Background and public health relevance
Research evidence
Considerations for using the decision-making tool to determine a response to a positive E. coli result in drinking water
Putting evidence into practice
References

You are a drinking water officer who has been notified of a positive result for Escherichia coli (E.coli) in a drinking water system that serves less than 500 people. Only one sample was taken during the week of notification. You are wondering whether you need to call for boil water advisory (BWA). What actions do you take to determine whether public health may be at risk? More specifically, you would like to clarify:

• What are the consequences of over-reacting as opposed to failing to act on the positive result?
• What quick actions can I take to facilitate decision making regarding the E. coli result?

The purpose of this document is to aid small water system operators and health authority officials dealing with uncertainties associated with a single positive E. coli result.

Background and public health relevance

Finding a true positive sample in drinking water for the microbial indicator E. coli is a definite indication of recent faecal contamination in drinking water systems, and the possible presence of pathogens. However, in some instances a single positive test for E. coli in drinking water does not necessarily result in a public health risk; as it may be a false positive due to sampling or laboratory error. Uncertainty concerning the public health significance of presumptive results may lead to two types of inadequate responses: over-reactions when there is not public health threat (false positive error⁶) or failure to act when there is a real hazard (false negative error⁶). These circumstances leave decision makers with the challenge of determining whether or not to act. By the time microbiological test results are released to the operator (24-48h), the water has likely been consumed, which adds more pressure to the decision maker. A quick and effective response to unacceptable bacteriological quality is a critical part in the provision of safe drinking water. This can be done by issuing a BWA,¹ which is the most common response to adverse water quality results. Boil water advisories are public announcements issued by the responsible authority advising the public that they should boil their drinking water for 1 minute (2 minutes elevations over 2000 m) prior to consumption.¹ Health Canada provides some guidance on the decision-making process for issuing BWA, including steps and criteria to consider;² however, there is no standardised approach for deciding when to issue BWAs across Canada and between provinces.³

Decision makers often face the dilemma of taking immediate action (based on presumptive microbiological results) by issuing a precautionary BWA or waiting another 24–48 hours to confirm initial laboratory results. Knowing when to call for a BWA based on one positive E. coli result is challenging, especially for Small Drinking Water Systems (SDWS). As defined by Health Canada, SWDS serve 5,000 people or less.⁴ The exact number in Canada is not known, but Statistics Canada estimated that in 2007, 28 million Canadians received their water from systems serving less than 300 people.⁵ Small drinking water systems are subject to unique challenges including high staff turn-over, missing historical water system records³, and a the lack of financial resources. These unique challenges render the interpretation of adverse water quality results difficult.

For additional background information on microbial indicators and their public health significance, please refer to Health Canada guidelines² and the NCCEH primer on the use and limitations of microbial indicators in drinking water.⁷

Research evidence

The direct consequences of failing to act on drinking water hazards were illustrated by the Walkerton, Ontario outbreak in 2000, during which, 7 people died and 2300 became ill because of poor management decisions.⁶ Specifically, responsible authorities failed to take appropriate action when positive E. coli tests occurred due to a chlorination failure.6 In contrast, authorities over-reacted during the 1998 water crisis in Sydney, Australia and issued a BWA based on unreliable laboratory results. They spent tens of millions of public dollars on the investigation.⁶ Experts finally concluded that the public’s health was not at risk.⁶

In addition to unnecessary cost, unwarranted BWAs may create panic among public and lead to a loss of public confidence in water systems.³ Repetitive BWAs impact their effectiveness as people are less likely to follow advice when the frequency of BWAs increases.^3,6 Other unintended consequences include the possible risk of burn injuries from boiling water at home³ and economic losses to businesses associated with water supply adaptations and lost revenue. A case study in Boston showed that businesses serving or manufacturing food and beverages incurred the highest economic cost, ranging from $100 to $400,000.⁸

From a public Health perspective, the consequences of over-reacting (issuing a BWA that is not warranted) are less harmful than failing to act⁶ (not issuing a BWA when it is needed). However, the fear of failing to act may lead water operators to call for BWAs as precautionary action when, in fact, no public health threat exists.

To make the best evidence based decision, it is important to investigate beyond the microbiological results. Weather events, operational parameters at the treatment plant, recent water quality complaint reports, or deviations from the sampling protocol are all important parameters to consider (see Figure 1). As suggested by Gammie (2001),⁹ these actions should be part of a pre-established protocol to better assess the available evidence at the time of decision-making and optimize response to a potential crisis.

Figure 1 provides a decision-making tool to guide the response to a positive E. coli result in drinking water. Please note that only multiple factors will point towards a faecal contamination event or a sampling/laboratory error. The tool was produced in collaboration with field experts (see acknowledgements).

Considerations for using the decision-making tool to determine a response to a positive E. coli result in drinking water

Benefits/advantages of using the decision making tool

• Because the turn around time for the analysis of water samples for microbiological tests is 24-48 hrs, the water has likely been consumed when the test result is released. Waiting another 24-48 hr to confirm the result may put public health at risk. Actions described in the toolkit to rule out a false positive E. coli provide immediate results and would not significantly delay the call for a BWA; therefore, using the toolkit
  • May improve the effectiveness of future advisories as it could prevent BWAs that are not warranted

Risks and potential unintended consequences of using the decision making tool

• Actions described in the toolkit to rule out a false positive E. coli provide immediate results and would not significantly delay the call for a BWA; therefore, if appropriately used, the risks of using this tool are minimal.

Cost

• Minimal, as tests are cheap to perform and can be done on site.

Technical feasibility of using the decision making tool

• It is necessary to possess a good understanding of the water system.
• Familiarity with the tool will ensure a quick and effective response.
• Basic knowledge of microbial indicators and physical parameters and their limitations is also required.

Equity and ethical considerations of using the decision making tool

• This tool may bring more consistent approaches to dealing with a single positive E. coli result, regardless of the system size and location.
• The actions described in the tool kit require minimal expenses allowing communities with limited resources to use the tool kit.

Deciding what actions to take based on one positive E. coli sample can be challenging. Although BWAs need to be called when there are real risks to public health, frequent BWAs can lead to decreased compliance, decreased public trust of the safety of their drinking water, and impacts to the local economy (e.g., tourism, restaurants). Operational and physical parameters can be useful in this type of situation because they provide real time data and can help ascertain whether the result indicates a real risk to public health. Good communication between the operator of the small water system, the laboratory that analysed the sample and local physicians is essential in gathering and interpreting available evidence. It can also be helpful to consult with long term health care facilities for increased incidence of disease that could be waterborne as their population is generally more vulnerable to diseases. A pre-established response protocol for potential crisis situations is recommended to expedite and provide guidance for the response.

Putting evidence into practice

Figure 1

References

1. Health Canada. Guidance for issuing and rescinding boil water advisories. Ottawa, ON: Ministry of Health, Federal-Provincial-Territorial Committee on Drinking Water, Federal-Provincial-Territorial Committee on Health and the Environment; 2009. Available from: http://www.hc-sc.gc.ca/ewh-semt/pubs/water-eau/boil_water-eau_ebullition/index-eng.php.
2. Health Canada. Guidance for issuing and rescinding drinking water avoidance advisories in emergency situations. Ottawa, ON: Ministry of Health, Federal-Provincial-Territorial Committee on Drinking Water, Federal-Provincial-Territorial Committee on Health and the Environment; 2009 Feb. Available from: http://www.hc-sc.gc.ca/ewh-semt/pubs/water-eau/avoid-annul/index-eng.php.
3. Grover R. Boil, boil, toil and trouble. The trouble with boil water advisories in British Columbia. Vancouver, BC: University of British Columbia; 2011. Available from: https://circle.ubc.ca/bitstream/id/118324/ubc_2011_spring_grover_renuka.pdf.
4. Health Canada. Guidance for providing safe drinking water in areas of federal jurisdiction - Version 1. Ottawa, ON: Environmental and Workplace Health; 2005 Aug 23. Available from: http://www.hc-sc.gc.ca/ewh-semt/pubs/water-eau/guidance-federal-conseils/index-eng.php
5. Statistics Canada. Survey of drinking water plants: 2005 to 2007. Ottawa, ON: Minister of Industry, Government of Canada; 2009 Dec. Available from: http://www.statcan.gc.ca/pub/16-403-x/16-403-x2009001-eng.pdf.
6. Hrudey SE, Hrudey EJ. Safe drinking water: lessons from recent outbreaks in affluent nations. London, UK: IWA Publishing; 2004.
7. Verhille S. Understanding microbial indicators for drinking water asessment: interpretation of test results and public health significance. Vancouver: National Collaborating Centre for Environmental Health 2013. Available from: http://www.ncceh.ca/en/major_projects/drinking_water.
8. Heflin C, Jensen J, Miller K. Community resilience: Understanding the economic impacts of disruptions in water service. Policy brief. Columbia, MO: University of Missouri, Institute of Public Policy; 2013 Jan. Available from: http://ipp.missouri.edu/files/ipp/attachments/05-2013_community_resilience_understanding_the_impacts_of_disruptions_in_water_service_0.pdf.
9. Gammie L. Chapter 2. Setting up a water utility based surveillance program-a proactive approach Waterborne gastrointestinal disease outbreak detection. Denver, CO: American Water Works Association; 2001.

Acknowledgements

We would like to thank the following individuals for their valuable input and review of this document: Nelson Fok for the question and context; Lorraine McIntyre, Joanne Edwards, Nelson Fok, Sylvia Struck, and Mona Shum for review and comments.