chlorine residual in the distribution network between the studies, 2.1 mg L-1
in the American study and only 0.5 mg L-1 at the first consumers’ tap in the
Canadian study by Payment et al. The system under study in Paper II had a chlorine
residual of 0.2 mg L-1 when leaving the plant. Since Payment et al. (1997) concluded
that the distribution system appeared to be the major source of contamination, the
59
Therese Westrell
chlorine residual is naturally important in reducing the viability of pathogens
reaching the consumers.
7.5
Microbial health risks in decentralised systems
The centralised treatment of water and wastewater has been the only solution
implemented for cities since the introduction of piped water and sewerage in
Sweden by the end of the 19th century. A question remains as to whether this is this
the only possible way to go. In this thesis the hygienic risks associated with
decentralised treatment of drinking water and local collection of faeces have been
investigated.
In Paper I it was concluded that decentralised drinking water treatment with
membranes was competitive with the centralised conventional treatment regarding
environmental impacts and health. Local drinking water treatment may in one sense
give a more robust infrastructure, since high concentrations of pathogens in the raw
water or sewage ingress and other contamination within the distribution system will
be counteracted by the treatment at the point-of-use.
Failures in decentralised systems will, in contrast to failures in centralised systems
(Paper II), not affect the whole population at the same time. Those failures that
could have an impact in the separate systems were identified as indoor crossconnections and membrane failures. The normally negligible risk in an end-point
membrane system could be increased by several orders of magnitude if the integrity
of the membrane was not maintained, but could still be lower than the risk from the
conventional system. Integrity testing is therefore a key component in membrane
filtration applications however may be less feasible to carry out in many small units.
The number of barriers is also important for safety. As shown in Paper I, the
decentralised system with only ultrafiltration could constitute a high risk for viruses
and bacteria if the membrane was compromised. This highlights the importance of
having several subsequent barriers. On the other hand the risks from exposure to
water in showers was much higher in the membrane scenario with only
microfiltration of the water due to the larger pore size of the membrane.
There is also a risk of cross-connections of water with different qualities within the
buildings. This happened in a dual water supply in the Netherlands where partlytreated surface water was supplied to households for garden watering, toilet flushing
and laundry use. The cross-connection with potable water was not discovered until
after a week and by that time an outbreak of gastroenteritis affecting 200 people had
occurred (Sinclair, 2003a). A very important issue is therefore that of operation and
maintenance. A system with many small treatment units spread over the city may be
difficult to keep under satisfactory control for a centralised organisation, which may
suggest that the management would have to be outsourced to sub-contractors.
60
Microbial risk assessment in urban water systems
The microbial health risks in systems with local collection of faeces are highly
dependent on the incidence of disease in the population, as shown in Paper V.
Wastewater from a large city will always contain pathogens, e.g. in a population of
100 000 people several hundreds would excrete Campylobacter during a year (see
Table 3-1). In systems based on a household level however, most households would
not have any pathogens in their faeces bins during a year, and in households where
a person was infected, the risks for other family members of acquiring an infection
would most likely be higher via other transmission routes, e.g. person-to-person
contact or via fomites.
The study in Paper V was based on a theoretical system, however an existing
system in Sweden has also been evaluated in a similar way (unpublished). In the
latter, in Urban Water called the ‘urban enclave’, the faeces from each household
was emptied into a common outdoor compost (although no composting process
has been observed in terms of temperature increase) with sufficient capacity for
several years of storage. The possibility of long-term storage will render possible
sufficient die-off of potential pathogens in order to reduce risks to tolerable levels.
On the other hand, the bins had to be emptied much more frequently than in the
system in Paper V (on average every month to every third month compared to
once a year), which results in more frequent exposure to potentially hazardous
material.
A fundamental difference between the systems is that the people in the ‘urban
enclave’ are not only exposed to the faeces from their own family but also to those
of thirty other families, which increases the likelihood of exposure to pathogenic
microorganisms when emptying bins or reusing the material in the area. In single
households knowledge about prior diarrhoeal illnesses could entail increased
awareness when handling the faeces – information that would not be available in a
system encompassing several households. If systems like this should be
implemented in urban areas, it is crucial that they are properly operated and
maintained and that routines for handling the hazardous material are set up and
followed.
7.6
Outbreaks versus sporadic cases
In Paper II it was estimated that a larger number of people were at risk of
becoming infected from pathogens passing the treatment under normal operating
conditions rather than at failure incidents. It is unlikely that these cases would be
reported to epidemiological statistics since the cases would be sporadic both in time
and place and the link to drinking water therefore not be evident from a medical
viewpoint (see Figure 1-3). While the incidents in the waterworks did not have a
substantial effect on the yearly risk of infection due to their short duration and
61
Therese Westrell
mixing of water produced during failure with that produced during normal
operating conditions, the incidents in the distribution network were more likely to
cause evident outbreaks. The reasons for this were that affected persons were of
high risk of becoming infected and because a local contamination would affect the
residents in a limited area. This geographical cluster of cases could lead to the
suspicion of a common source.
To conclude, although the risk of becoming infected during a failure event can be
high, the resulting annual number of cases could be higher from tap water produced
during normal operating conditions. This issue is very important when it comes to
risk management and risk minimisation and the question remains as to whether the
focus should be on reduction of the total number of infections or on the prevention
of events that can lead to outbreaks.
Risk experts often consider hazards that affect many people at a single occasion and
hazards that affect many people but sporadically as comparable (Covello, 1983).
From this point-of-view the effort that would have the highest impact on the
number of infections would be preferred. Non-experts however often consider
hazards that affect many people at the same time as being more dangerous (Covello,
1983). It is therefore probable to believe that the public would want to prevent
outbreaks, no matter how small or rare these are.
Considering the impact a waterborne disease outbreak has on society, this may
actually be the right priority. Beside the negative health impact for affected people
and the possibility of lost consumer confidence in the drinking water, waterborne
disease outbreaks also cost the society a substantial amount of money. In the
Milwaukee outbreak in 1993, where 400 000 people were estimated to be affected,
the total costs of illness has been estimated at US $96 million; $32 million in medical
costs and $64 million in productivity losses (Corso et al., 2003). The costs for
outbreak investigations and follow-up, i.e. trying to determine what pathogens and
what technical defects caused the outbreak, how many people were affected, how to
prevent further impact and so on, can also be substantial. Andersson et al. (1997)
estimated the cost of a waterborne outbreak with more than 3 000 cases of
Campylobacter in Sweden to be 4.8 million SEK (US $675 000). The total cost of
waterborne outbreaks in Sweden has not been assessed, however for comparison
the annual cost of foodborne illnesses in Sweden has been estimated at 1 082
million SEK (US $152 million) (Lindqvist et al., 2001).
7.7
Sensitive sub-populations
As mentioned in Section 4.2 sensitive populations can constitute a large part of the
population and are likely to increase (Gerba et al., 1996a). These groups of
individuals would be at the greatest risk of serious illness and mortality from water
62
Microbial risk assessment in urban water systems
and foodborne enteric microorganisms. There may therefore be a need to take these
individuals into special consideration in the risk management of drinking water and
different reuse practices of wastewater and sludge.
In Paper IV it was concluded that the sensitive subpopulations, the young, the
elderly and the sick had higher unheated tap water intake than other groups in
Sweden. If the drinking water is contaminated with pathogens these groups are at
increased risk of becoming infected or to develop more severe symptoms than
would healthy individuals. This may call for a raised concern regarding risk
assessment and management of drinking water systems. Some countries have special
water advice for the immunocompromised population (CDC, 2004) a situation that
has mainly arisen as a result of the high sensitivity to and fatality in cryptosporidiosis
for HIV patients (Aragón et al. 2003). Aragón et al. (2003) found that the proportion
of cases of cryptosporidiosis in AIDS patients attributable to tap water
consumption could be as high as 85%.
Since decentralised treatment of drinking water in Paper I was shown to be feasible
both from a health risk and from environmental effects perspective, additional
water treatment could be provided to sensitive consumer groups such as day-care
centres, nursing homes and hospitals. Such decisions should preferably be made in a
forum of decision makers and the public in general.
Nwachuku and Gerba (2004) suggest that children should be taken into separate
consideration in QMRA. This is because children are of increased risk of infection
to enteric pathogens, but also because they may be more environmentally exposed.
The hand-to-mouth and object-to-mouth contact is much greater among children
than adults (Nwachuku and Gerba, 2004) for example. This was considered in the
QMRA in Papers V and VI where separate soil ingestion rates for children were
used. Exposures were identified specifically for children such as playing at the
sludge storage and more frequent transfer of water-to-mouth in the wetland
compared to adults. Several outbreaks with EHEC from recreational waters have
mainly affected children and the attack rates have been higher when swallowing
water and submerging the head (Feldman et al., 2002). The same observation was
made at the outbreak with noroviruses in Gothenburg in the summer of 2004.
7.8 Application of HACCP in the management of wastewater and
biosolids treatment and reuse
HACCP was recently proposed for the management of hazards in biosolids reuse
on agricultural land in the UK (Water UK, 2004). The proposed procedure covers
different types of hazards, not only microbiological. The focus is however only on
the end product and does not include the management of hazards for staff or other
63
Therese Westrell
people involved in the treatment and handling of biosolids. These issues were
however addressed in Paper VI.
In Paper VI the calculated risks of infection of staff working at the wastewater
treatment plant were exceptionally high. Although staff may have acquired
immunity to some of the pathogens encountered in wastewater, this should not be a
reason for accepting high risks. Gastrointestinal symptoms, airway symptoms, joint
pains, unusual tiredness and toxic pneumonitis have been found in significantly
higher frequency among operational staff at sewage treatment plants in Sweden
compared to controls (Thorn et al., 2002; Thorn and Beijer, 2004). Although some
of these symptoms may be associated with exposure to endotoxins, somewhat
higher antibody levels to adenovirus and enterovirus were also found (Thorn and
Beijer, 2004). Risk reducing measures would however be relatively easily to
implement.
The reuse of sludge or biosolids in agriculture resulted in low risks from crops in
Paper VI. The use of treated sludge in agriculture probably constitutes lower
microbial hazards than the common use of animal slurry and manure, however
chemical hazards are most likely higher in sewage sludge. To be on the safe side
when reusing wastewater fractions, risks should be reduced substantially at
treatment to ensure a safe product for further handling. For this to be possible,
wastewater treatment plants must be designed such that the desired effluent quality
is consistently achieved (Cooper, 1991).
The ranking of hazardous exposures based on increase in endemic disease was
found useful in setting the risks in a public health perspective. The amounts and
frequencies of exposure were however uncertain in several of the exposures and
need further investigation. The QMRA and HACCP procedure was in Paper VI
applied to conditions of normal operation of the treatment plant. Worst-case
scenarios or hazardous events need to be further evaluated in order to propose a
final management system. Such events could be flooding, a major failure in the
wastewater treatment or sudden peaks based on treatment variability. The remaining
steps in HACCP should also be addressed with the participation of the managers
and staff of the treatment plant.
7.9
Risk communication – experts and users
As mentioned in Section 1.2 people generally regard exposures to involuntary risks
as less acceptable than voluntary risks. Considering this, people are less likely to
accept hazards in their drinking water since that is something they cannot affect and
since they are dependent on drinking water for their daily life. People may on the
other hand accept high microbial risks from a system they have chosen themselves,
as seen for example regarding the faeces collection in the ‘urban enclave’ mentioned
64
Microbial risk assessment in urban water systems
above. In that system the users appeared to be aware of and take precautionary
measures to avoid some of the obvious hazardous exposures such as when
emptying bins (H. Krantz, pers. comm.). Other hazardous exposures identified by the
risk assessor were however not considered such as children playing at the faeces
compost or the occurrence of flies in some apartments originating from the faeces.
Although QMRA require a lot of effort in data collection and skills in modelling,
the communication of risks to the public may be the most difficult task in risk
analysis. The way that experts use quantitative assessments, with advanced
computer programs and models, to identify, estimate and assess risks is very
different from the intuitive risk assessment or risk perceptions that people use in
everyday life (Slovic, 1987). The question is how experts and the public should
communicate with each other when they use such different languages and methods.
One possible way to proceed regarding the communication of microbial risks is by
the use of health indices, such as DALYs (described in Section 4.5) since the
number of years lost or lived with disability gives more information on the severity
of a hazard compared to measures of infection risk. In the Urban Water program
multi-criteria decision aid is proposed for use in decision making of urban water
systems involving different stakeholders (Söderberg and Kain, 2002; van Moeffaert,
2003). These should take all sustainability aspects into consideration, out of which
health is one of the most important.
65
Microbial risk assessment in urban water systems
8 Conclusions
The most important conclusions in relation to the formulated research questions
were:
• Decentralised drinking water treatment with membranes can be competitive
with centralised conventional treatment regarding environmental impacts and
health. Even in systems encompassing treatment processes with very high
removal efficiency such as membranes, more than one barrier is required to
minimise health risks in case of failures. In order to safeguard the health of
consumers in decentralised systems it is crucial that system operation and
management is well established.
• Failures in drinking water distribution involving microbial contamination are
likely to result in waterborne disease outbreaks. Short-term failures in
treatment may not have a large impact on the health of consumers, especially
if the water produced during failure is mixed with water produced during
normal operating conditions. Averaged over a year, the total number of
people at risk of becoming infected with pathogens from drinking water
could be higher from water produced during normal operating conditions
than during failures.
• The concentrations of noroviruses in surface water can be estimated by a
most-probable-number approach. These viruses show substantial fluctuations
in concentration over the year with peak concentrations in the wintertime.
These peaks could be further resolved into smaller peaks, possibly resulting
from outbreaks in the human population or different types of events.
Fluctuating concentrations of noroviruses or other pathogens could be
predicted by time series analysis and used as an early warning system if
complemented by regular monitoring.
• The direct, or cold, tap water consumption in Sweden differs between groups
in the population and to that reported from several other countries. Groups
already sensitive to infection, i.e. the elderly, the sick and children, consume
higher volumes of cold tap water than the rest of the population. This may
call for special attention in the risk management of drinking water systems.
• The infection risks associated with local handling and reuse of human faeces
is highly dependent on the incidence of infectious diseases in the population
and would be low for the majority of households in Denmark and Sweden.
In order to reduce risks to acceptable levels over the system as a whole,
sufficient die-off of the most persistent pathogens must be secured, which
67
Therese Westrell
requires facilities for the long-term storage of locally-collected faeces in case
no other treatment is provided. Without the assurance of proper maintenance
and following of directions, these kinds of systems should not be
recommended on a larger scale, especially if the material is to be applied to
soil.
• Microbial health risks associated with the handling and reuse of wastewater
and sludge can be successfully addressed and controlled within the
management system Hazard Analysis and Critical Control Points (HACCP).
The infection risks of hazards and hazardous exposure pathways can be put
into an epidemiological framework in order to facilitate decision-making
regarding hazard prioritisation. Many exposure pathways can be controlled
through easy measures. Worst-case scenarios or hazardous events need to be
included in a final risk management system.
QMRA has the potential to be used in many different contexts in the future, not the
least in the comparison between different water and wastewater systems. Although
many important questions have been resolved in this thesis many others remain to
be answered. One for example is the resolution needed in the models to make
accurate estimations of microbial health risks. As the need for underlying data is
great the field would benefit from international co-operation and data exchange. On
the other hand, site-specific pathogen monitoring was shown to be very important
for the accurate estimation of risks. The implementation of new solutions of water
and wastewater/sanitation systems was also shown to incorporate new exposure
pathways to pathogens. These need further attention regarding where, when and
how often they would occur and whom they will affect.
68
Microbial risk assessment in urban water systems
9 Acknowledgements
I know that many people were surprised when they heard that someone who is
personifying the Swedish expression “a little dirt cleans the stomach” would do a
PhD in microbial risk assessment. To all of you I can say that I have realised the
importance of protecting the population, and especially sensitive persons, from
unwanted and unnecessary hazards. I am today also willing to recognise the
importance of throwing out food that consists of more mould than original
contents...
There are many persons that in different ways have contributed to this thesis with
knowledge, ideas and fruitful discussions, but not the least with encouraging words
and actions. I would especially like to thank the following persons:
Thor Axel Stenström for supervising a stubborn PhD student like me. You have an
exceptional ability of producing good ideas and have supported me when I needed
it the most. My co-supervisor Nick Ashbolt who came up with the original project
idea. Thank you for bringing the latest news from the risk assessment world during
you visits to Sweden. It has been inspiring to follow your ‘evolution of ideas’.
The water section at SMI, a group of people with enormous hearts, for always
caring and sharing. Görel, Anette, Emma, Anna and Sabina, you are the backbones of
our section and the scouts of pathogens in water - always prepared! Jakob, Caroline,
Anneli, Annika and Johan are much appreciated for doing the other strange stuff
involving the ‘not-so-clean’ samples. It’s a dirty job but someone’s got to do it! To
my roommates Jonas and Mick for freezing and fighting together with me in (I quote
Uffe) “the best room at the whole of SMI”. A special thanks to Mick for his
invaluable language assistance. I also appreciate the pleasant company at lunches
and coffee breaks of other people at SMI and MTC.
My home institution, the department of Water and Environmental Studies at
Linköping University, for providing an open and stimulating atmosphere. My
deepest gratitude to Ian -‘The rocker’- Dickson. Without your computer aid this
thesis would certainly have disappeared into cyberspace a long time ago. My fellow
PhD students in Linköping and in the Urban Water program for making this
journey a lot easier, more enjoyable and more interesting. A particularly big hug to
Helena K who always lends an ear when I want to complain about the life as a
graduate student in general or discuss thoughts about life in particular. When shall
we publish our rhymes?
Peter Teunis for his mathematical superiority and humbleness. Ana Maria de Roda
Husman for her enthusiasm and for being full of new ideas to try out. Thank you
both for letting me work with you at RIVM, I had a great time!
69
Therese Westrell
I would like to acknowledge the financial support provided by the Swedish research
foundation MISTRA within the Sustainable Urban Water Management Program,
the Swedish Research Council for Environment, Agricultural Science and Spatial
Planning (Formas) and the EU project MicroRisk (contract EVK1-CT-2002-00123).
Without their contributions this project would not have been possible. I have
especially appreciated the possibilities for scientific exchange at international
conferences and national and international collaborations. Per-Arne Malmqvist and
Henriette Söderberg have done a great job running the whole Urban Water program.
Thank you both and also Jan-Olof Drangert for letting us study such interesting
subjects in the research school. The mega city travels have been enormously
rewarding and clearly gives another perspective on the conditions of urban water in
Sweden compared to other countries.
Rickard who convinced me to apply for the PhD position in the first place and
always came up with solutions to my problems. My friends that made sure that I
had causes for rejoicing on the side of my studies.
And last a great thanks to my family for all their love. You have inspired me with
the feeling that anything is possible! I love you!
70
Microbial risk assessment in urban water systems
10
References
Ackman, D., Marks, S., Mack, P., Caldwell, M., Root, T. and Birkhead, G. (1997). Swimmingassociated haemorrhagic colitis due to Escherichia coli O157:H7 infection: evidence of
prolonged contamination of a fresh water lake. Epidemiology and Infection 119(1): 1-8.
Amahmid, O., Asmama, S. and Bouhoum, K. (1999). The effect of waste water reuse in irrigation
on the contamination level of food crops by Giardia cysts and Ascaris eggs. International
Journal of Food Microbiology 49: 19-26.
Anderson, E. J. and Weber, S. G. (2004). Rotavirus infection in adults. Lancet Infectious Diseases 4(2):
91-99.
Anderson, J. M. (1996). Current water recycling initiatives in Australia: Scenarios for the 21st
century. Water Science and Technology 33(10-11): 37-43.
Andersson, Y. and Bohan, P. (2001). Disease surveillance and waterborne outbreaks. In: Water
quality: Guidelines, standards and health. L. Fewtrell and J. Bartram (ed.). London, UK,
IWA Publishing.
Andersson, Y., Jong, B. d. and Studahl, A. (1997). Waterborne Campylobacter in Sweden: the cost of
an outbreak. Water Science and Technology 35(11-12): 11-14.
Anonymous (2004). Smittskyddslag. (The Swedish Infectious Disease Protection Act). SFS 2004:168.
Aragón, T. J., Novotny, S., Enanoria, W., Vugia, D. J., Khalakdina, A. and Katz, M. H. (2003).
Endemic cryptosporidiosis and exposure to municipal tap water in persons with acquired
immunodeficiency syndrome (AIDS): a case-control study. BMC Public Health 3(1): 2.
Armon, R., Gold, D., Brodsky, M. and Oron, G. (2002). Surface and subsurface irrigation with
effluents of different qualities and presence of Cryptosporidium oocysts in soil and on crops.
Water Science and Technology 46(3): 115-122.
Arnbjerg-Nielsen, K., Hansen, L., Kjølholt, J., Stuer-Lauridsen, F., Hasling, A. B., Hansen, N. J.,
Stenström, T. A., Schönning, C., Westrell, T., Carlsen, A. and Halling-Sørensen, B. (2004).
Risikovurdering af anvendelse af helt eller delvist opsamlet komposteret human fæces i
private havebrug. (Risk assessment of partly composted faeces for use in private gardens),
Miljøstyrelsen, Denmark.
Arnesen, T. and Nord, E. (1999). The value of DALY life: problems with ethics and validity of
disability adjusted life years. British Medical Journal 319: 1420-1425.
Asano, T. and Sakaji, R. H. (1990). Virus risk analysis in wastewater reclamation and reuse. In:
Chemical Water and Wastewater Treatment. H. H. Hahn and R. Klute (ed.). Berlin
Heidelberg, Springer-Verlag: 483-496.
Ashbolt, N. J., Grabow, W. O. K. and Snozzi, M. (2001). Indicators of microbial water quality. In:
Water quality: Guidelines, standards and health. L. Fewtrell and J. Bartram (ed.). London,
UK, IWA Publishing: 289-315.
Ashbolt, N. J., Petterson, S. E., Roser, D. J., Westrell, T., Ottoson, J., Schönning, C. and
Stenström, T. A. (2004). Microbial risk assessment tool to aid in the selection of
sustainable urban water systems. Proceedings from 2nd IWA Leading -Edge Conference
on Sustainability in Water Limited Environments, 8 - 10 November 2004, Sydney,
Australia.
Ashbolt, N. J., Riedy, C. and Haas, C. N. (1997). Microbial health risk at Sydney's coastal bathing
beaches. Proceedings from the 17th Federal Convention of AWWA, Melbourne.
Atherholt, T. B., LeChevallier, M. W., Norton, W. D. and Rosen, J. S. (1998). Effect of rainfall on
Giardia and Crypto. Journal of American Water Works Association 90(9): 66-80.
AWWA (1999). Manual of water supply practices - M48: Waterborne Pathogens. Denver,
American Water Works Association.
Balbus, J. M. and Embrey, M. A. (2002). Risk factors for waterborne enteric infections. Current
Opinion in Gastroenterology 18(1): 46-50.
71
Barwick, R. S., Levy, D. A., Craun, G. F., Beach, M. J. and Calderon, R. L. (2000). Surveillance for
waterborne-disease outbreaks – United States, 1997-1998. Morbidity and Mortality Weekly
Report 49(SS-4): 1-35.
Bitton, G. (1994). Wastewater microbiology. New York, Wiley-Liss.
Black, R. E., Levine, M. M., Clements, M. L., Hughes, T. P. and Blaser, M. J. (1988). Experimental
Campylobacter jejuni infection in humans. Journal of Infectious Diseases 157(3): 472-479.
Blackburn, B. G., Craun, G. F., Yoder, J. S., Hill, V., Calderon, R. L., Chen, N., Lee, S. H., Levy,
D. A. and Beach, M. J. (2004). Surveillance for waterborne-disease outbreaks associated
with drinking water – United States, 2001-2002. MMWR Surveillance Summaries 53(8): 23-45.
Blumenthal, U. J., Cifuentes, E., Bennett, S., Quigley, M. and Ruiz-Palacios, G. (2001). The risk of
enteric infections associated with wastewater reuse: the effect of season and degree of
storage of wastewater. Transactions of the Royal Society of Tropical Medicine and Hygiene 95(2):
131-137.
Blumenthal, U. J., Strauss, M., Mara, D. D. and Cairncross, S. (1989). Generalised model of the
effect of different control measures in reducing health risks from waste reuse. Water Science
and Technology 21: 567-577.
Bofill-Mas, S., Pina, S. and Girones, R. (2000). Documenting the epidemiologic patterns of
polyomaviruses in human populations by studying their presence in urban sewage. Applied
and Environmental Microbiology 66(1): 238-245.
Bruce, M. G., Curtis, M. B., Payne, M. M., Gautom, R. K., Thompson, E. C., Bennett, A. L. and
Kobayashi, J. M. (2003). Lake-associated outbreak of Escherichia coli O157:H7 in Clark
County, Washington, August 1999. Archives of Pediatrics and Adolescent Medicine 157(10):
1016-1021.
CAC (1999). Principles and Guidelines for the Conduct of Microbiological Risk Assessment.
CAC/GL-30. Codex Alimentarius Commission (FAO/WHO).
Carey, C. M., Lee, H. and Trevors, J. T. (2004). Biology, persistence and detection of
Cryptosporidium parvum and Cryptosporidium hominis oocyst. Water Research 38(4): 818-862.
Carlander, A. (2002). Occurrence, reduction and risk of faecal pathogen transmission during
wastewater irrigation of willow vegetation filters. Licentiate thesis at the Department of
Short Rotation Forestry, Uppsala University, Uppsala.
Carlander, A. and Stenström, T. (2001). Irrigation with pretreated wastewater on short rotation
willow croppice - a sanitary study in Sweden. Proceedings from the International
Ecological Engineering Conference, Cristchurch, New Zealand.
Carlander, A. and Westrell, T. (1998). A microbiological and sociological evaluation of urinediverting double-vault latrines in Cam Dum, Vietnam. Uppsala, Swedish University of
Agricultural Sciences.
Carrique-Mas, J. (2001). Cryptosporidieinfektion i Sverige - toppen på ett isberg? (Cryptosporidium
infections in Sweden - the tip of an iceberg?) Smittskydd (1): 6-7.
Carrique-Mas, J., Andersson, Y., Petersen, B., Hedlund, K., Sjögren, N. and Giesecke, J. (2003). A
Norwalk-like virus waterborne community outbreak in a Swedish village during peak
holiday season. Epidemiology and Infection 131(1): 737-744.
Casani, S. and Knøchel, S. (2002). Application of HACCP to water reuse in the food industry.
Food Control 13: 315–327.
CDC (2004). Preventing cryptosporidiosis: a guide for people with compromised immune
systems. Fact sheet from U.S. Centers for Disease Control and Prevention.
Chauret, C., Sprinthorpe, S. and Sattar, S. (1999). Fate of Cryptosporidium oocysts, Giardia cysts, and
microbial indicators during wastewater treatment and anaerobic sludge digestion. Canadian
Journal of Microbiology 45(3): 257-262.
72
Microbial risk assessment in urban water systems
Chick, S. E., Koopman, J. S., Soorapanth, S. and Brown, M. E. (2001). Infection transmission
system models for microbial risk assessment. The Science of the Total Environment 274(1-3):
197-207.
Cifuentes, E. (1998). The epidemiology of enteric infections in agricultural communities exposed
to wastewater irrigation: perspectives for risk control. International Journal of Environmental
Health Research 8(3): 203-213.
Coleman, M. E. and Marks, H. M. (1999). Qualitative and quantitative risk assessment. Food
Control 10: 289-297.
Coleman, M. E., Marks, H. M., Golden, N. J. and Latimer, H. K. (2004). Discerning strain effects
in microbial dose-response data. Journal of Toxicology and Environmental Health-Part a-Current
Issues 67(8-10): 667-685.
Colley, D. G. (1996). Widespread foodborne cyclosporiasis outbreaks present major challenges.
Emerging Infectious Diseases 2(4): 354-356.
Cooper, R. C. (1991). The hygienic aspects of wastewater reuse. Waste Management and Research 9:
373-377.
Corso, P. S., Kramer, M. H., Blair, K. A., Addiss, D. G., Davis, J. P. and Haddix, A. C. (2003).
Cost of illness in the 1993 waterborne Cryptosporidium outbreak, Milwaukee, Wisconsin.
Emerging Infectious Diseases 9(4): 426-431.
Couch, R. B., Cate, T. R., Gerone, P. J., Fleet, W., Lang, D., Griffith, W. and Knight, V. (1966).
Production of illness with a small-particle aerosol of adenovirus type 4. Bacteriological
Reviews 30: 517-528.
Covello, V. T. (1983). The perception of technological risks: a literature review. Technological
Forecasting and Social Change 23: 285-297.
Crabtree, K. D., Gerba, C. P., Rose, J. B. and Haas, C. N. (1997). Waterborne adenoviruses: a risk
assessment. Water Science and Technology 35(11-12): 1-6.
Crockett, C. S., Haas, C. N., Fazil, A., Rose, J. B. and Gerba, C. P. (1996). Prevalence of shigellosis
in the U.S.: consistency with dose-response information. International Journal of Food
Microbiology 30: 87-99.
Cutter, S. L. (1993). Living with risk: the geography of technological hazards. London, Edward
Arnold.
de Motes, C. M., Clemente-Casares, P., Hundesa, A., Martin, M. and Girones, R. (2004).
Detection of bovine and porcine adenoviruses for tracing the source of fecal
contamination. Applied and Environmental Microbiology 70(3): 1448-1454.
Deere, D., Stevens, M., Davison, A., Helm, G. and Dufour, A. (2001). Management strategies. In:
Water Quality: Guidelines, Standards and Health. L. Fewtrell and J. Bartram (ed.).
London, IWA publishing: 257-288.
DeRegnier, D. P., Cole, L., Schupp, D. G. and Erlandsen, S. L. (1989). Viability of Giardia cysts
suspended in lake, river, and tap water. Applied and Environmental Microbiology 55(5): 11231229.
Dewettinck, T., van Houtte, E., Geenens, D., van Hege, K. and Verstraete, W. (2001). HACCP
(Hazard Analysis and Critical Control Points) to guarantee safe water reuse and drinking
water production - a case study. Water Science and Technology 43(12): 31-38.
Doller, P. C., Dietrich, K., Filipp, N., Brockmann, S., Dreweck, C., Vonthein, R., WagnerWiening, C. and Wiedenmann, A. (2002). Cyclosporiasis outbreak in Germany associated
with the consumption of salad. Emerging Infectious Diseases 8(9): 992-994.
DuPont, H. L., Chappel, C. L., Sterling, C. R., Okhuysen, P. C., Rose, J. B. and Jakubowski, W.
(1995). The infectivity of Cryptosporidium parvum in healthy volunteers. The New England
Journal of Medicine 332(13): 855-859.
DuPont, H. L., Hornick, R. B., Dawkins, A. T., Snyder, M. J. and Formal, S. B. (1969). The
response of man to virulent Shigella flexneri 2a. Journal of Infectious Diseases 119(3): 296-299.
73
DuPont, H. L., Hornick, R. B., Snyder, M. J., Libonati, J. P., Formal, S. B. and Gangarosa, E. J.
(1972). Immunity in shigellosis. II. Protection induced by oral live vaccine or primary
infection. Journal of Infectious Diseases 125(1): 12-16.
Eisenberg, J. N., Brookhart, M. A., Rice, G., Brown, M. and Colford, J. M., Jr. (2002). Disease
transmission models for public health decision making: analysis of epidemic and endemic
conditions caused by waterborne pathogens. Environmental Health Perspectives 110(8): 783790.
Eisenberg, J. N., Seto, E. Y. W., Olivieri, A. W. and Spear, R. C. (1996). Quantifying water
pathogen risk in an epidemiological framework. Risk Analysis 16(4): 549-563.
Eisenberg, J. N. S., Soller, J. A., Scott, J., Eisenberg, D. M. and Colford, J. M. (2004). A dynamic
model to assess microbial health risks associated with beneficial uses of biosolids. Risk
Analysis 24(1): 221-236.
Enriquez, C. E., Hurst, C. J. and Gerba, C. P. (1995). Survival of the enteric adenoviruses 40 and
41 in tap, sea, and waste water. Water Research 29(11): 2548-2553.
FAO (1997). Hazard analysis and critical control point (HACCP) system and guidelines for its
application. Annex to CAC/RPC 1-1969, Rev. 3 (1997), Amendment 1 (1999).
Fattal, B., Wax, Y., Davies, M. and Shuval, H. I. (1986). Health risks associated with wastewater
irrigation: an epidemiological study. American Journal of Public Health 76(8): 977-979.
Feachem, R. G., Bradley, D. J., Garelick, H. and Mara, D. (1983). Sanitation and disease - health
aspects of excreta and wastewater management. Chichester, Wiley & sons.
Feldman, K. A., Mohle-Boetani, J. C., Ward, J., Furst, K., Abbott, S. L., Ferrero, D. V., Olsen, A.
and Werner, S. B. (2002). A cluster of Escherichia coli O157: nonmotile infections associated
with recreational exposure to lake water. Public Health Reports 117(4): 380-385.
Ferguson, C., de Roda Husman, A. M., Altavilla, N., Deere, D. and Ashbolt, N. (2003). Fate and
transport of surface water pathogens in watersheds. Critical Reviews in Environmental Science
and Technology 33(3): 299-261.
Finley, B., Proctor, D., Scott, P., Harringtoin, N., Paustenbach, D. and Price, P. (1994).
Recommended distributions for exposure factors frequently used in health risk
assessment. Risk Analysis 14(4): 533-551.
Forhammar, M., Bruce, Å. and Erlandsson, B. (1986). Dricksvattenkonsumtionen i olika åldrar.
(Drinking water consumption by different age groups). Vår Föda 38(1): 86-91.
Frey, H. and Burmaster, D. E. (1999). Methods for characterizing variability and uncertainty:
comparison of bootstrap simulation and likelihood-based approaches. Risk Analysis 19(1):
109-130.
Frey, H. C. and Patil, S. R. (2002). Identification and review of sensitivity analysis methods. Risk
Analysis 22(3): 553-578.
Gale, P. (2000). Risk assessment model for a waterborne outbreak of cryptosporidiosis. Water
Science and Technology 41(7): 1-7.
Gale, P. (2003). Using event trees to quantify pathogen levels on root crops from land application
of treated sewage sludge. Journal of Applied Microbiology 94(1): 35-47.
Gannon, V. P., Graham, T. A., Read, S., Ziebell, K., Muckle, A., Mori, J., Thomas, J., Selinger, B.,
Townshend, I. and Byrne, J. (2004). Bacterial pathogens in rural water supplies in
Southern Alberta, Canada. Journal of Toxicology and Environmental Health A 67(20-22): 16431653.
Gantzer, C., Gaspard, P., Galvez, L., Huyard, A., Dumouthier, N. and Schwartzbrod, J. (2001).
Monitoring of bacterial and parasitological contamination during various treatment of
sludge. Water Research 35(16): 3763-3770.
Gerba, C. P., Rose, J. B. and Haas, C. N. (1996a). Sensitive populations: who is at the greatest
risk? International Journal of Food Microbiology 30: 113-123.
74
Microbial risk assessment in urban water systems
Gerba, C. P., Rose, J. B., Haas, C. N. and Crabtree, K. D. (1996b). Waterborne rotavirus: a risk
assessment. Water Research 30(12): 2929-2940.
Gibbs, R. A., Hu, C. J., Ho, G. E. and Unkovich, I. (1997). Regrowth of faecal coliforms and
salmonellae in stored biosolids and soil amended with biosolids. Water Science and Technology
35(11-12): 269-275.
Graham, D. Y., Jiang, X., Tanaka, T., Opekun, A. R., Madore, H. P. and Estes, M. K. (1994).
Norwalk virus infection of volunteers: new insights based on improved assays. Journal of
Infectious Diseases 170(1): 34-43.
Guendert, D. (2004). Singapore success. Water21 June 2004: 23-23.
Haas, C. N. (1983). Estimation of risk due to low doses of microorganisms: a comparison of
alternative methodologies. American Journal of Epidemiology 118(4): 573-582.
Haas, C. N., Crockett, C. S., Rose, J. B., Gerba, C. P. and Fazil, A. M. (1996). Assessing the risk
posed by oocysts in drinking water. Journal of American Water Works Association 88(9): 131136.
Haas, C. N., Rose, J. B., Gerba, C. and Regli, S. (1993). Risk assessment of virus in drinking water.
Risk Analysis 13(5): 545-552.
Haas, C. N., Rose, J. B. and Gerba, C. P. (1999). Quantitative microbial risk assessment. New
York, John Wiley & Sons, Inc.
Haas, C. N., Thayyar-Madabusi, A., Rose, J. B. and Gerba, C. P. (2000). Development of a doseresponse relationship for Escherichia coli O157:H7. International Journal of Food Microbiology
56(2-3): 153-159.
Hansen, A. and Stenström, T.-A. (1998). Kartläggning av Giardia och Cryptosporidium i svenska
ytvattentäkter. (Mapping of Giardia and Cryptosporidium in Swedish surface water sources). The
Swedish Institute for Infectious Disease Control (SMI) and the Swedish National Food
Administration (SLV).
Hansen, J. S. and Ongerth, J. E. (1991). Effects of time and watershed characteristics on the
concentration of Cryptosporidium oocysts in river water. Applied and Environmental
Microbiology 57(10): 2790-2795.
Hardy, M. E. (1999). Norwalk and "Norwalk-like viruses" in epidemic gastroenteritis. Clinics in
Laboratory Medicine 19(3): 675-690.
Havelaar, A. H. (1994). Application of HACCP to drinking water supply. Food Control 5(3): 145152.
Havelaar, A. H., de Hollander, A. E., Teunis, P. F., Evers, E. G., van Kranen, H. J., Versteegh, J.
F., van Koten, J. E. and Slob, W. (2000a). Balancing the risks and benefits of drinking
water disinfection: disability adjusted life-years on the scale. Environmental Health Perspectives
108(4): 315-321.
Havelaar, A. H., de Wit, M. A. S. and van Koningsveld, R. (2000b). Health burden in the
Netherlands (1900-1995) due to infections with thermophilic Campylobacter species.
Bilthoven, RIVM.
Hellard, M. E., Sinclair, M. I., Forbes, A. B. and Fairley, C. K. (2001). A randomized, blinded,
controlled trial investigating the gastrointestinal health effects of drinking water quality.
Environmental Health Perspectives 109(8): 773-778.
Hijnen, W. A. M., Beerendonk, E., Smeets, P. and Medema, G. J. (2004). Elimination of microorganisms by water treatment processes. Nieuwegein, the Netherlands, Kiwa Water
Research.
Hijnen, W. A. M., Medema, G. J. and van der Kooij, D. (2002). Quantitative assessment of the
removal of indicator bacteria by full-scale treatment. Proceedings from the International
Symposium on Waterborne Pathogens, Cascais/Lisbon, Portugal, IWA/AWWA.
75
Hoebe, C. J., Vennema, H., de Roda Husman, A. M. and van Duynhoven, Y. T. (2004). Norovirus
outbreak among primary schoolchildren who had played in a recreational water fountain.
Journal of Infectious Diseases 189(4): 699-705.
Hrdy, D. B. (1987). Epidemiology of rotaviral infection in adults. Reviews of Infectious Diseases 9(3):
461-469.
Hrudey, S. E., Payment, P., Huck, P. M., Gillham, R. W. and Hrudey, E. J. (2003). A fatal
waterborne disease epidemic in Walkerton, Ontario: comparison with other waterborne
outbreaks in the developed world. Water Science and Technology 47(3): 7-14.
Huang, P., Farkas, T., Marionneau, S., Zhong, W., Ruvoen-Clouet, N., Morrow, A. L., Altaye, M.,
Pickering, L. K., Newburg, D. S., LePendu, J. and Jiang, X. (2003). Noroviruses bind to
human ABO, Lewis, and secretor histo-blood group antigens: identification of 4 distinct
strain-specific patterns. Journal of Infectious Diseases 188(1): 19-31.
Hutchison, M. L., Walters, L. D., Avery, S. M., Synge, B. A. and Moore, A. (2004). Levels of
zoonotic agents in British livestock manures. Letters in Applied Microbiology 39(2): 207-214.
Höglund, C., Stenström, T. A. and Ashbolt, N. (2002). Microbial risk assessment of sourceseparated urine used in agriculture. Waste Management and Research 20(2): 150-161.
Höller, C. (1988). Quantitative and qualitative investigations on Campylobacter in the sewage system
of a big town. Zentralblatt Fur Bakteriologie Mikrobiologie Und Hygiene Serie B-Umwelthygiene
Krankenhaushygiene Arbeitshygiene Praventive Medizin 185(4-5): 307-325.
Höller, C., Koschinsky, S. and Witthuhn, D. (1999). Isolation of enterohaemorrhagic Escherichia
coli from municipal sewage. Lancet 353(9169): 2039-2039.
Hörman, A., Rimhanen-Finne, R., Maunula, L., von Bonsdorff, C. H., Torvela, N., Heikinheimo,
A. and Hänninen, M. L. (2004). Campylobacter spp., Giardia spp., Cryptosporidium spp.,
noroviruses, and indicator organisms in surface water in southwestern Finland, 2000-2001.
Applied and Environmental Microbiology 70(1): 87-95.
IWA (2004). The Bonn Charter for Safe Drinking Water. London, UK, International Water
Association.
Jagals, C. and Jagals, P. (2004). Application of HACCP principles as a management tool for
monitoring and controlling microbiological hazards in water treatment facilities. Water
Science and Technology 50(1): 69-76.
Jiang, S. C. and Chu, W. (2004). PCR detection of pathogenic viruses in southern California urban
rivers. Journal of Applied Microbiology 97(1): 17-28.
Jolis, D., Pitt, P. and Hirano, R. (1999). Risk assessment for Cryptosporidium parvum in reclaimed
water. Water Research 33(13): 3051-3055.
Katzenelson, E., Buium, I. and Shuval, H. I. (1976). Risk of communicable disease infection
associated with wastewater irrigation in agricultural settlements. Science 194(4268): 944-946.
Keene, W. E., McAnulty, J. M., Hoesly, F. C., Williams, L. P., Jr., Hedberg, K., Oxman, G. L.,
Barrett, T. J., Pfaller, M. A. and Fleming, D. W. (1994). A swimming-associated outbreak
of hemorrhagic colitis caused by Escherichia coli O157:H7 and Shigella sonnei. New England
Journal of Medicine 331(9): 579-584.
Keswick, B. H., Satterwhite, T. K., Johnson, P. C., DuPont, H. L., Secor, S. L., Bitsura, J. A.,
Gary, G. W. and Hoff, J. C. (1985). Inactivation of Norwalk virus in drinking water by
chlorine. Applied and Environmental Microbiology 50(2): 261-264.
Kistemann, T., Classen, T., Koch, C., Dangendorf, F., Fischeder, R., Gebel, J., Vacata, V. and
Exner, M. (2002). Microbial load of drinking water reservoir tributaries during extreme
rainfall and runoff. Applied and Environmental Microbiology 68(5): 2188-2197.
Koenraad, P. M. F. J., Jacobs-Reitsma, W. F. and Beumer, R. R. (1996). Short-term evidence of
Campylobacter in a treatment plant and drain water of a connected poultry abattoir. Water
Environment Research 68: 188-193.
76
Microbial risk assessment in urban water systems
Koenraad, P. M. F. J., Rombouts, F. M. and Notermans, S. H. W. (1997). Epidemiological aspects
of thermophilic Campylobacter in water-related environments: a review. Water Environment
Research 69(1): 52-63.
Koff, R. S. (1998). Hepatitis A. Lancet 351(9116): 1643-1649.
Koivunen, J., Siitonen, A. and Heinonen-Tanski, H. (2003). Elimination of enteric bacteria in
biological-chemical wastewater treatment and tertiary filtration units. Water Research 37(3):
690-698.
Koopmans, M. and Duizer, E. (2004). Foodborne viruses: an emerging problem. International
Journal of Food Microbiology 90(1): 23-41.
Kramer, M. H., Herwaldt, B. L., Craun, G. F., Calderon, R. L. and Juranek, D. D. (1996).
Surveillance for waterborne-disease outbreaks – United States, 1993-1994. MMWR
Surveillance Summaries 45(1): 1-33.
Kukkula, M., Arstila, P., Klossner, M. L., Maunula, L., Bonsdorff, C. H. and Jaatinen, P. (1997).
Waterborne outbreak of viral gastroenteritis. Scandinavian Journal of Infectious Diseases 29(4):
415-418.
Kukkula, M., Maunula, L., Silvennoinen, E. and von Bonsdorff, C. H. (1999). Outbreak of viral
gastroenteritis due to drinking water contaminated by Norwalk-like viruses. Journal of
Infectious Diseases 180(6): 1771-1776.
Kärrman, E., Bergstedt, O., Westrell, T., Heinicke, G., Stenström, T.-A. and Hedberg, T. (2004).
Systemanalys av dricksvattenförsörjning med avseende på mikrobiologiska barriärer och
miljöpåverkan. (Systems analysis of drinking water systems including microbial barriers and
environmental impact). Stockholm, VA-Forsk Report 2004-12.
LeChevallier, M. W., Abbaszadegan, M., Camper, A. K., Hurst, C. J., Izaguirre, G., Marshall, M.
M., Naumovitz, D., Payment, P., Rice, E. W., Rose, J., Schaub, S., Slifko, T. R., Smith, D.
B., Smith, H. V., Sterling, C. R., Stewart, M. and Cont, A. R. D. M. (1999a). Committee
report: Emerging pathogens - bacteria. Journal of American Water Works Association 91(9):
101-109.
LeChevallier, M. W., Abbaszadegan, M., Camper, A. K., Hurst, C. J., Izaguirre, G., Marshall, M.
M., Naumovitz, D., Payment, P., Rice, E. W., Rose, J., Schaub, S., Slifko, T. R., Smith, D.
B., Smith, H. V., Sterling, C. R., Stewart, M. and Cont, A. R. D. M. (1999b). Committee
report: Emerging pathogens - viruses, protozoa, and algal toxins. Journal of American Water
Works Association 91(9): 110-121.
LeChevallier, M. W., Norton, W. D. and Lee, R. G. (1991). Occurrence of Giardia and
Cryptosporidium spp. in surface water supplies. Applied and Environmental Microbiology 57(9):
2610-2616.
Lemon, S. M. (1997). Type A viral hepatitis: epidemiology, diagnosis, and prevention. Clinical
Chemistry 43(8 Pt 2): 1494-1499.
Levine, M. M., DuPont, H. L., Formal, S. B., Hornick, R. B., Takeuchi, A., Gangarosa, E. J.,
Snyder, M. J. and Libonati, J. P. (1973). Pathogenesis of Shigella dysenteriae 1 (Shiga)
dysentery. Journal of Infectious Diseases 127(3): 261-270.
Lidskog, R., Sandstedt, E. and Sundqvist, G. (1997). Samhälle, risk och miljö : sociologiska
perspektiv på det moderna samhällets miljöproblem. (Society, risk and environment: sociological
perspectives on the environmental problems of modern society). Lund, Studentlitteratur.
Lindesmith, L., Moe, C., Marionneau, S., Ruvoen, N., Jiang, X., Lindbland, L., Stewart, P.,
LePendu, J. and Baric, R. (2003). Human susceptibility and resistance to Norwalk virus
infection. Nature Medicine 9(5): 548-553.
Lindqvist, R., Andersson, Y., Lindback, J., Wegscheider, M., Eriksson, Y., Tidestrom, L.,
Lagerqvist-Widh, A., Hedlund, K. O., Lofdahl, S., Svensson, L. and Norinder, A. (2001).
A one-year study of foodborne illnesses in the municipality of Uppsala, Sweden. Emerging
Infectious Diseases 7(3): 588-592.
77
Lodder, W. J., Vinjé, J., van de Heide, R., de Roda Husman, A. M., Leenen, E. J. T. M. and
Koopmans, M. P. G. (1999). Molecular detection of Norwalk-like caliciviruses in sewage.
Applied and Environmental Microbiology 65(12): 5624-5627.
Lopman, B., Vennema, H., Kohli, E., Pothier, P., Sanchez, A., Negredo, A., Buesa, J., Schreier, E.,
Reacher, M., Brown, D., Gray, J., Iturriza, M., Gallimore, C., Bottiger, B., Hedlund, K.,
Torven, M., von Bonsdorff, C., Maunula, L., Poljsak-Prijatelj, M., Zimsek, J., Reuter, G.,
Szucs, G., Melegh, B., Svennson, L., van Duijnhoven, Y. and Koopmans, M. (2004).
Increase in viral gastroenteritis outbreaks in Europe and epidemic spread of new
norovirus variant. Lancet 363(9410): 682-688.
Lopman, B. A., Reacher, M. H., van Duijnhoven, Y., Hanon, F. X., Brown, D. and Koopmans, M.
(2003). Viral gastroenteritis outbreaks in Europe, 1995-2000. Emerging Infectious Diseases
9(1): 90-96.
Ma, W. T., Sun, Z. C., Wang, Z. X., Feng, Y. B., Wang, T. C., Chan, U. S., Miu, C. H. and Zhu, S.
C. (1998). Application of membrane technology for drinking water. Desalination 119(1-3):
127-131.
Mac Kenzie, W. R., Hoxie, N. J., Proctor, M. E., Gradus, M. S., Blair, K. A., Peterson, D. E.,
Kazmierczak, J. J., Addiss, D. G., Fox, K. R., Rose, J. B. and Davis, J. P. (1994). A massive
outbreak in Milwaukee of Cryptosporidium infection transmitted through the public watersupply. New England Journal of Medicine 331(3): 161-167.
Macler, B. A. and Regli, S. (1993). Use of microbial risk assessment in setting US drinking water
standards. International Journal of Food Microbiology 18(4): 245-256.
Marshall, J. A., Salamone, S., Yuen, L., Catton, M. G. and Wright, J. P. (2001). High level
excretion of Norwalk-like virus following resolution of clinical illness. Pathology 33(1): 5052.
Maunula, L., Kalso, S., Von Bonsdorff, C. H. and Ponka, A. (2004). Wading pool water
contaminated with both noroviruses and astroviruses as the source of a gastroenteritis
outbreak. Epidemiology and Infection 132(4): 737-743.
Maurer, A. M. and Sturchler, D. (2000). A waterborne outbreak of small round structured virus,
Campylobacter and Shigella co-infections in La Neuveville, Switzerland, 1998. Epidemiology and
Infection 125(2): 325-332.
Mawdsley, J. L., Bardgett, R. D., Merry, R. J., Pain, B. F. and Theodorou, M. K. (1995). Pathogens
in livestock waste, their potential for movement through soil and environmental pollution.
Applied Soil Ecology 2(1): 1-15.
McAlearny, A. S., Schweikhart, S. B. and Pathak, D. S. (1999). Quality-Adjusted Life-Years and
other health indices: a comparative analysis. Clinical Therapeutics 21(9): 1605-1629.
McCarthy, N. and Giesecke, J. (2001). Incidence of Guillain-Barré syndrome following infection
with Campylobacter jejuni. American Journal of Epidemiology 153(6): 610-614.
McCarthy, T. A., Barrett, N. L., Hadler, J. L., Salsbury, B., Howard, R. T., Dingman, D. W.,
Brinkman, C. D., Bibb, W. F. and Cartter, M. L. (2001). Hemolytic-uremic syndrome and
Escherichia coli O121 at a lake in Connecticut, 1999. Pediatrics 108(4): E59.
McCullough, N. B. and Eisele, C. W. (1951a). Experimental human salmonellosis. I. Pathogenicity
of strains of Salmonella meleagridis and Salmonella anatum obtained from spray-dried whole
egg. Journal of Infectious Diseases 88(3): 278-289.
McCullough, N. and Eisele, C. W. (1951b). Experimental human salmonellosis. III. Pathogenicity
of strains of Salmonella newport, Salmonella derby, and Salmonella bareilly obtained from spraydried whole egg. Journal of Infectious Diseases 89(3): 209-213.
McCullough, N. and Eisele, C. W. (1951c). Experimental human salmonellosis. IV. Pathogenicity
of strains of Salmonella pullorum obtained from spray-dried whole egg. Journal of Infectious
Diseases 89(3): 259-265.
McGann, B. (2004). Are three water qualities better than one? Water21 April 2004: 35-35.
78
Microbial risk assessment in urban water systems
McKone, T. E. and Daniels, J. I. (1991). Estimating human exposure through multiple pathways
from air, water, and soil. Regulatory Toxicology and Pharmacology 13(1): 36-61.
Mead, P. S., Slutsker, L., Dietz, V., McCaig, L. F., Bresee, J. S., Shapiro, C., Griffin, P. M. and
Tauxe, R. V. (1999). Food-related illness and death in the United States. Emerging Infectious
Diseases 5(5): 607-625.
Medema, G. J., Bahar, M. and Schets, F. M. (1997). Survival of Cryptosporidium parvum, Escherichia
coli, faecal enterococci and Clostridium perfringens in river water: influence of temperature
and autochthonous microorganisms. Water Science and Technology 35(11-12): 249-252.
Medema, G. J., Hoogenboezem, W., Veer, A. J. v. d., Ketelaars, H. A. M., Hijnen, W. A. M. and
Nobel, P. J. (2003). Quantitative risk assessment of Cryptosporidium in surface water
treatment. Water Science and Technology 47(3): 241–247.
Medema, G. J., Teunis, P. F. M., Havelaar, A. H. and Haas, C. N. (1996). Assessment of the doseresponse relationship of Campylobacter jejuni. International Journal of Food Microbiology 30: 101111.
Melloul, A., Amahmid, O., Hassani, L. and Bouhoum, K. (2002). Health effect of human wastes
use in agriculture in El Azzouzia (the wastewater spreading area of Marrakesh city,
Morocco). International Journal of Environmental Health Research 12(1): 17-23.
Meng, Q. S. and Gerba, C. P. (1996). Comparative inactivation of enteric adenoviruses, poliovirus
and coliphages by ultraviolet irradiation. Water Research 30(11): 2665-2668.
Michel, P., Wilson, J. B., Martin, S. W., Clarke, R. C., McEwen, S. A. and Gyles, C. L. (2000).
Estimation of the under-reporting rate for the surveillance of Escherichia coli O157:H7
cases in Ontario, Canada. Epidemiology and Infection 125: 35-45.
MISTRA (1999). Revised programme plan for the MISTRA programme Sustainable Urban Water
Management.
Mounts, A. W., Ando, T., Koopmans, M., Bresee, J. S., Noel, J. and Glass, R. I. (2000). Cold
weather seasonality of gastroenteritis associated with Norwalk-like viruses. Journal of
Infectious Diseases 181 Suppl 2: S284-S287.
Murray, C. J. and Lopez, A. D. (1997). Global mortality, disability, and the contribution of risk
factors: Global Burden of Disease Study. Lancet 349: 1436-1442.
Murray, C. J. L. and Lopez, A. D. (1996). Evidence-based health policy - lessons from the Global
Burden of Disease Study. Science 274: 740-743.
Nakajima, H., Nakagomi, T., Kamisawa, T., Sakaki, N., Muramoto, K., Mikami, T., Nara, H. and
Nakagomi, O. (2001). Winter seasonality and rotavirus diarrhoea in adults. Lancet
357(9272): 1950.
Notermans, S., Gallhoff, G., Zwietering, M. H. and Mead, G. C. (1994). The HACCP concept:
specification of criteria using quantitative risk assessment. Food Microbiology 11: 397-408.
Nwachuku, N. and Gerba, C. P. (2004). Microbial risk assessment: don't forget the children.
Current Opinions in Microbiology 7(3): 206-209.
Nygård, K., Andersson, Y., Røttingen, J. A., Svensson, Å., Lindbäck, J., Kistemann, T. and
Giesecke, J. (2004a). Association between environmental risk factors and Campylobacter
infections in Sweden. Epidemiology and Infection 132(2): 317-325.
Nygård, K., Torvén, M., Ancker, C., Knauth, S. B., Hedlund, K. O., Giesecke, J., Andersson, Y.
and Svensson, L. (2003). Emerging genotype (GGIIb) of norovirus in drinking water,
Sweden. Emerging Infectious Diseases 9(12): 1548-1552.
Nygård, K., Vold, L., Halvorsen, E., Bringeland, E., Rottingen, J. A. and Aavitsland, P. (2004b).
Waterborne outbreak of gastroenteritis in a religious summer camp in Norway, 2002.
Epidemiology and Infection 132(2): 223-229.
Nylen, G., Dunstan, F., Palmer, S. R., Andersson, Y., Bager, F., Cowden, J., Feierl, G., Galloway,
Y., Kapperud, G., Megraud, F., Molbak, K., Petersen, L. R. and Ruutu, P. (2002). The
79
seasonal distribution of Campylobacter infection in nine European countries and New
Zealand. Epidemiology and Infection 128(3): 383-390.
Obiri-Danso, K. and Jones, K. (1999). Distribution and seasonality of microbial indicators and
thermophilic Campylobacters in two freshwater bathing sites on the River Lune in northwest
England. Journal of Applied Microbiology 87(6): 822-832.
Ogoshi, M., Suzuki, Y. and Asano, T. (2001). Water reuse in Japan. Water Science and Technology
43(10): 17-23.
Ong, C., Moorehead, W., Ross, A. and Isaac-Renton, J. (1996). Studies of Giardia spp. and
Cryptosporidium spp. in two adjacent watersheds. Applied Environmental Microbiology 62(8):
2798-2805.
Ottoson, J., Hansen, A., Westrell, T., Johansen, K., Norder, H. and Stenström, T. A. (submitted).
Removal of noro- and enteroviruses, Giardia cysts, Cryptosporidium oocysts and faecal
indicators at four secondary wastewater treatment plants in Sweden.
Ottoson, J. and Stenström, T. A. (2003). Faecal contamination of greywater and associated
microbial risks. Water Research 37(3): 645-655.
Pai, C. H., Kelly, J. K. and Meyers, G. L. (1986). Experimental infection of infant rabbits with
verotoxin-producing Escherichia coli. Infection and Immunity 51(1): 16-23.
Parkin, R. T., Soller, J. A. and Olivieri, A. W. (2003). Incorporating susceptible subpopulations in
microbial risk assessment: pediatric exposures to enteroviruses in river water. Journal of
Exposure Analysis and Environmental Epidemiology 13(2): 161-168.
Parrino, T. A., Schreiber, D. S., Trier, J. S., Kapikian, A. Z. and Blacklow, N. R. (1977). Clinical
immunity in acute gastroenteritis caused by Norwalk agent. New England Journal of Medicine
297(2): 86-89.
Payment, P. (1999). Poor efficiency of residual chlorine disinfectant in drinking water to inactivate
waterborne pathogens in distribution systems. Canadian Journal of Microbiology 45: 709-715.
Payment, P., Berte, A., Prévost, M., Ménard, B. and Barbeau, B. (2000). Occurrence of pathogenic
microorganisms in the Saint Lawrence River (Canada) and comparison of health risks for
populations using it as their source of drinking water. Canadian Journal of Microbiology 46:
565-575.
Payment, P., Siemiatycki, J., Richardson, L., Renaud, G., Franco, E. and Prévost, M. (1997). A
prospective epidemiological study of gastrointestinal health effects due to the
consumption of drinking water. International Journal of Environmental Health Research 7(1): 531.
Petterson, S. R., Ashbolt, N. J. and Sharma, A. (2001). Microbial risks from wastewater irrigation
of salad crops: a screening-level risk assessment. Water Environment Research 73(6): 667-672.
Pina, S., Puig, M., Lucena, F., Jofre, J. and Girones, R. (1998). Viral pollution in the environment
and in shellfish: human adenovirus detection by PCR as an index of human viruses.
Applied and Environmental Microbiology 64(9): 3376-3382.
Pouillot, R., Beaudeau, P., Denis, J. B. and Derouin, F. (2004). A quantitative risk assessment of
waterborne cryptosporidiosis in France using second-order Monte Carlo simulation. Risk
Analysis 24(1): 1-17.
Rao, V. C., Metcalf, T. G. and Melnick, J. L. (1987). Removal of indigenous rotavirus during
primary settling and activated-sludge treatment of raw sewage. Water Research 21(2): 171177.
Rao, V. C., Seidel, K. M., Goyal, S. M., Metcalf, T. G. and Melnick, J. L. (1984). Isolation of
enteroviruses from water, suspended solids and sediments from Galveston bay: survival of
poliovirus and rotavirus adsorbed to sediments. Applied and Environmental Microbiology 48(2):
404-409.
Raphael, R. A., Sattar, S. A. and Springthorpe, V. S. (1985). Long-term survival of human
rotavirus in raw and treated river water. Canadian Journal of Microbiology 31(2): 124-128.
80
Microbial risk assessment in urban water systems
Regli, S., Rose, J. B., Haas, C. N. and Gerba, C. P. (1991). Modeling the risk from Giardia and
viruses in drinking water. Journal of American Water Works Association 83(11): 76-84.
Rendtorff, R. C. (1954). The experimental transmission of human intestinal protozoan parasites.
II. Giardia lamblia cysts given in capsules. American Journal of Hygiene 59(2): 209-220.
Robertson, L. J. and Gjerde, B. (2001). Occurrence of Cryptosporidium oocysts and Giardia cysts in
raw waters in Norway. Scandinavian Journal of Public Health 29(3): 200-207.
Robertson, L. J., Johannessen, G. S., Gjerde, B. K. and Loncarevic, S. (2002). Microbiological
analysis of seed sprouts in Norway. International Journal of Food Microbiology 75(1-2): 119-126.
Rockx, B., de Wit, M., Vennema, H., Vinje, J., de Bruin, E., van Duynhoven, Y. and Koopmans,
M. (2002). Natural history of human calicivirus infection: a prospective cohort study.
Clinical Infectious Diseases 35(3): 246-253.
Rose, J. B., Dickson, L. J., Farrah, S. R. and Carnahan, R. P. (1996). Removal of pathogenic and
indicator microorganisms by a full-scale water reclamation facility. Water Research 30(11):
2785-2797.
Rose, J. B. and Gerba, C. P. (1991a). Assessing potential health risks from viruses and parasites in
reclaimed water in Arizona and Florida, USA. Water Science and Technology 23(Kyoto): 20912098.
Rose, J. B. and Gerba, C. P. (1991b). Use of risk assessment for development of microbial
standards. Water Science and Technology 24(2): 29-34.
Roseberry, A. M. and Burmaster, D. E. (1992). Lognormal distributions for water intake by
children and adults. Risk Analysis 12(1): 99-104.
Rosef, O., Rettedal, G. and Lageide, L. (2001). Thermophilic Campylobacter in surface water: a
potential risk of campylobacteriosis. International Journal of Environmental Health Research
11(4): 321-327.
Sahlström, L., Aspan, A., Bagge, E., Tham, M. L. and Albihn, A. (2004). Bacterial pathogen
incidences in sludge from Swedish sewage treatment plants. Water Research 38(8): 19891994.
Samadpour, M., Stewart, J., Steingart, K., Addy, C., Louderback, J., McGinn, M., Ellington, J. and
Newman, T. (2002). Laboratory investigation of an E. coli O157:H7 outbreak associated
with swimming in Battle Ground Lake, Vancouver, Washington. Journal of Environmental
Health 64(10): 16-20, 26, 25.
Schaub, S. A. and Oshiro, R. K. (2000). Public health concerns about caliciviruses as waterborne
contaminants. Journal of Infectious Diseases 181 Suppl 2: S374-S380.
Schvoerer, E., Bonnet, F., Dubois, V., Cazaux, G., Serceau, R., Fleury, H. J. and Lafon, M. E.
(2000). PCR detection of human enteric viruses in bathing areas, waste waters and human
stools in Southwestern France. Research in Microbiology 151(8): 693-701.
Schönning, C. (2003). Risker för smittspridning via avloppsslam - redovisning av
behandlingsmetoder och föreskrifter. (Risks of disease transmission via sewage sludge - treatment
methods and regulations). Stockholm, Naturvårdsverket.
Shin, G. A. and Sobsey, M. D. (2003). Reduction of Norwalk virus, poliovirus 1, and
bacteriophage MS2 by ozone disinfection of water. Applied and Environmental Microbiology
69(7): 3975-3978.
Shuval, H., Lampert, Y. and Fattal, B. (1997). Development of a risk assessment approach for
evaluating wastewater reuse standards for agriculture. Water Science and Technology 35(11-12):
15-20.
Shuval, H. I. (1991). Effects of wastewater irrigation of pastures on the health of farm animals and
humans. Revue Scientifique et Technique de L’Office International de Epizooties 10(3): 847-866.
Sinclair, M. (2003a). Setback for Netherlands dual supplies. Health Stream (30): 5.
Sinclair, M. (2003b). US waterborne disease study. Health Stream (31): 1-3.
Slovic, P. (1987). Perception of risk. Science 236: 280-285.
81
SLV (2001). Livsmedelsverkets föreskrifter om dricksvatten. (Regulations and general advices on
drinking water from the National Food Administration). SLVFS 2001:30.
SLV (2002). Campylobacter i kött och vatten - kartläggning av Campylobacter i rått kött och råvatten
till dricksvatten i Sverige år 2000. (Campylobacter in meat and water – mapping of Campylobacter
in raw meat and raw water to drinking water in Sweden 2000). Uppsala, Swedish National Food
Administration.
SLV (2003). Vägledning till Livsmedelsverkets föreskrifter om dricksvatten. (Guidance to the national
regulations on drinking water from the National Food Administration ).
SMI (2004). Epidemiologisk årsrapport 2003. (Epidemiological yearly report 2003). Solna, Swedish
Institute for Infectious Disease Control (SMI).
SNV (1996). Statens naturvårdsverks föreskrifter om strandbadvatten. (Regulations on recreational
waters from the Swedish Environmental Protection Agency). SNFS (1996:6).
Soller, J. A., Olivieri, A. W., Crook, J., Cooper, R. C., Tchobanoglous, G., Parkin, R. T., Spear, R.
C. and Eisenberg, J. N. S. (2003). Risk-based approach to evaluate the public health
benefit of additional wastewater treatment. Environmental Science and Technology 37(9): 18821891.
Srikanth, R. and Naik, D. (2004). Prevalence of giardiasis due to wastewater reuse for agriculture
in the suburbs of Asmara City, Eritrea. International Journal of Environmental Health Research
14(1): 43-52.
Stenström, T. A. (1987). Kommunalt avloppsvatten från hygienisk synpunkt - mikrobiologiska
undersökningar. (Municipal wastewater from a hygienic point-of-view – microbiological investigations).
Stockholm, Swedish Environmental Protection Agency.
Stenström, T. A., Boisen, F., Georgsson, F., Lahti, K., Lund, V., Andersson, Y. and Ormerod, K.
(1994). Vattenburna infektioner i Norden. (Waterborne infections in the Nordic countries).
Copenhagen, Nordic Councel of Ministers.
Stenström, T. A. and Carlander, A. (1999). Mikrobiella risker för smittspridning och sjukdomsfall slamspridning och behandling. (Microbial risks with disease transmission and disease cases - sludge
application and treatment). Stockholm, Swedish Environmental Protection Agency.
Stenström, T. A. and Carlander, A. (2001). Occurrence and die-off of indicator organisms in the
sediment in two constructed wetlands. Water Science and Technology 44(11-12): 223-230.
Stenström T. A., Wiman M. and Tullander V. (1985). Hässleholms avloppsreningsverk,
intensivundersökning 19-21 april 1982. (Hässleholm’s wastewater treatment plant, intensive
investigation April 19-21, 1982). Report SNV PM 1944, Sewage Works Evaluation Project,
Sweden.
Strachan, N. J., Fenlon, D. R. and Ogden, I. D. (2001). Modelling the vector pathway and
infection of humans in an environmental outbreak of Escherichia coli O157. FEMS
Microbiology Letters 203(1): 69-73.
Svenskt Vatten (2000). Facts on water supply and sanitation in Sweden. Accessible at
http://www.svensktvatten.se/main/main.asp?objectID=376, The Swedish Water and
Wastewater Association.
Svenungsson, B., Lagergren, Å., Ekwall, E., Evengård, B., Hedlund, K. O., Kärnell, A., Löfdahl,
S., Svensson, L. and Weintraub, A. (2000). Enteropathogens in adult patients with diarrhea
and healthy control subjects: a 1-year prospective study in a Swedish clinic for infectious
diseases. Clinical Infectious Diseases 30: 770-778.
Söderberg, H. and Kain, J.-H. (2002). Integrating knowledge or aggregating data - multicriteria
approaches for sustainable water and waste systems. Proceedings from the 3rd
International Conference on Decision-Making in Urban and Civil Engineering, 6-8
November 2002, London.
82
Microbial risk assessment in urban water systems
Tanaka, H., Asano, T., Shroeder, E. D. and Tchobanoglous, G. (1998). Estimating the safety of
wastewater reclamation and reuse using enteric virus monitoring data. Water Environment
Research 70(1): 39-51.
Tani, N., Dohi, Y., Kurumatani, N. and Yonemasu, K. (1995). Seasonal distribution of
adenoviruses, enteroviruses and reoviruses in urban river water. Microbiology and Immunology
39(8): 577-580.
Tchobanoglous, G. and Burton, F. L. (1991). Wastewater engineering: treatment, disposal, and
reuse /Metcalf & Eddy, Inc. New York, McGraw-Hill.
Tessier, J. L. and Davies, G. A. L. (1999). Giardiasis. Primary Care Update for OB/GYNS 6(1): 8-11.
Teunis, P., Chappell, C. and Okhuysen, P. (2002a). Cryptosporidium dose response studies: variation
between isolates. Risk Analysis 22(1): 175-183.
Teunis, P., Takumi, K. and Shinagawa, K. (2004). Dose response for infection by Escherichia coli
O157:H7 from outbreak data. Risk Analysis 24(2): 401-407.
Teunis, P. F. M., Chappell, C. L. and Okhuysen., P. C. (2002b). Cryptosporidium dose-response
studies: variation between hosts. Risk Analysis 22(3): 475-486.
Teunis, P. F. M. and Havelaar, A. H. (2000). The Beta-Poisson dose-response model is not a
single-hit model. Risk Analysis 20(4): 513-520.
Teunis, P. F. M., van der Heijden, O. G., van der Giessen, J. W. B. and Havelaar, A. H. (1996).
The dose-response relation in human volunteers for gastro-intestinal pathogens.
Bilthoven, The Netherlands, National Institute of Public Health and the Environment
(RIVM).
Teunis, P. F. M., Medema, G. J., Kruidenier, L. and Havelaar, A. H. (1997). Assessment of the risk
of infection by Cryptosporidium or Giardia in drinking water from a surface water source.
Water Research 31(6): 1333-1346.
Thorn, J. and Beijer, L. (2004). Work-related symptoms and inflammation among sewage plant
operatives. International Journal of Occupational and Environmental Health 10(1): 84-89.
Thorn, J., Beijer, L. and Rylander, R. (2002). Work related symptoms among sewage workers: a
nationwide survey in Sweden. Occupational and Environmental Medicine 59(8): 562-566.
Thurston-Enriquez, J. A., Haas, C. N., Jacangelo, J. and Gerba, C. P. (2003). Chlorine inactivation
of adenovirus type 40 and feline calicivirus. Applied and Environmental Microbiology 69(7):
3979-3985.
Uhnoo, I., Wadell, G., Svensson, L., Olding-Stenkvist, E., Ekwall, E. and Möllby, R. (1986).
Aetiology and epidemiology of acute gastro-enteritis in Swedish children. Journal of Infection
13: 73-89.
Urban Water (2002). Sustainable Urban Water Management: Programme Plan 2003.
USEPA (1997). Exposure factor handbook. Washington, DC, Office of Research and
Development, National Center for Environmental Assessment, U.S. Environmental
Protection Agency.
USEPA (2000). Estimated per capita water ingestion in the United States. Based on the United
States Department of Agriculture's 1994-96 Continuing Survey of Food Intakes by
Individuals, United States Environmental Protection Agency, Office of Water.
Van, R., Wun, C. C., O'Ryan, M. L., Matson, D. O., Jackson, L. and Pickering, L. K. (1992).
Outbreaks of human enteric adenovirus types 40 and 41 in Houston day care centers.
Journal of Pediatrics 120(4 Pt 1): 516-521.
van Breemen, L. W. C. A., Ketelaars, H. A. M., Hoogenboezem, W. and Medema, G. (1998).
Storage reservoirs - a first barrier for pathogenic micro-organisms in the Netherlands.
Water Science and Technology 37(2): 253-260.
van Moeffaert, D. (2003). Multi criteria decision aid in sustainable urban water management.
Proceedings from the 2nd International Symposium on Ecological Sanitation, 7-11 April,
Lübeck, Germany.
83
Vernozy-Rozand, C., Montet, M. P., Lequerrec, F., Serillon, E., Tilly, B., Bavai, C., Ray-Gueniot,
S., Bouvet, J., Mazuy-Cruchaudet, C. and Richard, Y. (2002). Prevalence of verotoxinproducing Escherichia coli (VTEC) in slurry, farmyard manure and sewage sludge in France.
Journal of Applied Microbiology 93(3): 473-478.
Vinnerås, B., Björklund, A. and Jönsson, H. (2003a). Thermal composting of faecal matter as
treatment and possible disinfection method - laboratory-scale and pilot-scale studies.
Bioresource Technology 88(1): 47-54.
Vinnerås, B., Holmqvist, A., Bagge, E., Albihn, A. and Jönsson, H. (2003b). The potential for
disinfection of separated faecal matter by urea and by peracetic acid for hygienic nutrient
recycling. Bioresource Technology 89(2): 155-161.
Wadell, G. (1984). Molecular epidemiology of human adenoviruses. Current Topics in Microbiology
and Immunology 110: 191-220.
Wahlström, H., Tysen, E., Olsson Engvall, E., Brandström, B., Eriksson, E., Mörner, T. and
Vågsholm, I. (2003). Survey of Campylobacter species, VTEC O157 and Salmonella species in
Swedish wildlife. Veterinary Record 153(3): 74-80.
Waldenström, J., Broman, T., Carlsson, I., Hasselquist, D., Achterberg, R. P., Wagenaar, J. A. and
Olsen, B. (2002). Prevalence of Campylobacter jejuni, Campylobacter lari, and Campylobacter coli
in different ecological guilds and taxa of migrating birds. Applied and Environmental
Microbiology 68(12): 5911-5917.
Wang, G. and Doyle, M. P. (1998). Survival of enterohemorrhagic Escherichia coli O157:H7 in
water. Journal of Food Protection 61(6): 662-667.
Ward, R. L., Bernstein, D. I., Young, E. C., Sherwood, J. R., Knowlton, D. R. and Schiff, G. M.
(1986). Human rotavirus studies in volunteers: determination of infectious dose and
serological response to infection. Journal of Infectious Diseases 154(5): 871-880.
Water UK (2004). The application of HACCP procedures in the water industry: biosolids
treatment and use on agricultural land. London, Water UK, prepared by WRc.
WCED (1987). Our common future, World Commission on Environment and Development,
Oxford University Press, Oxford, UK.
Wheeler, J. G., Sethi, D., Cowden, J. M., Wall, P. G., Rodrigues, L. C., Thompkins, D. S., Hudson,
M. J. and Roderick, P. J. (1999). Study of infectious intestinal disease in England: rates in
the community, presenting to general practice, and reporting to national surveillance. BMJ
British Medical Journal 318: 1046-1050.
WHO (2003). Guidelines for Safe Recreational Waters: Volume 1 - Coastal and Fresh-Waters.
Geneva, World Health Organization.
WHO (2004). Guidelines for Drinking Water Quality, 3rd edition. Geneva, World Health
Organization.
WHO and UNICEF (2004). Meeting the MDG drinking-water and sanitation target: a mid-term
assessment of progress. Geneva, World Health Organization and United Nations
Children's Fund.
Xiao, L. H., Fayer, R., Ryan, U. and Upton, S. J. (2004). Cryptosporidium taxonomy: Recent
advances and implications for public health. Clinical Microbiology Reviews 17(1): 72-97.
Yoder, J. S., Blackburn, B. G., Craun, G. F., Hill, V., Levy, D. A., Chen, N., Lee, S. H., Calderon,
R. L. and Beach, M. J. (2004). Surveillance for waterborne-disease outbreaks associated
with recreational water – United States, 2001-2002. MMWR Surveillance Summaries 53(8): 122.
Zwietering, M. H. and van Gerwen, S. J. C. (2000). Sensitivity analysis in quantitative microbial
risk assessment. International Journal of Food Microbiology 58(3): 213-221.
84
Was this manual useful for you? yes no
Thank you for your participation!

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

Download PDF

advertisement