information to be used in evaluating Escherichia coli for inclusion in the

Public health practitioners and water providers have been testing water for microbial contaminants for more than a century. In the United States, public water testing is regulated by federal law. The specifics of the federal law are enforced through such rules as the Total Coliform Rule (TCR). A negotiated rule-making process to amend the TCR is under way. As this process proceeds, it is important that stakeholders have access to the latest information pertinent to the discussion. This article provides information to be used in evaluating Escherichia coli for inclusion in the amended rule as a potential indicator for drinking water monitoring. Advances in the techniques used for coliform and E. coli testing, new understanding about distribution system biofilms, newly emerged waterborne pathogens, and international changes in the use of E. coli as an indicator all combine to underscore the importance of reconsidering the role of E. coli as an indicator organism. E. coli as a public health indicator of drinking water quality BY JON STANDRIDGE E scherichia coli, a gram-negative, rod-shaped bacterium measuring 0.5 by 2.5 µm, is a subset of a group of gram-negative enteric bacteria known collectively as coliforms (Standard Methods, 1998). Total coliforms are classically described as nonspore-forming, gram-negative rods capable of hydrolyzing lactose to acid and gas end products within 48 h at 35 o C. Alternatively, coliforms can be identified by a positive test result for the enzyme galactosidase. E. coli is identified in the laboratory using a suite of reactions including growth at 44.5 o C, hydrolysis of lactose, inability to use citrate as a carbon source, production of indole from tryptophan, and a negative Voges-Proskauer test. Alternatively, E. coli can be identified by a positive test result for the enzymes galactosidase and glucuronidase. BACKGROUND Differentiation of pathogenic E. coli from normal flora E. coli. E. coli is one of the major bacterial inhabitants of the gut of healthy humans and warm-blooded animals (Ewing, 1986). Some 90 99% of all coliforms in fecal material are E. coli bacteria (Stevens et al, 2003; Edberg et al, 2000; Dufour, 1977). Because of the organism s high concentration as normal flora in the gut of healthy animals, E. coli has been identified as an ideal fecal indicator. STANDRIDGE 100:2 JOURNAL AWWA PEER-REVIEWED FEBRUARY 2008 65

In contrast, most bacteriological pathogens do not have strains that can normally inhabit the human gut. For example, the presence of Salmonella or Shigella in the human gut is almost always associated with current or recent intestinal disease (Edberg et al, 2000). In the case of E. coli, however, only certain variants of the organism can cause serious disease. Clinical microbiologists have developed a complex system based on serological and virulence characteristics for categorizing the few strains of E. coli capable of producing disease (Paton & Paton, 1998). These categories include enteropathogenic E. coli, enterotoxigenic E. coli, enteroinvasive E. coli, enteroaggregative E. coli, diffuse adherent E. coli, and enterohemorrhagic E. coli (also known as shiga toxin-producing The Escherichia coli organisms detected as the result of regulatory testing are, in fact, normal flora organisms incapable of causing disease. E. coli). The most commonly occurring of these is the enterohemorrhagic strain known as E. coli O157, which has been the causative agent in several waterborne disease outbreaks (Barwick et al, 2000; Swerdlow et al, 1992; CDC, 1991; O Mahony et al, 1986; Rosenberg et al, 1977; Schroeder et al, 1968). The pathogenic E. coli strains produce symptoms ranging from mild diarrhea to severe watery or bloody diarrhea, cramping, headache, and in some cases hemolytic uremic syndrome, which may result in kidney failure or even death (Barwick et al, 2000; Paton & Paton, 1998). The fact that E. coli has both pathogenic and harmless variants presents a dilemma for the public health community. The E. coli test methods used for routine testing of water samples are neither designed nor intended to detect the pathogenic strains (Standard Methods, 1998). The E. coli organisms detected as the result of regulatory testing are, in fact, normal flora organisms incapable of causing disease. Their presence indicates fecal contamination but not imminent E. coli disease, a fact that often gets lost in the flurry of activity that follows detection of E. coli in a water supply. The public, the news media, and sometimes even utility officials erroneously believe and report that the pathogenic strain has been detected and thus is a threat requiring action. The public notification language in the Public Notification Rule (PNR; USEPA, 2000) exacerbates this tendency toward misinformation by coupling the detection of indicator E. coli with a list of enteric disease symptoms. To avoid disseminating erroneous information, the public health and water industry communities must do a better job of understanding and communicating the concept of indicator testing to protect public health. LITERATURE REVIEW The issues and historic perspectives on the use of indicator bacteria for protecting drinking water supplies have been extensively debated in the literature over the past several years. In their extensive review article, Edberg and colleagues (2000) began and concluded with an unequivocal statement that E. coli should be the indicator of choice. These researchers provided strong historic evidence that E. coli was always the target organism because E. coli makes up > 95% of the intestinal coliform flora but that inadequate detection technology resulted in the creation of more easily performed E. coli surrogates such as total and fecal coliforms. Microbiologists in the first decades of the twentieth century understood and described the deficiencies of the total coliform test. Nonetheless, they recommended its adoption solely on the basis of the simplicity of performing the procedure and the associated likelihood of the test s adoption and widespread use. Edberg and co-workers argued that once a rapid, reliable, easy-to-perform enzymatic test specific for E. coli became available, the surrogates must be abandoned. Other review articles followed a similar format and referenced the same literature as Edberg et al (2000). For example, the Australian government commissioned a discussion paper examining possible changes in the use of coliforms as indicators of drinking water quality (Stevens et al, 2003). It provided extensive justification for discontinuing the use of total coliform as a drinking water indicator. The report concluded that only enterococci and E. coli are adequate indicators of fecal contamination and recommended the use of E. coli as the primary indicator of fecal pollution, supported by other measurements such as heterotrophic bacteria, chlorine residual, and turbidity. Leclerc and colleagues (2001) reviewed advances in the bacteriology of coliforms, focusing on their use in drinking water protection from an international perspective. The researchers concluded that advances in the understanding of glucuronidase-positive enterobacteriaceae dictated the abolition of the fecal coliform designation. Other researchers, discussing the indicator issue from the Canadian perspective, concluded that E. coli appears to provide the best bacterial indication of faecal contamination in drinking water (Tallon et al, 2005). This conclusion was based on several themes that recurred throughout the literature and may be summarized as follows. 66 FEBRUARY 2008 JOURNAL AWWA 100:2 PEER-REVIEWED STANDRIDGE

Newly developed enzymatic methods of detecting E. coli are simple, rapid, specific, and sensitive. Unlike total coliform, E. coli does not multiply in the drinking water environment. E. coli is the only coliform exclusively associated with a fecal source. Thermotolerant coliforms are unacceptable as fecal indicators because of the inclusion of a large number of nonfecal environmental organisms (such as Klebsiella) in the group. In addition to the review articles published over the past five years, several research papers on the issue of drinking water microbial indicators have been published. Regulatory moves toward reliance on E. coli as the fecal indicator of choice have spawned several studies evaluating the efficacy of various testing methodologies. One study demonstrated that the enzymatic tests approved by the US Environmental Protection Agency (USEPA) for enumerating E. coli were equivalent or better than the previously approved fecal coliform tests (Hamilton et al, 2005). A European study offered similar findings of the superiority of the enzymatic methods in Italian laboratories (Bernasconi et al, 2006). In a study of 139 samples from 22 water bodies, Finnish researchers evaluated the ability of the fecal-indicator thermotolerant coliforms, E. coli, Clostridium perfringens, and F-specific ribonucleic acid (F-RNA) bacteriophages to serve as surrogates for specific pathogen testing for Campylobacter spp., Giardia spp., Cryptosporidium spp., and noroviruses (Hörman et al, 2004). The report concluded that the presence of thermotolerant coliforms and E. coli had significant predictive value for the presence of the enteropathogens studied. The absence of the indicators was shown to be predictive of the absence of the pathogens. One of the concerns associated with any fecal contamination indicator is that the organism can be found multiplying and emanating from locations where no fecal contamination has occurred. This is the leading reason total coliform has been challenged as an ideal indicator, and recent articles have voiced this concern about E. coli as well. Beauchamp and colleagues (2006) described environments in the paper and wood pulp industry where E. coli has been shown to multiply at various points in the paper-making process. Complementing this discovery, the researchers then demonstrated the ability of wood chips to promote multiplication of E. coli in laboratory experiments. Other research also reported on detection of E. coli and total coliforms in wood pulping operations in which no fecal contamination had occurred (Tamplin, 2003). Work sponsored by the Canadian Paper Research Institute demonstrated similar findings (Gauthier & Archibald, 2001). Another study published additional confounding findings of E. coli bacteria becoming naturalized members of the soil microbial community (Ishii et al, 2006). That study included bench-top experiments proving that E. coli could easily multiply when inoculated into soils and incubated at 30 37 o C. In addition, Ishii and co-workers pointed out that soil not directly contaminated with fecal material could become a significant source of E. coli detections in tropical climates. Power and colleagues discovered an encapsulated strain of E. coli that has the ability to bloom in Australian lakes serving as drinking water reservoirs (Power et al, 2005). When the water temperature was > 18 o C, E. coli concentrations could quickly rise from 150 cfu/100 ml to 10,000 cfu/100 ml, in the absence of any fecal contamination. This encapsulated strain with the ability to thrive in a relatively clean environment appears to be rare. Although these cases are important to consider when evaluating E. coli as an indicator, the regrowth phenomena reported all stem from unusual situations and occur rarely. STATUS OF E. coli AS AN INDICATOR IN USEPA REGULATIONS Safe Drinking Water Act (SDWA). The SDWA, the federal statute that requires regulation of the drinking water industry, was originally passed by the US Congress in 1974 and subsequently amended in 1986 and 1996 (USEPA, 1996; 1984; 1974). The SDWA defines public water supplies, requires USEPA to identify and prioritize pollutants, specifies maximum contaminant level goals (MCLGs), and establishes enforceable maximum contaminant levels (MCLs) for the prioritized pollutants. The provisions of the SDWA and its amendments are carried out and enforced through the rule-making process. Total Coliform Rule (TCR). In 1989, the USEPA finalized the TCR, which set an MCL for total coliforms and established monitoring and analytical requirements (USEPA, 1989). The rule sets an MCLG of zero, eliminating specific coliform counts from the MCL. The rule introduced the presence/absence concept, i.e., many samples are tested for the presence of coliforms in 100-mL samples and decisions about the microbiological safety of the water system are made by looking at test results as a whole over time. The TCR sets the MCL for systems analyzing 40 or more samples per month such that no more than 5% of samples tested may be positive for total coliforms. For those systems testing fewer than 40 samples per month, no more than one sample per month may be positive. The rule notes the importance of rapid reporting of contamination events and incorporates language toward that end. The TCR was promulgated during an era of change in the way the scientific and engineering communities viewed waterborne disease and the use of indicators for public health protection. The conflicts surrounding the acceptance and assimilation of the scientific changes are often apparent in the document. In the introductory sections of the TCR, the USEPA strongly defends the continued use of total coliforms in setting the MCL. It references the decades of previous use, STANDRIDGE 100:2 JOURNAL AWWA PEER-REVIEWED FEBRUARY 2008 67

the fact that total coliforms usually are present in animal and human feces, and the fact that total coliform detections are often associated with waterborne disease. As stated in the rule, [US]EPA believes that treatment which provides total coliform-free water will reduce fecal pathogens to minimal levels (1989). In spite of this strong statement found early in the document, later in the rule the agency challenges its own stated policy. In the discussion on setting the MCL, the TCR states, given that total coliforms are ubiquitous in water, [US]EPA believes that an infrequent single total coliform positive does not necessarily represent a health risk (1989). Further inconsistencies about the value of total coliform detection in protecting public health can be seen in the section on invalidation of positive test results. This section of the rule begins with discussions of documented situations in which coliform detections were not the result of fecal contamination. In light of these situations, the TCR includes an elaborate protocol that allows the primacy agency to invalidate positive total coliform test results. Criteria for invalidation include laboratory problems, total coliform positives that after completing upstream and downstream resampling are isolated to one sample tap, or any other condition in which the primacy agency determines that the presence of total coliforms does not reflect adverse water quality in the distribution system. During the TCR rule-making process, USEPA received numerous comments suggesting that fecal coliforms and/or E. coli be made the indicator of choice for protection from waterborne disease. In response to these comments, the rule briefly states that fecal coliforms and E. coli are subsets of the total coliform group and therefore are included in the total coliform MCL, thus eliminating the need for separate MCLs. In apparent conflict with this statement, the USEPA for the first time included a requirement for additional testing for fecal coliform or E. coli on all sample cultures testing positive for total coliforms. A positive fecal coliform or E. coli result coupled with a repeat or original sample testing positive for either E. coli/fecal or total coliform precipitates an acute MCL violation that shortens the reporting time to the primacy agency by 24 h and requires public notification within 72 h. In the section discussing these new requirements, the TCR states the presence of fecal coliforms is strong evidence of recent sewage contamination and indicates an urgent public health problem probably exists (USEPA, 1989). The section on public notification language when fecal coliform/e. coli is detected contains strong language about the importance and significance of finding these fecal indicators. Even though fecal coliform/e. coli detection is not specifically listed as an MCL violation, the required notification language states that [US]EPA has determined that the presence of fecal coliform or E. coli is a serious health threat and that more urgent public health notice language is needed when fecal coliforms or E. coli are detected (USEPA, 1989). A close reading of the 1989 rule leaves the impression that USEPA had recognized some of the shortcomings of the public health protection value of total coliform testing but concluded there was insufficient evidence to abandon its use. The TCR also introduced the potential value of fecal coliform/e. coli testing. With the 1989 TCR, USEPA opened discussions of abandoning fecal coliform testing in favor of specific E. coli testing. The promulgation of the TCR occurred during a period in which rapid enzymatic tests specific for E. coli were becoming widely accepted. In response to this occurrence, USEPA requested comments on whether E. coli results could be accepted in lieu of fecal coliform testing. The agency noted that the vast majority of comments advocated elimination of fecal coliform testing in favor of specific E. coli testing. Nevertheless USEPA determined that because fecal coliform testing was simple, inexpensive, and not less protective of public health, its continued use was approved. The TCR marked the first time that tests specific to detecting E. coli are listed for drinking water. It includes extensive detail for add-ons to existing membrane filter and multiple-tube fermentation tests that allow for detecting E. coli. In addition, the rule allows the use of enzymebased ortho-nitrophenyl- -d-galactopyranoside 4-methylumbelliferyl- -d-glucuronide (ONPG MUG) rapid-test systems that simultaneously detect both total coliform and E. coli. PNR. With the PNR, the USEPA began to move toward increasing the importance and role of E. coli or fecal coliform detection in protecting drinking water while decreasing the reliance on and importance of total coliforms (USEPA, 2000). The rule notes the importance of informing drinking water customers about the quality of their water but recognizes the need to differentiate between critical situations that require immediate notification from those that are less pressing. Instances in which people need to know immediately that their drinking water has become contaminated are designated tier 1 notification events. Tier 1 events are defined as violations and situations with significant potential to have serious adverse effects on human health as a result of short-term exposure. Contamination occurrences that do not pose an immediate health risk are classified as tier 2 or tier 3 events. The rule also defines specific language and time frames for the required notification actions. In the PNR, the USEPA reversed its TCR opinion on the public health significance of total coliform detections in drinking water. In spite of numerous comments supporting the continuation of total coliforms as an acute indictor of public health risk, the agency determined that total coliforms are not sufficiently strong or predictable indicators of significant potential of risk from short-term effects (USEPA, 2000). The final rule codifies this thinking by placing the detection of fecal indicators or failure to test for fecal indicators as the only microbiological 68 FEBRUARY 2008 JOURNAL AWWA 100:2 PEER-REVIEWED STANDRIDGE

parameter included in the tier 1 notification category. The PNR specifically states that tier 1 notification is required for violation of the MCL for total coliforms when fecal coliforms or E. coli are present in the water distribution system. Tier 1 notifications must be disseminated within 24 hours and must include the following language: Fecal coliforms and E. coli are bacteria whose presence indicates that the water may be contaminated with human or animal wastes. Microbes in these wastes can cause short-term effects, such as diarrhea, cramps, nausea, headaches, or other symptoms. They may pose a special health risk for infants, young children, some of the elderly, and people with severely compromised immune systems (USEPA, 2000). The PNR defines the right of any primacy agency including the states to elevate any To avoid disseminating erroneous information, the public health and water industry communities must do a better job of understanding and communicating the concept of indicator testing to protect public health. contamination events to tier 1 when it is deemed appropriate. For example, the state of Wisconsin takes a conservative approach in which all total coliform detections in transient noncommunity systems are automatically elevated to tier 1. Detections of total coliforms in the absence of fecal indicators are deemed tier 2 violations. The PNR specifies that TCR notifications of tier 2 violations be made within 30 days and include the following language: Coliforms are bacteria that are naturally present in the environment and are used as an indicator that other, potentially harmful bacteria may be present. Coliforms were found in more samples than allowed, and this was a warning of potential problems (USEPA, 2000). The PNR essentially cleared up the ambiguity concerning the public health significance of total coliform detections found in the 1989 TCR. Surface Water Treatment Rule (SWTR) and associated rules. The Long Term 1 Enhanced SWTR (LT1ESWTR) and Long Term 2 Enhanced SWTR (LT2ESWTR) were promulgated under the SDWA to further protect public health against Cryptosporidium and other microbial pathogens in drinking water (USEPA, 2006b; 2002b). The LT2ESWTR furthered USEPA s progression toward designation of E. coli as the indicator of choice for detecting fecal contamination. The rule specifically requires E. coli testing in addition to and sometimes in place of Cryptosporidium testing. The selection of E. coli as the only indicator allowed is prefaced by a discussion regarding the use of alternatives, including total and fecal coliforms. After a thorough analysis of Information Collection Rule data, the advisory group and USEPA determined that E. coli could be used to identify some of the water sources that were unlikely to exceed a Cryptosporidium concentration of 0.075 oocysts/l, i.e., the level at which filtered public water supply systems must provide additional treatment under the LT2ESWTR. The data were robust enough to set different levels for reservoirs and flowing source waters and allow only E. coli testing for small public water supplies that provide filtration. Ground Water Rule (GWR). The USEPA promulgated the GWR to increase protection against microbial pathogens, specifically viral and bacterial pathogens emanating from fecal contamination at the source (well; USEPA, 2006c). The introductory summary to the GWR asserts that the new rule was necessary because the TCR did not provide adequate fecal contamination protection at the source. The rule states that source water fecal testing is needed to differentiate between distribution system intrusion and contamination at the source. The GWR requires that all sources (wells) be tested for fecal contamination when a total coliform is detected anywhere in the system. Detection of fecal indicators in source water triggers a requirement of four logs of treatment reduction for viruses. The GWR also amends the PNR by adding the detection of fecal indicators in a source water sample as a trigger of a tier 1 acute notification. As a result of the rule, disinfection requirements were placed on many groundwater systems that until this point had no or minimal disinfection in place. The GWR explains the reasoning behind the choice of total coliform detection as the triggering event, noting that total coliform testing was already required by the TCR and therefore provided a no-cost screening tool. The rule cites literature demonstrating that total coliform was indeed present in some well-studied virus contamination events (Borchardt et al, 2004; USEPA, 1998a). In contrast with the LT2ESWTR and PNR, the GWR includes several strong statements emphasizing the importance of total coliform detections in protecting public health. For example, in the discussion of trigger events, the rule states, [US]EPA believes that a total coliform positive sample in the distribution system is an indication of potential microbial contamination of the groundwater system that may have originated from the groundwater source. This is a potentially serious public health risk that warrants followup action (USEPA, 2006c). STANDRIDGE 100:2 JOURNAL AWWA PEER-REVIEWED FEBRUARY 2008 69

The GWR includes a section justifying the use of fecal indicators for the detection of fecal contamination in source water. In an apparent disconnect with the earlier justification of the detection of total coliform as the trigger event for the rule, the rule cites the incidence of fecal contamination of groundwater being the source of viral waterborne disease, thus justifying using indicators of fecal pollution for protecting public health. The rule allows the primacy agency to choose any one of E. coli, enterococci, or coliphage as the fecal indicator for mandated testing. The TCR and GWR include provisions for the primacy agency to discount total coliform positive results emanating from known distribution problems such as biofilm. Total coliform positive samples stemming from The issues and historic perspectives on the use of indicator bacteria for protecting drinking water supplies have been extensively debated in the literature over the past several years. problems in the distribution system (as determined by the primacy agency) would not trigger fecal contaminant testing of the source water. USEPA put this exception in place in response to comments that a single total coliform positive result in the distribution system was too conservative a trigger to force testing for fecal contamination in the source water (USEPA, 2006c). Disinfectants/Disinfection Byproducts Rule (D/DBPR). The Stage 1 D/DBPR and the Stage 2 D/DBPR were developed to protect water consumers from carcinogenic and toxic chemicals formed when common disinfectants used to control microbial pollutants in drinking water combine with naturally occurring organic compounds normally found in water (USEPA, 2006a; 1998b). These negotiated rules were informed by a National Academy of Sciences Federal Advisory Committee Agreement (FACA) (USEPA, 2000). The FACA designated the use of E. coli as the only indicator to be used as a fecal pathogen indicator but provided no background regarding this decision. Clean Water Act. The Federal Water Pollution Control Act, commonly known as the Clean Water Act, directs states, territories, and authorized tribes, with oversight by USEPA, to adopt water quality standards to protect the public health and welfare (USEPA 2002a). The act was amended in 2000 with the Beach Act, which defined coastal waters to include the Great Lakes and set a deadline of April 2004 for all states and territories to establish water quality standards for water bodies within their jurisdiction. These acts are implemented through the rulemaking process. The Water Quality Standards for Coastal and Great Lakes Recreation Waters Final Rule was published in 2004 and issues guidance on the indicators to be used for testing mandated under the Beach Act (USEPA, 2004). The rule s introduction provides an excellent discussion of indicators. USEPA and its predecessor, the Public Health Service, had abandoned the use of total coliforms for indicating the suitability of recreational waters during the 1960s because of their ubiquitous occurrence in surface waters. Instead, fecal coliforms were designated as the indicator of choice. The 2004 rule requires state regulators to abandon fecal coliform testing in favor of testing methods specific for E. coli. The rule describes the evolution of scientific information that forced the change from fecal coliforms to E. coli and enterococci as the indicators of choice for public health protection testing at beaches. The National Pollutant Discharge Elimination System (NPDES) program requires wastewater and other dischargers to test for specific pollutants that comply with the Clean Water Act. In support of the NPDES program, the USEPA published guidelines establishing approved test procedures. In 2002, the agency issued analytical guidelines that relied heavily on referencing standard methods already published by such groups as the US Geological Survey, the American Society for Testing and Materials, and the American Public Health Association (USEPA, 2002c). In 2005, USEPA published an additional guidance document focusing specifically on biological analysis for wastewater and sludge samples (USEPA, 2005). This guidance document, rather than referencing published standard methods, includes both membrane filter and enzymatic tests for E. coli and enterococci submitted as alternative test procedures. Although the document does not eliminate testing for fecal coliforms, it does provide two new methods for E. coli testing in sewage. E. coli REGULATION IN OTHER US FEDERAL AGENCIES US Department of Agriculture (USDA). With regard to fecal contamination of consumable products, the food industry and its regulators face many of the same issues confronting the drinking water community. The estimated 5 million annual cases of foodborne disease and the associated 4,000 annual deaths in the United States have resulted in significant regulation of the food industry from the USDA Food Safety Inspection Service. The regulatory responsibility of the USDA is focused primarily on facilities processing meat, poultry, and eggs. Before 1997, 70 FEBRUARY 2008 JOURNAL AWWA 100:2 PEER-REVIEWED STANDRIDGE

processing plants relied on testing for specific pathogens such as Salmonella, E. coli O157:H7, Campylobacter, and Listeria monocytogenes to protect the food supply. In the early 1990s, this approach was challenged as not protective of public health. Several hearings and scientific conferences were convened, resulting in significant regulatory changes published in the Federal Register (USDA FSIS, 1996). The department concluded that quantitative testing for generic E. coli would be the most effective means of controlling enteric pathogens introduced through fecal contamination in slaughter plants. In the final rule, the USDA cites numerous comments and reports supporting this change, not the least of which is a National Academies of Science report that states at present, E. coli testing is the best indicator of fecal contamination among the commonly used fecal-indicator organisms. The materials supporting the department s final rule are summarized and referenced in the Federal Register document (USDA FSIS, 1996). US Food and Drug Administration (USFDA). The USFDA Center for Food Safety and Applied Nutrition is part of the US Department of Health and Human Services. The center s primary responsibility is to ensure the safety of food products, not including meat, poultry, and egg products regulated by the USDA. For most regulated products, the focus is on enforcing good manufacturing practice controls through mandatory Hazard Analysis and Critical Control Point (HACCP) performance criteria. For example, in the juice industry, the general goal is to use accepted sanitation practices that ensure a 5-log reduction in pathogens from raw foodstuffs to shelf-ready products. USFDA rules provide specific guidance for implementing HACCP criteria that will achieve the 5-log reduction (USFDA, 1998). Worker hygiene, plant design, air handling, cleanup procedures, sanitizers, cooler temperatures, pest control, packaging, and transportation are all outlined in great detail. The USFDA s focus on HACCP implementation, which includes no testing of the final product for fecal contamination, continued with the recent publication of the Guide to Minimize Microbial Food Safety Hazards of Fresh-cut Fruits and Vegetables, Draft Final Guidance (USFDA, 2007). This document includes recommendations for environmental testing of production facilities to monitor the effectiveness of postprocessing sanitation practices but recommends no product testing for fecal contamination. INTERNATIONAL STATUS OF E. coli AS AN INDICATOR OF FECAL CONTAMINATION During the past five years, Australia, Canada, the European Union, and the World Health Organization have all reviewed and considered changes for the role of E. coli as a fecal contamination indicator. The recent regulatory changes occurring internationally are summarized in Table 1. Australia. In Australia, the responsibility for regulation and testing of water supplies lies with the regional governments. Recognizing the advantages of establishing a national policy on drinking water quality, the regional governments formed a coalition to set standards. In 1994, the Council of Australian Governments agreed on a water reform agenda to work toward reform in the water industry at the national level. In 2000, the Australian Government National Health and Medical Research Council (NHMRC) Drinking Water Review Coordinating Group recognized the increasing uncertainty of using traditional indicators as measures of fecal contamination of drinking water. The next year, the NHMRC commissioned a discussion paper examining possible changes in the use of coliforms as indicators of drinking water quality (Stevens et al, 2003). Although not peer-reviewed, this publication provides an excellent review of the issues from both international and Australian perspectives. The key recommendations were that total coliform be removed as an indicator of fecal contamination and replaced with E. coli in the Australian Drinking Water Guidelines (ADWG). In 2004, the guidelines were revised per these recommendations (NHMRC & NRMMC, 2004). The revised guidelines state that total coliforms are no longer to be used as fecal indicator organisms in drinking water but can TABLE 1 Summary of recent regulatory changes occurring internationally Year Entity Total Coliform Fecal Indicator Implemented Other Required Tests Australia Not required Thermotolerant coliforms 2004 None or Escherichia coli Canada Used to detect operational E. coli 2006 None failures or nonfecal intrusions European Union Nonmandatory check E. coli or enterococci 2003 Heterotrophic plate parameter count, Clostridium perfringens World Health Not acceptable Thermotolerant coliforms 2006 None Organization or E. coli STANDRIDGE 100:2 JOURNAL AWWA PEER-REVIEWED FEBRUARY 2008 71

continue to be used for operational monitoring. The ADWG recommends testing for E. coli and/or thermotolerant coliforms to indicate the presence of fecal contamination. Thermotolerant coliforms are defined as gramnegative facultative anaerobic bacilli that can ferment lactose at 44.5 ± 0.2 o C with the production of acid in 24 h in media containing bile salts (i.e., fecal coliforms in US terminology). The document states that although thermotolerant coliform testing may be simpler to perform, E. coli is a better indicator, given that up to 10% of the thermotolerant organisms detected may not be of fecal origin. Australia, the World Health Organization, and the European Union have all undertaken comprehensive reviews of the status of microbial drinking water indicators and have concluded that total coliform testing should be abandoned and only Escherichia coli should be used. Canada. The Canadian government produces national guidelines for drinking water monitoring. In 2004, the Federal Provincial Territorial Committee on Drinking Water prepared a document for public comment, Microbiological Quality of Drinking Water: Escherichia coli, which was revised two years later (Health Canada, 2006a). In this document, E. coli is characterized as a fecal contamination detection test and total coliforms are described as a means of detecting operational failures or nonfecal intrusions into the system. The latest regulatory guidance document, Guidelines for Canadian Drinking Water Quality (GCDWQ), was published in 2006 (Health Canada, 2006b). The guidelines are similar to the USEPA s TCR. They require both total coliform and E. coli testing and establish a standard of zero E. coli in a 100-mL sample and a standard of < 10% of samples testing positive for total coliforms. These Canadian federal standards have been adopted and codified by the provincial governments. For example, Alberta follows the GCDWQ as outlined in Alberta Regulation 277/2003, Environmental Protection and Enhancement Act (Alberta, 2003). British Columbia also adheres to the GCDWQ with the exception that fecal coliform analysis is allowed as a surrogate for E. coli (British Columbia, 2001). European Union. In 1998, the Council of the European Union published Council Directive 98/83/EC prescribing requirements on the quality of water intended for human consumption (European Union, 1998). The directive states that by Dec. 25, 2003, all members of the European Union are required to meet the criteria set in the directive and are free to implement more stringent standards. Microbiological parameters are divided into two categories. The first category is the audit group of required tests, which includes E. coli and/or enterococci. The directive establishes a standard of zero organisms per 100 ml for this category. The second category, termed the check parameter (nonmandatory) test list, includes C. perfringens, heterotrophic bacteria grown at 22 o C, and total coliforms. World Health Organization. The World Health Organization recently published new guidelines for protecting drinking water in third-world countries (WHO, 2006). In the chapter on microbial aspects of drinking water, the document recommends testing for E. coli or thermotolerant coliforms in any water used directly for consumption, water leaving treatment facilities, and water within distribution systems. The guideline values for verification of microbiological quality specify that these organisms must not be detectable in any 100- ml sample. The guidelines specifically state that total coliform bacteria are not acceptable indicators of the sanitary quality of water supplies, particularly in tropical areas, where many bacteria of no sanitary significance occur in almost all untreated supplies (WHO, 2006). VALUE OF FECAL INDICATORS IN DETECTING OTHER CONTAMINANTS Among the goals in choosing an indicator organism for drinking water monitoring is that the indicator predict not only the presence of fecal contaminants but also the presence of other waterborne disease organisms. One argument for continuing total coliform testing is that its presence may be indicative of other nonfecal microorganisms of potential health concern. According to the operative premise, because most of the nonfecal pathogens found in water distribution systems emanate from biofilms and because coliforms can be associated with biofilms, the presence of total coliform is a good indicator of nonfecal pathogens. This premise must be carefully considered. The most important nonfecal waterborne pathogens are those capable of multiplying in the water distribution system; these include Legionella, Mycobacteria, and the protozoans Acanthamoeba and Naegleria fowleri. Research has shown that Mycobacteria enter the distribution system through leaking pipes, valves, joints and seals, crossconnections and backflows, finished water storage vessels, improper treatment of materials, or during water treatment breakthroughs (Vaerewijck et al, 2005). These 72 FEBRUARY 2008 JOURNAL AWWA 100:2 PEER-REVIEWED STANDRIDGE

organisms grow and persist when conditions are favorable and are often associated with biofilms. Although coliforms can be present in a biofilm community, they are seldom the predominant organism. In fact, biofilms may have few or no coliforms associated with them (Silhan et al, 2006; LeChevallier et al, 1996). Reliance on total coliform testing to detect biofilm occurrence would be a risky public health practice. Other options, such as specific testing for the free-living pathogens of concern, may be more appropriate. Depending on fecal indicators or total coliform testing to protect the public from the free-living pathogens or the opportunistic bacterial pathogens Pseudomonas, Flavobacterium, Acinetobacter, Klebsiella, Serratia, and Aeromonas creates a false sense of security. CONCLUSION Caveats regarding the use of E. coli as a drinking water indicator organism. According to the commonly accepted definition of an ideal indicator of drinking water pollution, the indicator must occur in proportion to the pollution, never be present in safe water, always be present in contaminated water, not multiply in the environment, die off in the environment more slowly than pathogens, be easily detected in the laboratory, and be safe to work with. E. coli generally fulfills these criteria for an ideal indicator for drinking water monitoring, but it is not without flaws. Until recently, it was thought that E. coli would not multiply appreciably outside the gut of warmblooded animals. Research summarized in this article challenges this assumption. Organism multiplication must be considered in light of the Australian report of an encapsulated E. coli blooming in warm reservoirs (Powers et al, 2005) as well as reports of E. coli growing in paper mills (Beauchamp et al, 2006; Gauthier & Archibald, 2001). Concerns raised about E. coli dying off at vastly different rates in some aquatic environments warrant additional thought. The confusion that results from the fact that E. coli occurs as both a pathogen and a fecal indicator also merits attention. Any decision to use E. coli as the primary indicator of drinking water quality must be accompanied by a commitment to educate the public in order to allay any concerns stemming from this confusion. Yet another shortcoming of E. coli as an indicator is that it does not indicate the presence of nonfecalorigin pathogens such as Legionella or Mycobacteria. REFERENCES Alberta, 2003. Alberta Regulation 277, Environmental Protection and Enhancement Act, Potable Water Regulation. Saskatchewan, Alta., http://www.qp.gov.ab.ca/documents/regs/ 2003_277.cfm?frm_isbn=0779723023. Barwick, R.S.; Levy, D.A.; Craun, G.F.; Beach, M.J.; & Calderon, R.L., 2000. Surveillance for Waterborne-Disease Outbreaks United States, 1997 1998. Morbidity & Mortality Weekly Rept., 49:1. Beauchamp, C.J.; Simao-Beaunoir, A.-M.; Beaulieu, C.; Chalifour, F.-P., 2006. Confirmation of E. coli Among Other Thermotolerant Coliform Bacteria in Paper Mill Effluents, Wood Chips Screening Rejects, and Paper Sludges. Water Res., 40:12:2452. Bernasconi, C.; Volponi, G.; & Bonadonna, L., 2006. Comparison of Three Different Media for the Detection of E. coli and Coliforms in Water. Water Sci. & Technol., 54:3:141. Borchardt, M.A.; Haas, N.L.; & Hunt, R.J., 2004. Vulnerability of Drinking Water Wells in LaCrosse, Wisconsin, to Enteric-Virus Contamination From Surface Water Contributions. Appl. & Envir. Microbiol., 70:10:5937. British Columbia, 2001. British Columbia Drinking Water Protection Act (SBC 2001), Chapter 9. Victoria, B.C., http://www.qp.gov.bc.ca/statreg/stat/d/01009_01.htm. CDC (Centers for Disease Control and Prevention), 1991. Waterborne Disease Outbreaks, 1989 1990. Morbidity & Mortality Weekly Rept., 40:1. Dufour, P., 1977. Escherichia coli : The Fecal Coliform. Spec. Tech. Publ. 65, ASTM, Philadelphia. Edberg, S.C.; Rice, E.W.; Karlin, R.J.; & Allen, M.J., 2000. Escherichia coli : The Best Biological Drinking Water Indicator for Public Health Protection. Jour. Appl. Microbiol., 88:106S. European Union, 1998. Council Directive 98/83/EC of 3 November 1998 on the Quality of Water Intended for Human Consumption. Official Jour. European Communities, 5.12.98:L330:32. Ewing, W.H., 1986. Edward s and Ewing s Identification of Enterobacteriaciae (4th ed.). Elsevier Science Publ., New York. Gauthier, F., & Archibald F., 2001. Ecology of Fecal Indicator Bacteria Commonly Found in Pulp and Paper Mill Water Systems. Water Res., 35:9:2207. Hamilton, W.P.; Kim, M.; & Thackston, E.L., 2005. Comparison of Commercially Available Escherichia coli Enumeration Tests: Implications for Attaining Water Quality Standards. Water Res., 39:20:4869. Health Canada, 2006a. Microbiological Quality of Drinking Water: Escherichia coli. http://www.hc-sc.gc.ca/ewh-semt/pubs/watereau/doc_sup-appui/escherichia_coli/index_e.html. Health Canada, 2006b. Guidelines for Canadian Drinking Water Quality. http://www.hc-sc.gc.ca/ewh-semt/pubs/water-eau/doc_supappui/sum_guide-res_recom/index_e.html. Hörman, A.; Rimhanen-Finne, R.; Maunula, L.; von Bonsdorff, C.-H.; Torvela, N.; Heikinheimo, A.; & Hänninen, M.-L., 2004. Campylobacter spp., Giardia spp., Cryptosporidium, Noroviruses, and Indicator Organisms in Surface Water in Southwestern Finland, 2000 2001. Appl. & Envir. Microbiol., 70:1:87. Ishii, S.; Ksoll, W.B.; Hicks, R.E.; & Sadowsky, M.J., 2006. Presence and Growth of Naturalized Escherichia col in Temperate Soils From Lake Superior Watersheds. Appl. & Envir. Microbiol., 72:1:612. STANDRIDGE 100:2 JOURNAL AWWA PEER-REVIEWED FEBRUARY 2008 73

E. coli has been labeled inferior to total coliforms because it does not have the sensitivity of total coliforms to indicate distribution anomalies such as biofilms or nonfecal contamination occurrences. With the introduction and acceptance of enzymebased test systems, Escherichia coli testing is now simple, fast, and inexpensive. Literature support of E. coli as a drinking water indicator organism. The body of literature, including review articles and research reports, overwhelmingly supports the use of E. coli as the indicator of choice for protecting drinking water. This support may be attributable, in part, to the inadequacies of total coliform as an ideal indicator. Reasons for discounting total coliforms include their ability to grow in distribution systems, their ubiquitous presence in natural environments (such as soil) with no evidence of recent fecal contamination, and the fact that they have not been proven to be sensitive indicators in actual waterborne outbreaks. The literature commonly cites the following reasons to support the choice of E. coli as the best indicator: The technology available for detecting E. coli has vastly improved. Coliform testing was originally chosen because there were simple tests available, whereas specific E. coli testing involved complex multiple steps. With the introduction and acceptance of enzyme-based test systems, E. coli testing is now simple, fast, and inexpensive. E. coli is a major constituent of the fecal flora of warm-blooded animals. E. coli survives but does not multiply in the aquatic environment. Few false-negative results occur; fecal contamination rarely occurs without detectable E. coli. Few false-positive results occur; almost all E. coli detections in drinking water can be traced to contamination. International acceptance of E. coli as a drinking water indicator organism. Australia, the World Health Organi- LeChevallier, M.W.; Welch, N.J.; & Smith, D.B., 1996 Full-scale Studies of Factors Related to Coliform Regrowth in Drinking Water. Appl. & Envir. Microbiol., 62:7:2201. Leclerc, H.; Mossel, D.A.A.; Edberg, S.C.; & Struijk, C.B., 2001. Advances in Bacteriology of the Coliform Group: Their Suitability as Markers of Microbial Water Safety. Ann. Rev. of Microbiol., 55:201. NHMRC (National Health and Medical Research Council) & NRMMC (Natural Resource Management Ministerial Council), 2004. Australian Drinking Water Guidelines. ISBN 1864961244, NHMRC and NRMMC, Canberra, Australia. O Mahony, M.C.; Noah, N.D.; Evans, B.; Harper, D.; Rowe, B.; Lowes, J.A.; Pearson, A.; & Goode, B., 1986. An Outbreak of Gastroenteritis on a Passenger Cruise Ship. Jour. Hyg., (Cambridge) 97:229. Paton, J.C. & Paton, A.W., 1998. Pathogenesis and Diagnosis of Shigatoxin Producing Escherichia coli Infections. Clinical Microbiol. Rev., 11:3:450. Power, M.L.; Littlefield-Wyer, J.; Gordon, D.M.; Veal, D.A.; & Slade, M.B., 2005. Phenotypic and Genotypic Characterization of Encapsulated Escherichia coli Isolated From Blooms in Two Australian Lakes. Envir. Microbiol., 7:5:631. Rosenberg, M.L.; Koplan, J.P.; Wachsmuth, I.K.; Wells, J.G.; Gangarosa, E.J.; Guerrant, R.L.; & Sack, D.A., 1977. Epidemic Diarrhea at Crater Lake From Enterotoxigenic Escherichia coli. A Large Waterborne Outbreak. Ann. Intern. Med., 86:714. Schroeder, S.A.; Caldwell, J.R.; Vernon, T.M.; White, P.C.; Granger, S.I.; & Bennett, J.V., 1968. A Waterborne Outbreak of Gastroenteritis in Adults Associated With Escherichia coli. Lancet, 6:737. Silhan, J.; Corfitzen, C.B.; & Albrechtsen, H.J., 2006. Effect of Temperature and Pipe Material on Biofilm Formation and Survival of Escherichia col in Used Drinking Water Pipes: A Laboratorybased Study. Water Sci. & Technol., 54:3:49. Standard Methods for the Examination of Water and Wastewater, 1998 (20th ed.). APHA, AWWA, WEF, Washington. Stevens, M.; Ashbolt, N.; & Cunliffe, D., 2003. Recommendations to Change the Use of Coliforms as Microbial Indicators of Drinking Water Quality. National Health and Medical Research Council, ISBN: 1864961651, online ISBN: 1864961597, http://www.nhmrc.gov.au/publications/synopses/_files/eh32.pdf. Swerdlow, D.L.; Woodruff, B.A.; Brady, R.C.; Griffin, P.M.; Tippen, S.; Donnell, H.D.; Geldereich, E.; Payne, B.J.; Meyer, A.; & Wells, G.J., 1992. A Waterborne Outbreak in Missouri of Escherichia coli O157:H7 Associated With Bloody Diarrhea and Death. Ann. Intern. Med., 117:812. Tallon, P.; Magajna, B.; Lofranco, C.; & Leung, K., 2005. Microbial Indicators of Faecal Contamination in Water: A Current Perspective. Water, Air & Soil Pollution, 166:1 4:139. Tamplin, M.L., 2003. Application and Suitability of Microbiological Tests for Fecal Bacteria in Pulp Mill Effluents: A Review. Water Quality Res. Jour. of Canada, 38:2:211. USDA (US Department of Agriculture) FSIS (Food Safety Inspection Service), 1996. Pathogen Reduction; Hazard Analysis and Critical Control Point (HACCP) Systems; Final Rule. Fed. Reg., 9 CFR Part 304, et al. 61:144:38806, http://www.fsis.usda.gov/oppde/ rdad/frpubs/93-016f.pdf. USEPA (US Environmental Protection Agency), 2006a. National Primary Drinking Water Regulations: Stage 2 Disinfectants and Disinfection Byproducts Rule; Final Rule. 40 CFR Parts 9, 141, and 142, Fed. Reg., 71:2:388. 74 FEBRUARY 2008 JOURNAL AWWA 100:2 PEER-REVIEWED STANDRIDGE