A gastroenteritis illness outbreak associated with swimming in campground lake.

On Monday, June 14, 1993, the Oakland County Health Division (OCHD) was notifie that 13 of 23 children had become ill after attending a camping outing the previous weekend (June 11-13, 1993). None of the adults accompanying the children were reported to have become ill. The predominant symptoms experienced included vomiting, diarrhea, nausea, and abdominal cramps, with the majority of symptoms continuing at the time of reporting. Between June 4-25, eight groups were identified as camping near Lake #5, a 10-acre, man-made lake. Six of the groups camped from June 11-13 and were the focus of this investigation. Telephone interviews of five groups were conducted to obtain detailed exposure and illness histories, resulting in 111 completed interviews. A preliminary evaluation of illness reports revealed that four groups swimming in Lake #5 reported illnesses. The remaining group did not swim at Lake #5, nor did they report any illnesses. The beach at Lake #5 was closed the morning of June 15 pending water testing and an investigation.

METHODS

Epidemiologic Investigation

Data collection

Questionnaires were administered by telephone to five groups, comprised of approximately 140 individuals, attending the camp from June 11-13. The questionnaire assessed camp activities, symptoms experienced, and the onset and duration of illness. Attempts were made to interview individuals personally, with parents providing information on children.

Case definition

A "case" was defined as an individual attending the camp between June 11 and June 13 who experienced vomiting and/or diarrhea or two of the following: nausea, abdominal cramps, or headaches.

A "primary case" was defined as an individual with an onset of illness within 7 hours of arriving at camp.

A "subsequent case" was defined as an individual with an onset of illness greater than 72 hours after arriving at camp who had contact with a primary case.

Data analysis

Data were entered and analyzed using Epi Info version 5.01. Frequency distributions and means were computed for selected variables. Bivariate analyse were conducted using chi-square tests at p = 0.05 to assess associations betwee exposure and illness.

Environmental Investigation

Water samples were collected at Lake #5 on June 14, 16, and 21. In addition, weekly sampling continued as part of the Oakland County Health Division's (OCHD swimming beach program. The 626-acre campground contains 1,100 campsites available for day use and seasonal camping. The campground has six man-made lakes ranging in size from four to 33 acres. Lake #5 is the smallest lake used for swimming and is fed by two water wells. The well pumping rate and frequency is adjusted based on lake level, air temperature, and the expected number of people swimming at the lake.

Water samples were collected from seven drinking water wells in the Lake #5 camping area. The campground has approximately 150 hand pump wells, spread throughout the campground, as well as mechanical wells servicing comfort stations (restrooms). The comfort stations are also serviced by a municipal sewer and were inspected for indications of sewage leakage as were a number of vault privies. An inspection of other campground facilities (e.g., pool, food service area) was also conducted to identify potential sources of lake or well water contamination.

Laboratory Investigation

Stool specimens were requested of individuals experiencing symptoms at the time of the interviews to be cultured by the OCHD laboratory for Salmonella and Shigella species.

Well water samples were cultured for total coliform bacteria, and beach water samples were cultured for fecal coliform bacteria by the OCHD laboratory.

RESULTS

Epidemiologic Investigation

One-hundred eleven telephone interviews were completed between June 16-23. Thos interviewed were 67% female and ranged in age from 2 to 48 years (mean: 19.0 years, median: 11.5 years). There were 70 children (15 years of age and under) and 41 adults.

Illness histories

Fifty-three individuals (48%) met the case definition, 94% of whom experienced diarrhea and/or vomiting. Cases included 39 children and 14 adults. The mean ag was 16.5 years (median: 10.0 years). Dates of onset of symptoms ranged from Jun 12-20. The incubation periods for primary cases (n = 37) ranged from 11 to 68.5 hours (mean: 37.3 hours). The date and time of the first report of swimming while at camp was used as the exposure time to calculate incubation periods. Th duration of illness for all cases ranged from two to 95 hours (mean: 30 hours).

Forty-one cases (77%) experienced cramps followed by 37 (70%) with vomiting, 36 (68%) with nausea, 30 (57%) with diarrhea, and 17 (32%) with headaches. Twenty-eight cases (53%) experienced at least three of these symptoms, 50% who experienced four of the symptoms.

The frequency of symptoms analyzed by age group revealed significantly more children experiencing vomiting compared to adults (79% vs. 43%, chi-square, p [is less than] 0.05). These percentages were virtually opposite for diarrhea which 79% of adults experienced and 49% of children experienced (p [is less than] 0.10). Other symptoms were not significantly different by age group.

Exposure histories

The majority of camp activities involved swimming or some type of water recreation activity. Approximately 84% of individuals interviewed reported swimming activities. Of these, 50% reported swimming in Lake #5 only. Ten percent of individuals reported swimming in both lakes and the pool, 7% swam in the pool only, while 2% reported swimming in the lakes only. Significantly more children reported swimming than adults (96% vs. 63%, chi-square, p = 0.05).

Of those who became ill on or before June 15, swimming in the lakes and the pool, or swimming in Lake #5 only was significantly associated with illness (chi-square, p [is less than] .001). Seven individuals swam in the pool only an did not report any illness. Two individuals could not be classified as ill or non-ill and were excluded from analysis. Of the 37 ill individuals who swam, 31 were children and 6 were adults. Of the 17 individuals who did not swim, 14 wer adults and 3 were children.

Of those with an onset of illness after June 15, swimming was not associated with illness (chi-square, p [is greater than] 0.05). The three ill individuals not swimming were adults. In children, immersing the head or ingesting water wa significantly associated with illness (chi-square, p [is less than] .05).

When questioned about sources of drinking water, approximately 7% of individual reported their sole source of water as bottled water or water brought from home 50% of whom reported using well water for preparing cold drinks, coffee, or food, and 88% of whom reported using well water for washing dishes and brushing teeth. Each group brought its own supply of food for preparation at the camp. Only 16% of campers reported purchasing cold drinks or food from the concession stand during their stay.

Environmental Investigation

The lakes were reported to be in heavy use during the time period in question a the weather was hotter than usual and the campsites in the immediate area were the only ones open for camping. The prevailing winds and current flows in this lake are into the beach area. This combined with the limited beach area and a shallow depth is likely to keep any contamination in the beach area.

Samples collected from Lake #5 on June 14, 16, and 21 were all within acceptabl limits (i.e., less than 200 fecal coliforms per 100 ml). On June 14, two of the three samples had 10 fecal colonies and the third one had 50 colonies. On June 16, each of the three samples had less than 10 colonies of fecal coliforms. On June 18, finding no indication of a source of pollution in the lake and having acceptable water sample results, the beach was reopened at 3:00 p.m. On June 21 three water samples showed fecal coliform counts of 140, 150, and 170.

Lake #5 is a relatively small, 10-acre man-made lake with a limited beach area (approximately 250 feet). The lake's maximum depth is ten feet. The beach area is on the east side of the lake. The prevailing winds are from the west. The lake's primary water source is from two wells. One well is located at the south end of the lake and the second one is located in the northeast part of the lake north of the beach area. The water flow is south to north. Lake #5 outlets into Lake #6 through an underground pipe. As an adjunct to water quality at Lake #5, the campground agreed to keep the south end well pump running, to facilitate mixing of the lake twice a day via motorboat, and to monitor beach use to prevent crowding. The seven wells tested in the area of Lake #5 were satisfactory (i.e., no coliform TABULAR DATA OMITTED found). The campground indicated difficulty in the maintenance and servicing of its wells, both hand-type and mechanical wells. In addition, many of the wells are older and do not meet the current well code. Finally, inspection of municipal sewers servicing comfort stations and of vault privies revealed no indication of sewag overflow and leakage into the lake area.

Laboratory Investigation

One stool specimen submitted for culture was negative for Salmonella and Shigella. Preliminary viral studies of Lake #5 indicated the presence of poliovirus.

Beach samples met the standards for swimming (i.e., 200 fecal coliforms per 100 ml). The first two sets were less than 50 fecal coliforms per 100 ml.

Well water samples from seven wells in the Lake #5 area were satisfactory (i.e. zero total coliform count).

DISCUSSION

A number of organisms have been used as indicators of sewage pollution. Among these organisms are total coliforms, fecal coliforms, and enterococci. Each indicator organism/group has characteristics that limit its use as a sole or ideal identifier of human sewage pollution. Fecal coliform, as an indicator, is commonly used in evaluating bathing water.

The use of bacterial indicators have been used where relatively high dose exposure to a pathogen is necessary for illness to occur. Viral agents are pathogenic at relatively low dose levels. It is unclear, from this outbreak, bu the possibility exists that low levels of indicator organisms (e.g. fecal coliforms) may not adequately identify the probability of the presense of viral agents. In this outbreak, surface water samples showed acceptable water quality (based on fecal coliforms) but also that poliovirus (reservoir -- man, fecal) was present. This suggests the possibility of other viral agents being present, including the Norwalk virus.

Previous outbreaks of Norwalk-like gastroenteritis have been associated with a variety of environmental settings (1,2). A report by Baron, et al., associated recreational lake with an outbreak of viral gastroenteritis (3). In this outbreak, swimming in a small man-made lake during a three-day camping outing was epidemio-logically associated with an outbreak of gastroenteritis among multiple groups of campers. Although water samples collected from the beach are and from wells supplying the lake were satisfactory, they were not collected until after the outbreak period; therefore, the quality of the water at the tim of exposure cannot be confirmed. In addition, the lack of common food sources among groups and acceptable drinking water results suggest that these exposures are not as likely to have caused the outbreak. A survey of the lake area did no identify any probable environmental sources of contamination of the lake water. The water source for the lake is groundwater of high quality and was not a suspected source of any contamination. Finally, results of telephone interviews indicated that an incontinent child in one of the groups who went swimming reported an earlier onset of illness than the other individuals interviewed (i.e., approximately 2.5 hours after swimming). There is a possibility, therefore, that the water was fecally contaminated, and this contributed to the infection of other swimmers.

The epidemic curve for this outbreak showed a bimodal distribution suggesting the possibility of secondary transmission. The majority of groups interviewed consisted of both adult and child household members. All of the ill children went swimming while the majority of ill adults did not go swimming. Also, children were more likely than adults to report having immersed their head or ingested water. Within groups of campers, which were predominantly families, there were many instances of an onset of illness of 72 hours or less (i.e., on or before June 15) followed by a later onset. This would suggest the possibilit of transmission from a child to an adult household member. It should be noted, however, that these may not be true secondary cases as other common exposures a the camp may have occurred. Information on ill household members not attending camp was limited.

CONCLUSIONS

The reported symptoms, incubation periods, duration of illness, and potential secondary transmission involved in this outbreak suggest the involvement of a Norwalk-like agent. This is further demonstrated by the fact that the predominant symptoms in children and adults varied, specifically that children experienced vomiting significantly more than adults, who experienced significantly more diarrhea. As noted by White, et al., the symptoms are experienced by all age groups, but diarrhea is relatively more prevalent among adults, whereas a higher proportion of children experience vomiting (4). Negative stool culture results, as well as preliminary viral studies indicating the presence of poliovirus in Lake #5, further suggest the possibility of a vital etiology in this outbreak.

Recommendations made to the campground to prevent the occurrence of future outbreaks included continuous running of a well water pump at the south end of the lake, mixing of the lake water via motorboat, and monitoring the use of the beach to prevent overcrowding. No significant problems at the campground were reported subsequent to the investigation.

This investigation was complicated by the campgrounds varied and multiple provisions for water and sanitary facilities. The campground is currently on a program to update these facilities. This includes the abandonment of many of th hand-pump wells, upgrading of water supplies, and reduction in the number of vault privies.

References

1. Ctr. for Disease Control (1990)," Vital Agents of Gastroenteritis: Public Health Importance and Outbreak Management," MMWR 39: 1-24.

2. Blacklow, N.R., G. Cukor (1982), "Norwalk Virus: A Major Cause of Epidemic Gastroenteritis," AJPH 72: 1321-1323.

3. Baron, R.C., F.D. Murphy, H.B. Greenburg, C.E. Davis, D.J. Bregman, G.W. Gary, J.M. Hughes, L.B. Schoenberger (1982), "Norwalk Gastrointestinal Illness, Amer Journal of Epid 115: 163-172.

4. White, K.E., M.T. Osterholm, J.A. Mariotti, J.A. Korlath, D.H. Lawrence, T.L Ristinen, H.B. Greenberg (1986), "A Foodborne Outbreak of Norwalk Virus Gastroenteritis," Amer Journal of Epid 124: 120-126.

Alan Drenchen, B.S., M.A., Environmental Health Supervisor, Oakland Co. Health Division, 27725 Greenfield Rd., Southfield, MI 48076.