Report No DWI0724

Aug 1995

EXECUTIVE SUMMARY

In the United States, models have been developed to predict the risks of microbiological infection from drinking water supplies. These are used by the U.S. Environmental Protection Agency not only for the development of microbial standards but also for determining the level of drinking water treatment required to ensure an acceptable risk of infection. One model has predicted that there may be a lifetime risk of death as high as 1 in 20 from exposure to waterborne virus.

This report was commissioned by the Department of the Environment with the objectives of both reviewing such models world-wide and considering the development of models for application in the UK to specific pathogens.

The first part of this report reviews and critically assesses the risk assessment models developed world-wide for pathogens in drinking water. No information was found for countries other than the US. The major criticism of US models is that no account is taken of what proportion of consumers are exposed to what numbers of pathogens. Indeed, one manifestation of the assumptions made is that consumers are effectively either exposed to just one pathogen or to zero pathogens. This may not be appropriate to drinking water supplies where micro-organisms appear to be clustered. Clustering would render a smaller proportion of consumers exposed to much higher numbers of pathogens. It is concluded that by ignoring pathogen clustering, US models could overestimate the risk from more infectious pathogens (e.g., viruses and protozoa) but underestimate the risk from less infectious bacterial pathogens.

The remainder of this review considers the development of risk assessment models in the UK. It is suggested that any pathogen which presents a hazard through drinking water should be accommodated. Epidemiological information on individual pathogens highlights the necessity to customise UK risk assessment models for each pathogen. The most cost-effective approach for risk modelling in the UK is to further our understanding of pathogen numbers in the drinking water supply. In view of pathogen clustering and the highly infectious nature of certain pathogens, it is proposed that dose-response data may be redundant in drinking water models. In effect, infection could be modelled on the proportion of consumers exposed to a dose of one or more pathogens. Such an approach would provide a 'worst case scenario'.

Little information is available on pathogen numbers in the UK drinking water supply, because very large volumes need to be sampled to detect them. Here, a method is demonstrated for modelling pathogen numbers across the supply using raw water data and treatment removal rates. That would enable the cost-effectiveness of various risk reduction options to be modelled. In addition to effects of environmental inputs and treatment processes on health could be assessed. The overall conclusion of this review is that data and technology may be available to develop a UK risk assessment model for Cryptosporidium.

Copies of this report may be available as an Acrobat pdf download under the 'Find Completed Research' heading on the DWI website.