Page 4: Guidance on Waterborne Bacterial Pathogens

Part B. Supporting information

B.1 Waterborne faecal pathogens

Waterborne faecal pathogens are microorganisms that can occur in water as a result of contamination from human or animal faeces and cause gastrointestinal illness. They are often associated with bacteria from the Enterobacteriaceae family. Faecal bacteria that are well established as having a history of being responsible for waterborne outbreaks of gastrointestinal illness include pathogenic Escherichia coli, Salmonella, Shigella, Campylobacter and Yersinia.

B.1.1 Pathogenic Escherichia coli

Escherichia coli are bacteria found naturally in the digestive tracts of warm-blooded animals, including humans. As such, E. coli are used in the drinking water industry as the definitive indicator of recent faecal contamination of water. Escherichia coli are Gram-negative, facultative anaerobic, rod-shaped bacteria, approximately 0.5-2 µm in size (AWWA, 2006). Whereas most strains of E. coli are non-pathogenic, some possess virulence traits that enable them to cause serious diarrhoeal infections in humans. These pathogenic E. coli are divided into groups based on the mechanisms with which they interact with the human intestinal tract and cause symptoms (e.g., some produce specific types of toxin, whereas others invade, bind to or cause structural alterations of intestinal cells) (Percival et al., 2004). The six groups are enterohaemorrhagic (EHEC), enterotoxigenic (ETEC), enteroinvasive (EIEC), enteropathogenic (EPEC), enteroaggregative (EAEC) and diffuse adherent (DAEC) E. coli (Percival et al., 2004; AWWA, 2006). The EHEC group has emerged as a group that is of particular significance to the water industry (AWWA, 2006). This broad group contains many different serotypes that have been implicated as causes of human illness (Muniesa et al., 2006).

One member of the EHEC group, E. coli O157:H7, has been most commonly associated with pathogenic E. coli outbreaks worldwide (Muniesa et al., 2006) and has been implicated in a few waterborne outbreaks (Bruce-Grey-Owen Sound Health Unit, 2000; Schuster et al., 2005; Craun et al., 2006; Clark et al., 2010). In Canada, the Walkerton outbreak of 2000 was the first documented outbreak of E. coli O157:H7 infection associated with a Canadian municipal water supply and the largest multibacterial waterborne outbreak in the country to date (Bruce-Grey-Owen Sound Health Unit, 2000). Surveillance reports published for other countries have indicated that over the period from 1990 to the early 2000s, E. coli O157:H7 was identified as the causative agent of approximately 6% of the reported drinking water outbreaks in England and Wales (Smith et al., 2006) and roughly 7% of those reported in the United States (Craun et al., 2006).

Cattle and human sewage are the primary and secondary sources, respectively, of EHEC (Jackson et al., 1998; Percival et al., 2004; Gyles, 2007), but human sewage is the major source of the other pathogenic E. coli groups (Percival et al., 2004; AWWA, 2006). Transmission of pathogenic E. coli occurs through the faecal-oral route, and the primary routes of exposure are from contaminated food or water or by person-to-person transmission (Percival et al., 2004; AWWA, 2006). Pathogenic E. coli are not usually a concern in treated drinking water when treatment and distribution systems are properly operated and maintained. However, outbreaks of E. coli O157:H7 involving consumption of drinking water contaminated with human sewage or cattle faeces have been documented in North America (Olsen et al., 2002), including some fatal outbreaks(Swerdlow et al., 1992; Novello, 1999; Bruce-Grey-Owen Sound Health Unit, 2000). The probability of becoming ill depends on the number of organisms ingested, the health status of the person and the resistance of the person to the organism or toxin (LeChevallier et al., 1999).

With the exception of EHEC, most pathogenic E. coli require a high number of bacteria to be ingested in order to produce illness. Infectious dose estimates for non-EHEC strains range from 105 to 1010 organisms (Percival et al., 2004). EHEC strains, in contrast, have a very low infectious dose. It has been suggested that ingestion of fewer than 100 cells may be sufficient to cause infection (Percival et al., 2004; Pond, 2005). The onset and duration of pathogenic E. coli-related illness will be strain dependent, but symptoms can begin in as little as 8-12 hours and last from a few days up to a few weeks (Percival et al., 2004).

Pathogenic E. coli can cause diarrhoea that ranges in severity from mild and self-limiting to severe and life-threatening (Percival et al., 2004; AWWA, 2006). Most non-EHEC illness is marked by a watery diarrhoea that can be accompanied by vomiting, abdominal pain, fever and muscle pain, depending on the group or strain involved.

EHEC illness can begin with watery and bloody diarrhoea in combination with vomiting, but in some cases can progress to the more serious and potentially life-threatening symptoms of haemorrhagic colitis (grossly bloody diarrhoea) and haemolytic uraemic syndrome (kidney failure). These symptoms are caused by shiga-like toxins, potent toxins that are related to Shigella dysenteriae toxins (Percival et al., 2004; AWWA, 2006). It has been suggested that up to 10% of E. coli O157:H7 infections can progress to haemolytic uraemic syndrome (Moe, 1997; Sherman et al., 2010). Children and the elderly are most susceptible to the complications that arise from EHEC infections (Percival et al., 2004). One area of recent interest has been the possible long-term health effects in adults as a result of contracting haemolytic uraemic syndrome from E. coli O157:H7, as these to date have been largely unknown (Clark et al., 2010). Clark et al. (2010) reported on the results of a health study among persons who developed gastrointestinal illness or remained asymptomatic following exposure to E. coli O157:H7 and Campylobacter during the Walkerton outbreak in May 2000. The authors concluded that increases in the incidences of hypertension, cardiovascular disease and indicators of kidney impairment were evident in persons who experienced acute gastroenteritis during the outbreak. Further study in this area is required, but because such waterborne outbreaks are rare, there are very limited opportunities for such studies.

B.1.1.1 Treatment technology

In the majority of treatment and disinfection studies involving pathogenic E. coli, the EHEC strain O157:H7 has been selected as the model organism because of its health significance and prominence in foodborne and waterborne outbreaks. Regardless, review of the evidence generated to date suggests that the proper application of water treatment and disinfection technologies will be capable of controlling strains of pathogenic and non-pathogenic E. coli in drinking water (Percival et al., 2004; AWWA, 2006).

In terms of chlorine and monochloramine effectiveness, laboratory studies have demonstrated E. coli O157:H7 log inactivation capabilities of up to 4 log at concentrations and contact times that would be encountered in municipal drinking water treatment (Rice, 1999; Wojcicka et al., 2007; Chauret et al., 2008).

For ultraviolet (UV) disinfection, Zimmer-Thomas et al. (2007) observed log inactivations of 4.5 log or greater for E. coli O157:H7 at all tested doses of low-pressure and medium-pressure UV. These included UV doses commonly used in water disinfection (20 and 40 mJ/cm2, low pressure), as well as low doses intended to be representative of compromised UV dose delivery (5 and 8 mJ/cm2, low and medium pressure). In UV inactivation experiments, Sommer et al. (2000) observed considerable divergence in sensitivity of different pathogenic (including enterohaemorrhagic) strains of E. coli. The authors further demonstrated that a UV dose of 125 J/m2 (equivalent to 12.5 mJ/cm2) was sufficient to produce a 6 log inactivation of all of the strains under study.

Further information on treatment technologies for E. coli can be found in the E. coli guideline technical document (Health Canada, 2012).

B.1.1.2 Assessment

Studies have shown that the survival and susceptibility to disinfection of pathogenic E. coli strains approximate those of typical E. coli (LeChevallier et al., 1999; Rice, 1999). Also, although routine examination methods for E. coli are not designed to distinguish pathogenic E. coli strains from the general E. coli population, the latter will always occur in greater concentration in faeces, even during outbreaks. Pathogenic E. coli will not occur in the absence of generic E. coli. As a result, the presence of E. coli is the best available indicator of faecal contamination and the potential presence of faecal pathogens, but is not a specific signal for the presence of pathogenic E. coli.

B.1.2 Salmonella and Shigella

Salmonella and Shigella are agents of gastrointestinal illness that belong to the same microbiological family as E. coli, Enterobacteriaceae.

Salmonella are non-spore-forming, facultative anaerobic, Gram-negative bacilli that are 2-5 µm long and 0.8-1.5 µm wide (AWWA, 2006). Salmonella is a complex taxonomic genus consisting of over 2000 different varieties or serological types that can cause infections in animals and humans (AWWA, 2006). According to experts, the genus is officially made up of only two species, Salmonella enterica and Salmonella bongori (Percival et al., 2004; AWWA, 2006). Salmonella enterica is the species of most relevance for human infections, and it can be further broken down into six subspecies, of which one, Salmonella enterica subsp. enterica, contains the majority of serotypes that are associated with cases of human gastroenteritis (Percival et al., 2004). By convention, when referring to Salmonella serotypes, the serotype is adopted as the species name (e.g., Salmonella enterica subsp. enterica serovar enteritidis becomes Salmonella enteritidis).

The vast majority of Salmonella serotypes encountered in developed countries are zoonotic pathogens. Reservoirs for these organisms include poultry, pigs, birds, cattle, cats, dogs, rodents and turtles (AWWA, 2006). Infected humans and, as a result, sewage are also sources of Salmonella. Transmission of Salmonella occurs through the faecal-oral route, predominantly through food. By comparison, drinking water is not often implicated as a source of Salmonella infection (Percival et al., 2004). As Salmonella is a zoonotic pathogen, runoff from agricultural lands can provide a mechanism for the transfer of animal faecal wastes to source waters.

Shigella are facultative anaerobic, non-sporulating, non-motile, Gram-negatives rods 0.3-1.5 µm in diameter and 1-6.5 µm in length (AWWA, 2006). The taxonomy of Shigella is much simpler than that of Salmonella. The genus is categorized into four major serological groups: dysenteriae, flexneri, boydii and sonnei. Shigella sonnei and Shigella flexneri are the two species of importance as causes of gastrointestinal illness in developed countries (Percival et al., 2004). Infected humans are the only significant reservoir (AWWA, 2006). Transmission is faecal-oral, through drinking water or through food that has been contaminated with human faecal wastes. Person-to-person transmission is also a significant route of exposure for Shigella, particularly among children. Shigella is a human-specific pathogen and is not expected to be found in the environment (AWWA, 2006). Thus, contamination of water supplies is suggestive of a source of human faecal contamination, such as from sewage or on-site wastewater disposal systems.

Numerous outbreaks linked to contaminated drinking water have been reported worldwide (Boring et al., 1971; White and Pedersen, 1976; Auger et al., 1981; Arnell et al., 1996; Angulo et al., 1997; Alamanos et al., 2000; R. Taylor et al., 2000; Chen et al., 2001). Schuster et al. (2005) reported that Shigella and Salmonella were identified as the causative agents in 9 and 16 confirmed, proposed or suspected drinking water outbreaks in Canada, respectively, over the years 1974-2001. In the United States, Salmonella and Shigella accounted for approximately 2% and 5% of drinking water outbreaks reported from 1991 to 2002, according to U.S. surveillance data (Craun et al., 2006). Common causes of waterborne outbreaks by these organisms are poor source water, inadequate treatment or post-treatment contamination (e.g., by cross-connections) (AWWA, 2006). Both organisms give rise to acute, self-limiting gastrointestinal illness with symptoms of diarrhoea, vomiting and abdominal pain. Shigella-associated illness is more dysenteric in nature, marked by a more watery diarrhoea containing blood and mucus (AWWA, 2006). Once infected, recovering individuals may continue to shed either of these organisms in their faeces for days up to several weeks or months. Published reports regarding the median infective doses for these two organisms have suggested that they may be as low as 103-105 organisms for Salmonella serotypes and 102-103 organisms for Shigella flexneri and Shigella sonnei (Hunter, 1997; Kothary and Babu, 2001). The factors that contribute to the virulence of these organisms are still under investigation. Both possess mechanisms that enable the bacteria to invade, survive, replicate and disrupt the function of the human intestinal lining (Percival et al., 2004). In addition, Shigella sonnei and Shigella flexneri are known to produce an exotoxin that affects intestinal water absorption and retention (Percival et al., 2004).

B.1.2.1 Treatment technology

Salmonella and Shigella survival characteristics in water and their susceptibility to disinfection have been demonstrated to be similar to those of coliform bacteria, including E. coli (McFeters et al., 1974; Mitchell and Starzyk, 1975; Chang et al., 1985; Koivunen and Heinonen-Tanski, 2005). It is generally recognized that a properly operated facility will be sufficient in controlling Salmonella and Shigella in treated drinking water (AWWA, 2006).

B.1.2.2 Assessment

The absence of E. coli during routine verification should be an adequate indication of the sufficient removal and inactivation of Salmonella and Shigella.

B.1.3 Campylobacter and Yersinia

Campylobacter are pathogenic bacteria found primarily in the intestinal tracts of domestic and wild animals, especially birds. Poultry, cattle, sheep and pigs are considered significant reservoirs for these organisms (Percival et al., 2004; AWWA, 2006). Campylobacter are motile, Gram-negative, slender, curved rods 0.2-0.5 µm wide and 0.5-5 µm long. Yersinia can be found in the faeces of wild animals as well as domestic livestock such as cattle, pigs and sheep (Percival et al., 2004). Yersinia are facultative anaerobic, Gram-negative, non-sporulating rods 0.5-0.8 µm in diameter and 1-3 µm in length (AWWA, 2006). It is the Campylobacter species C. jejuni, C. coli and C. upsaliensis and the Yersinia species Y. enterocolitica that are most important to the water industry (AWWA, 2006). Human sewage also contains large numbers of both of these organisms.

Both Campylobacter and Yersinia enterocolitica are transmitted through the faecal-oral route, mostly through contaminated food and sometimes through water (Percival et al., 2004). Person-to-person transmission of Campylobacter or Yersinia enterocolitica is uncommon (Percival et al., 2004; AWWA, 2006).

Waterborne outbreaks of gastroenteritis involving Campylobacter jejuni and Yersinia enterocolitica have been recorded on numerous occasions, with improper treatment, post-treatment contamination or consumption of untreated water supplies being the most frequent causes (Eden et al., 1977; McNeil et al., 1981; Mentzing, 1981; Vogt et al., 1982; Taylor et al., 1983; Lafrance et al., 1986; Sacks et al., 1986; Thompson and Gravel, 1986). In a review of Canadian data on waterborne outbreaks for the period spanning from 1974 to 2001, Schuster et al. (2005) reported that Campylobacter was implicated in 24 outbreaks and was second only to Giardia (51 outbreaks) in outbreaks where a causative agent was identified. The most notable Canadian waterborne outbreak was the May 2000 Walkerton outbreak involving Campylobacter and E. coli O157:H7, where faecally contaminated well water was not properly treated before consumption (Clark et al., 2003). No outbreaks of Yersinia-related gastroenteritis have been reported for municipal drinking water supplies in North America over the past two decades (Schuster et al., 2005; Craun et al., 2006).

Gastroenteritis caused by Campylobacter typically presents as flu-like symptoms and/or abdominal pain, followed by a profuse watery diarrhoea caused by the presence of an enterotoxin similar to cholera toxin (AWWA, 2006). An important characteristic of Campylobacter is the high infectivity potential; as few as 1000 organisms can cause infection (Black et al., 1988; Hara-Kudo and Takatori, 2011). Yersinia enterocolitica can cause a variety of symptoms, depending on the age of the person infected, but the most commonly observed are gastrointestinal illness, fever and occasionally vomiting in children (AWWA, 2006). The gastrointestinal illnesses caused by both organisms are considered to be self-limiting (Percival et al., 2004).

B.1.3.1 Treatment technology

Studies have demonstrated the susceptibility of Campylobacter species and Yersinia enterocolitica to disinfectants commonly used in water treatment (Blaser et al., 1980; Wang et al., 1982; Sobsey, 1989; Lund, 1996; Rose et al., 2007). It is generally recognized that treatment technologies effective in the removal and inactivation of E. coli will be effective against these pathogenic bacteria (AWWA, 2006).

B.1.3.2 Assessment

Studies have suggested a lack of a correlation between indicator organisms (e.g., E. coli, total coliforms) and the presence of Campylobacter and Yersinia in raw surface water supplies (Carter et al., 1987; Lund, 1996; Hörman et al., 2004). Thus, E. coli may not be an adequate indicator of the presence of both C. jejuni and Y. enterocolitica in source waters at all times. However, as it is expected that properly operated treatment and disinfection technologies are effective in controlling these organisms in treated drinking water, it is expected that the E. coli guideline is sufficiently protective against their potential presence.

Report a problem or mistake on this page
Please select all that apply:

Thank you for your help!

You will not receive a reply. For enquiries, contact us.

Date modified: