OUTBREAK OF E. COLI O157:H7 ASSOCIATED WITH BATHING AT A PUBLIC BEACH IN THE MONTREAL-CENTRE REGION

Escherichia coli O157:H7 is usually transmitted through food, especially inadequately cooked ground beef. Waterborne outbreaks have also been attributed to drinking water and water used in aquatic activities. Outbreaks associated with swimming water in natural bodies of water are much less frequent^(1-4). We are reporting here the first swimming water-associated E. coli O157:H7 outbreak discovered in Montreal-Centre.

Between 15 August 2001, and 21 August 2001, four E. coli O157:H7 cases were reported to the Montreal-Centre Public Health Department. The cases involved four young boys between 3 and 7 years of age. The epidemiologic investigations indicated that the swimming water at a public beach in the Montreal area could be the exposure factor common to all cases. The four boys went in the water between 28 July and 12 August. They visited the beach between one and four times. The symptoms began between 5 August and 16 August. The boys lived in completely different areas of the city and did not know one another. On the basis of this information, the swimming water was suspected as being the most probable source of infection. An environmental investigation, pulsed-field electrophoresis of case isolates, and a case-control study were done.

Environmental investigation

The temperature in the Montreal area during the exposure period was much hotter than the usual averages, with high humidity, resulting in a humidex of approximately 40° C. The public beach was an area of approximately 200 metres by 500 metres, and the swimming area was smaller than the beach dimensions. The sanitary facilities were dry toilets; there were two fountains and a tap marked "washing". A shower with several open air outlets completed the facilities. At the time of the visit, there was no soap and no paper for hand wiping. The water drains flowed into a wood gutter, but it is not known where the wastewater ended up. In short, this swimming environment can be described as rustic with few sanitary facilities, especially for hand washing. There were no farms or animals in the vicinity. There was a restaurant nearby, but none of the cases had eaten food prepared in this restaurant. Parents mentioned during the investigation that there were many people at the beach and in the water when they were there.

During the summer of 2001, there was a swimming water monitoring program. The results of the microbiologic analyses for the months of July and August 2001 were as follows (City of Montreal shore water monitoring program):

Note: In the beach monitoring program of Quebec's environment department, the Ministère de l'environnement (MEVQ), beaches are closed when the coliform concentration exceeds 200 fecal coliforms/100 mL.

Microbiologic analyses

Pulsed-field electrophoresis was done on the four strains. Two strains were E. coli O157:H7 pulsotype 204. The other strains were pulsotype 208 and 209 (three bands and one band different from pulsotype 204 respectively). Additional analyses with restriction enzymes were done; they confirmed that three different strains were involved. Note that the basic criteria for E. coli O157:H7 are very strict and that an identical profile is needed to establish an association.

Follow-up epidemiologic analyses

Although we had found three different strains, we continued to suspect that the water exposure was the common source of the infection since the only exposure common to these reported cases over a brief period of time was the swimming water, and the affected individuals did not know each other and lived in four very different areas. To assess the probability that the swimming water exposure was not a coincidence, we did a case-control study to determine exposure at this public beach in the same age group in the general population. The controls were chosen from among cases of other reportable diseases. We chose pertussis because it does not share the same acquisition risk factors as an enteric disease and because it mainly affects young children. We wished to obtain four controls per case. However, owing to the low incidence of pertussis in the fall of 2001, we were able to obtain only 13 controls. The cases were matched for age. We asked the individual controls whether they had bathed at a public beach in the Montreal area during the risk period and, if so, which beach. The controls were recruited between the months of September 2001 and December 2001.

Among the controls, one of the 13 had been to the public beach during the target period. Therefore, the probability that the four cases had bathed at the public beach can be attributed to chance is minimal. Thus, we believe that the subjects probably acquired their infection through the swimming water, even though the pulsed-field electrophoresis yielded three different profiles.

Discussion

This is the first swimming water-associated outbreak of E. coli O157:H7 discovered in the Montreal area. Although E. coli O157:H7 is a reportable disease, the number of cases associated with this outbreak is probably underestimated, representing only the most severe cases: three of the four cases were hospitalized, but did not have hemolytic uremic syndrome.

The environmental conditions determined to have existed were favourable for transmission of this type. The outbreak reported here shows similarities to an outbreak reported in the United States. In the published study on that outbreak, the age of the subjects and the fact that they had ingested water in a larger proportion than had the controls, and that the analyses of the swimming water samples indicated fecal contamination (>50/mL) are elements associated with contamination. The concentration of the enterococci was lower in the deeper parts of the lake^(1,5). In the outbreak reported here, in two cases the subjects had swallowed water, the average age was 6.25 years and the water analyses showed fecal coliforms. The location where bathing was possible was shallow.

The last analysis done at the public beach indicated an elevated quantity of coliforms and we can presume that, if the analysis had been done during the days when there were many bathers, the number of coliforms might have been higher and have led to closure of the beach. The water was probably contaminated by children who had fecal accidents or individuals (children or adults) who were asymptomatic shedders. The fact that three strains of E. coli O157:H7 were found might suggest that a number of individuals contaminated the water. Since E. coli O157:H7 can survive several weeks in water, this factor could also explain the three strains found in the outbreak; it is possible that the contamination occurred over a prolonged period(1). It must not be forgotten that the infectious dose for E. coli O157:H7 is low, making transmission easy. Another argument in favour of exposure to water at the public beach as the source of infection is the case-control study done to determine the frequency of exposure at the public beach in the general population. Exposure in the general population is infrequent, contrary to the situation for these E. coli cases.

It is possible that recall bias is present among the selected controls; some of them were questioned several weeks following the outbreak. However, their response may reflect swimming habits in the Montreal population, and it seems that few Montrealers go to this beach.

This public beach has existed for several years, and this is the first time that an outbreak there has been reported. The heat wave during this period of the summer probably significantly increased the number of people using this beach. A large concentration of bathers, the shallowness of the water and the presence of young children promoted contamination of the water and transmission of E. coli O157:H7.

Acknowledgements

We thank S Asselin for reviewing the document.

References

1. Keene WE, McAnulty JM, Hoesly FC et al. A swimming-associated outbreak of hemorrhagic colitis caused by Escherichia coli O157:H7 and Shigella sonnei. N Engl J Med 1994;331(9):579-84.

2. Warrner M, Kuo K, Williams L et al. Lake-associated outbreak of E. coli O157:H7 - Illinois. MMWR 1996;45(21):437-39.

3. Ackman D, Marks S, Mack P et al. Swimming-associated hemorrhagic colitis due to Escherichia coli O157:H7 infection: evidence of prolonged contamination of a fresh water lake. Epidemiol Infect 1997;119(1):1-8.

4. Cransberg K, van den Kerkhof JH, Banffer JR et al. Four cases of hemolytic uremic syndrome - source: contaminated swimming water? Clin Nephrol 1996;46(1):45-9.

5. Wang G, Doyle MP. Survival of enterohemorrhagic Escherichia coli O157:H7 in water. J Food Protect 1998;61(6):662-67.

Source: A Bruneau, MD, MSc, H Rodrigue, BSc, Infectious Diseases Unit, Montreal-Centre Public Health Department; J Ismäel, BSc, R Dion, MD, Quebec Public Health Laboratory and the Quebec National Institute of Public Health; R Allard, MD, FRCPC, Infectious Diseases Unit, Montreal-Centre Public Health Department and the Quebec National Institute of Public Health