Volume 30-22
15 November 2004

[Table of Contents]

Escherichia coli O157:H7 outbreak at a summer hockey camp, Sudbury, 2004

Introduction

At 17:55 p.m. on 19 July 2004, a local paediatrician notified the Sudbury and District Health Unit of a case of bloody diarrhea being assessed in the local hospital's emergency department. The paediatrician also informed the health unit that the case was a camper and that she had heard of similar cases having been assessed over the preceding 24 hours.

Investigation by the health unit revealed that the outbreak began during the 12-16 July session at a rural summer hockey camp. This report summarizes the findings from the investigation, which confirmed that E. coli O157:H7 was the causative organism and that hamburger, purchased at a local retailer on 11 July and served undercooked at the camp on 16 July was the likely source of the outbreak. In total, 34 cases were investigated. The outbreak involved 27 camp-acquired cases and three cases due to secondary (family-like setting) transmission. These three secondary cases are not included in the camp outbreak analyses. Four additional non-camp-related community cases were discovered following a voluntary recall by the meat retailer.

Outbreak Cases

To determine the source and extent of the outbreak, a standardized questionnaire was administered in person, by e-mail, fax, or telephone to all campers (aged 6-13) and junior counsellors (aged 12-15) (collectively referred to as campers), to camp staff and operators (collectively referred to as staff) and to visitors. Follow-up interviews were conducted as required. The camp continued to operate until 24 July when it voluntarily ceased operation on the recommendation of the health unit. The sessions of 12-16 July (cohort 1) and 18-24 July (cohort 2) were investigated.

Campers were exclusive to either cohort 1 or 2. Most staff members worked during both weeks and were, therefore, included in both cohort 1 and 2. If a staff developed case-defining symptoms prior to or within 48 hours of the start of the second session (i.e. shorter than one minimum incubation period; before 21 July), they were only included in cohort 1 analysis. If a staff developed case-defining symptoms more than 8 days after the end of the first session (i.e. more than one maximum incubation period; after 24 July), they were included as a non-case for cohort 1 and as a case for cohort 2⁽¹⁾. If the onset was between 21 July and 24 July inclusive, staff cases were included as cases for both cohort 1 and 2 analyses.

Cases were categorized as suspect, probable or confirmed. Cases were defined as a camper or staff who attended the camp on or after 12 July and who had an onset of a specific clinical history or with positive laboratory results for E. coli O157:H7 on or after 16 July. The case definition is outlined in Table 1. All three categories of cases were included in the analysis.

All of the 123 campers, staff and visitors were contacted (56 from cohort 1 and 80 from cohort 2; 13 staff were in both cohorts). A total of 27 cases was identified of which four were suspect, 12 were probable, and 11 were confirmed. Of these cases, one was included in both cohort 1 and 2 analyses. There were 22 cases in cohort 1 and six cases in cohort 2. Overall rates for illness among campers in cohort 1 and 2 were 44% (15 of 34) and 11% (4 of 35) respectively, and among staff were 37% (7 of 19) and 15% (2 of 13). Of the 35 visitors to the camp, none became ill. The median age of cases was 12 (range 8 to 36). Onset dates ranged from 16 July through to 26 July. (Figure 1). The median duration of illness was 3.5 days (range 1 to 11). Fifteen cases (56%) had bloody diarrhea and nine (33%) were hospitalized. There was one report of hemolytic uremic syndrome (HUS) among the 27 camp-related cases.

Figure 1.
Epidemic curve for hockey camp, campers and staff, cohorts 1 and 2, Escherichia coli O157:H7 outbreak, Sudbury, 2004

Three additional cases related to secondary transmission within family-like settings were reported. Of these cases, one developed HUS.

Of the 19 camp cases for which clinical specimens were submitted, 12 had confirmation of E. coli O157:H7 (including three from cohort 2).

Site and Epidemiologic Investigation

Consumption of reportedly undercooked hamburger served at noon on 16 July was associated with substantially increased risk for illness for cohort 1 cases (attack rate: 57% [17 of 30] vs 20% [4 of 20]; relative risk: 2.83, 95% confidence interval 1.12 to 7.19). Hamburger consumption was unknown for one case and two non-cases.

The camp purchased the suspect ground beef from a local retailer on 11 July. Ground beef from the same lot was served as "sloppy joes" and beef fajitas during lunch and dinner on 14 July. Of the four cases that did not consume hamburger on 16 July, three consumed both sloppy joes and beef fajitas and one consumed only sloppy joes. For the one case with unknown hamburger consumption, both beef fajitas and sloppy joes were consumed.

Increased risk for illness was not associated with consumption of other foods or beverages at the camp. Although a significant relative risk was associated with the consumption of water from the cabin taps (RR 2.56; [1.75, 3.76]), all five cases also ate hamburgers at the 16 July noon meal. The drinking water supply was not identified as a possible source of infection. Investigations revealed no other environmental sources or explanations for E. coli O157:H7 exposure.

With an assumed exposure date of noon on 16 July, the first two cases had incubation periods of only 6 and 7 hours respectively. Both cases were laboratory confirmed. Investigation did not yield any explanation for these short incubation periods. Both cases ate the hamburger on 16 July. One case also ate the two ground beef meals on 14 July. The other case also ate sloppy joes on 14 July. The median incubation period was 2.2 days (range 6 hours to 8 days; mean 2.6 days).

Cohort 2 cases are presumed to be the result of person-to-person transmission. Upon investigation, all reported some potential contact with a symptomatic cohort 1 case.

Ground Beef and Community Cases

No ground beef specimens from the meal consumed on 16 July were available at the camp for testing. However, the source of the meat was verified as packaged and sold from a local retailer on 11 July. The retailer participated in a voluntary recall that yielded four symptomatic individuals ("community cases") from whom three repackaged ground beef samples were obtained. In addition, four more meat samples from asymptomatic individuals were also returned for testing. All seven meat samples tested positive for E. coli O157:H7.

Samples of raw ground beef purchased on 16 July and used to prepare hamburgers at the camp for consumption on 18 July remained at the camp. Samples of this beef were seized from the camp freezer on 22 July. Unopened packages of this meat tested negative. Raw hamburgers prepared at the camp tested positive for E. coli O157:H7. Although all food handlers were asymptomatic and tested negative, other symptomatic camp staff were on-site and reported some involvement in meal preparation and/or cleanup. It is assumed that symptomatic camp staff were the source of these contaminated hamburgers.

Laboratory Results

Serotyping on all E. coli O157:H7 was conducted at the Central Public Health Laboratory, Toronto and the National Microbiology Laboratory, Winnipeg, and a summary of the results is presented in Table 2.

Discussion

An outbreak of E. coli O157:H7 associated with the consumption of undercooked contaminated hamburgers occurred at a rural summer hockey camp. Of the 34 cases investigated, a total of 30 cases were directly or indirectly related to the outbreak. Investigation results support the conclusions that of the 30 cases, 21 were related to hamburger consumption, six to camp person-to-person transmission and three to secondary transmission.

None of the suspect ground beef was available at the camp for sampling. However, a voluntary community-wide recall on 11 July of suspect ground beef resulted in seven suspect beef returns, all of which were positive for E. coli O157:H7. The recall also yielded four clinical cases of E. coli O157:H7 infection of which one was laboratory confirmed. All laboratory-confirmed cases and meat samples had the same or closely related genetic typing.

Prompt action on the part of the health unit may have resulted in the prevention of further spread of E. coli O157:H7 infection both because contaminated meat was returned and because of the high profile of the outbreak in the community. The health unit ensured that all media messages included instructions on safe food handling practices.

Conclusions

The fact that the E. coli O157:H7 serotype of the ground beef samples from the community recall was identical or genetically closely related to that of the outbreak case stool samples supports a common source of infection. Specifically, the investigation supports the following conclusions:

1. Ground beef packaged and sold by one local retailer on 11 July was contaminated with E. coli O157:H7 prior to handling at the local hockey camp; and
2. Consumption of undercooked ground beef from the 11 July lot is the likely source of the camp-related outbreak and at least one community case of E. coli O157:H7 infection.

Acknowledgements

The authors thank the following for their assistance: Dean Middleton, Public Health Division; and the staff of the Central Public Health Laboratory, Ontario Ministry of Health and Long-Term Care.

Reference

1. Chin J., (ed.) Control of communicable diseases manual. 17th ed. Washington, DC: American Public Health Association, 2000.

Source: P Sutcliffe, MD, MHSc; L Picard, RN, MSc; B Fortin, BA, CPHI (C); D Malaviarachchi, BSc, MSc; J Hohenadel, MSc; and B O'Donnell, CPHI (C); Sudbury & District Health Unit, Ontario.

[Previous] [Table of Contents] [Next]