The Scottish Government

Main Findings

• A review of the literature revealed that because of the predominance of food associated infections and outbreaks, past and current research has focused on agricultural sources of E.coli O157 and how it passed into the food chain. There was a marked absence of research specifically directed at studying E.coli O157 in water
• There is no evidence to indicate that E.coli O157 is more persistent in the environment or more resistant to water treatment processes than the non pathogenic E.coli found in the gastro-intestinal tract.
• Except in the case of humans with a severe infection, the number of E.coli O157 in faecal material will be several orders of magnitude lower than the concentration of other E.coli species. It is tentatively concluded therefore that existing microbial standards for drinking water quality should be adequate. However, because the infective dose of E.coli O157 appears to be low, drinking water supplies which only just achieve compliance with coliform and E.coli standards should be regarded as being at risk.
• Private water supplies may require additional barriers to prevent contamination by E.coli O157 and other enteric pathogens. This should be achieved by better source protection, the use of point-of-use devices or a combination of both.
• Future research must concentrate on obtaining better information on the fate of E.coli O157 in the environment, ensuring that it is removed by normal water treatment processes and identifying the most acceptable and cost-effective means of protecting private water supplies.

Introduction

Escherichia coli O157 was first recognised as a pathogen in 1982 in the United States of America and since then the number of reported outbreaks and sporadic cases has increased. Undoubtedly, some of this has been attributable to increased clinical awareness and improved techniques for the isolation and identification of E.coli O157. However, even when allowances are made for this, there is still an increased incidence presumably due to proliferation in, and increased exposure of, the human population to E.coli O157.

The primary reservoirs of the organisms appear to be humans, healthy cattle (dairy and beef) and other farm animals such as sheep. Wastes which contain faecal material from these animals are therefore the contaminants of most relevance to water sources. These include sewage effluents and sewage sludge, particularly from treatment works receiving abattoir waste, surface water run-off from agricultural land to which farm slurries and agricultural 'dirty water' have been applied, and surface run-off from grazing pastures.

The incidence of infection with E.coli O157 is variable throughout the United Kingdom with the highest rate being reported in Scotland. Outbreaks associated with the consumption of contaminated food have raised further concerns about this geographical bias. While the consumption of contaminated food has been identified as a primary cause of infections with E.coli O157, water has also been implicated. The Scottish Office undertook to consider the need for further research aimed at clarifying the extent of the risks posed by E.coli O157 to water supplies and commissioned a scoping study.

Methodology

The review of past and current research revealed that there was little knowledge about the source and fate of E.coli O157 in water. Although somewhat theoretical, it was felt that risk assessment was the appropriate approach to take. E.coli O157 is one of a number of micro-organisms which are being termed 'emerging pathogens'. Although they are of increasing concern to the water industry, a lack of adequate isolation and analysis techniques means that currently there are insufficient data on which to base operational decisions. In spite of this lack of data, risk assessment has proved valuable in studying the threat posed by Cryptosporidium and the BSE agent.

Risk assessment can also be likened to Hazard Assessment Critical Control Point Analysis, a technique finding increasing application in the food industry and recommended in the Pennington Report for greater use in the control of contamination by E.coli O157. Additionally, a recent guidance note issued by The Scottish Office Agriculture, Environment and Fisheries Department on the protection of private water supplies, recommended that risk factors should be identified for each supply and priorities for action assessed on this basis.

Reported Waterborne Outbreaks of E.coli O157

To facilitate reviewing current and past research on E.coli O157, attention focused on the epidemiology of reported waterborne outbreaks of E.coli O157.

Three waterborne outbreaks of E.coli O157 are documented. Two occurred in Scottish villages and a third in Missouri, USA. In all three outbreaks the cause was the contamination of the water supply after treatment and during distribution. This tends to imply that the treatment barriers used were effective in removing threats posed by enteric disease but that another barrier (the integral distribution system) aimed at protecting the quality of the treated water failed.

Bacteriological examination of the water revealed the presence of relatively high concentrations of coliform organisms indicating that routine water quality parameters could have forewarned of the potential for an outbreak of disease.

Risk Assessment

Sources of E.coli O157 of Importance to Water Supplies

Currently studies have established that man and farm animals such as cattle, sheep and goats are significant reservoirs of E.coli O157 and as such these sources can be regarded as threats to the security of water supplies, particularly private supplies in which adequate treatment is not guaranteed.

The Loading from Animal Reservoirs of E.coli O157

Existing data on prevalence, loading and patterns of excretion are difficult to interpret in terms of assessing the risks to water supplies. However, it seems reasonable to assume that agricultural sources of E.coli O157 could represent a threat in situations where there are few or no treatment barriers. It is worth noting that, based on the available information, E.coli O157 will only be present in relatively small concentrations compared to other E.coli present in the faeces of animals.

Cattle consistently excrete 108 E.coli per gram of faecal material and other faecal indicator bacteria, such as enterococci, are excreted at higher concentrations. It can be concluded therefore that unless the survival of E.coli O157 is inordinately (100 to 1000 times) greater than that of these faecal indicator bacteria, the potential presence of this organism should be easily detected in water supplies which are regularly monitored.

Environmental Survival

Insufficient data is available on the survival of E.coli O157 in water on which to base a judgement as to whether it should be regarded as being of special concern. The survival of enteric bacteria in the environment is influenced by many factors including temperature, nutrient availability, predation and sunlight. There are however two important questions that need to be addressed.

Firstly, does E.coli O157 survive for longer periods of time than other E.coli or coliform organisms? If it does then the situation could be envisaged where routine monitoring of microbiological quality using faecal indicators would not indicate the potential presence of E.coli O157. Although it seems improbable, the available data cannot give a definite answer to this question. Further work on the existence of the viable, but non-culturable state of E.coli would therefore be useful to ascertain if current standards of drinking water quality provide adequate protection.

Secondly, does E.coli O157 survive long enough on land to make contamination of surface or groundwater sources, via surface run off, a possibility? The extended survival in the faecal material from cattle (56-70 days) would appear to make this a possibility in that significant rainfall would be anticipated in such a period and transit times in shallow aquifers may be less than 50 days.

Impact of Water Treatment Processes on E.coli O157

The incidence of E.coli O157 in treated drinking water is only likely to be a problem if there has been a failure of the chlorination process at the treatment works or contamination of the treated water in supply. Contamination in supply by cattle slurry or sewage is possible through cracked pipes etc.

Both conditions would be detected by high total and faecal coliform counts on routine water sampling and indicate that there had been a fundamental breakdown in the multiple barriers used in the treatment and supply of the drinking water.

The Infective Dose of E.coli O157 in Drinking Water

For the purposes of risk assessment it may be appropriate to consider that doses of single cells of E.coli O157 are capable of causing infection through water.

Outcome of Preliminary Risk Assessment

Public Supplies

Water treatment processes are effective in removing enteric bacteria through the physical processes of coagulation and filtration, and disinfection. No evidence could be found to suggest that E.coli O157 is particularly resilient to these forms of treatment and, as a consequence, it is concluded that water treatment based on multiple barriers should form an effective barrier against E.coli O157.

Private Supplies

The risk assessment suggests that E.coli O157, along with other enteric organisms, can pose a particular threat to private water supplies because treatment barriers will be minimal and presumably maintenance and monitoring are of a fairly low standard.

Future Research Needs

The output from a preliminary risk assessment on E.coli O157 has identified gaps in current knowledge about this organism and how it could be a threat to water supplies. Areas for further research to allow the threats to water supplies to be understood better and therefore minimised are:-

• Agricultural Routes of Transmission; since cattle and other farm animals are recognised as the primary reservoirs of E.coli O157, it is likely that the main sources of potential contamination will originate from the disposal of slurries and "dirty water" from farms and wastes from abattoirs. The effectiveness of treatment barriers against the spread of E.coli needs to be determined. This should allow the security of the agricultural outlet for this product to be assessed with regard to a) its safe use on pasture land and b) its potential as a source of contamination of surface water courses through surface run-off.
• Improved Source Protection; the installation of point of use or point of entry treatment devices is often considered as a means of improving the quality of private supplies. Whilst these devices can bring about significant improvements, the cost of purchase and the need for regular maintenance often discourage their use. An alternative approach would be to try to ensure better protection of the quality of the water both at its source and its point of abstraction.
• Survival in Water and Removal by Water Treatment; given the current shortcomings in the methodologies for analysing for E.coli O157 it is difficult to envisage how routine monitoring for the organism would be feasible and the use of a surrogate would have greater advantages in terms of cost, but more importantly in the frequency with which monitoring could be carried out. The aim would be to ascertain if E.coli O157 exhibits similar decay rates in natural waters as faecal indicator organisms. This would demonstrate whether or not monitoring microbiological quality using standard parameters such as coliforms, E.coli, faecal streptococci and clostridia could provide adequate indication of the potential presence of E.coli O157.