 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
| |
|
||||||||||||||||
| |
|
||||||||||||||||
 |
||
 |
||
|
Volume 27-22
|
WATERBORNE CRYPTOSPORIDIOSIS OUTBREAK, NORTH BATTLEFORD, SASKATCHEWAN, SPRING 2001
Introduction On 25 April, 2001 Saskatchewan Health requested assistance from Health Canada to investigate an outbreak of gastroenteritis in the Battlefords area of Saskatchewan. Cryptosporidium parvum was suspected of being the causative organism in the outbreak after several laboratory-confirmed cases were identified. The same day, a precautionary drinking water advisory was issued for the city of North Battleford and the town of Battleford. The epidemiologic investigation included: a descriptive study, a review of the sale of over-the-counter antidiarrheal medications, a randomized cross-sectional community study, and an overview of the municipal water treatment system in the city of North Battleford. The Battlefords health service area (BHSA) in northwestern Saskatchewan is comprised of the Battlefords, Twin Rivers, Northwest, and Lloydminster health districts. The city of North Battleford and the town of Battleford are located in the Battlefords Health District. The city of North Battleford is a community of approximately 14,000 persons and is situated on the northern side of the North Saskatchewan River, just across the river from the town of Battleford, a community of approximately 4,000 residents. At the time of the outbreak, the city of North Battleford had two principal sources of municipal drinking water: ground water supplied by a number of wells located along the North Saskatchewan River, and surface water drawn from the North Saskatchewan River, a few kilometres downstream of the well sites. The well water was treated with chlorine and underwent filtration to remove metals at the ground water treatment plant, while the surface water was subject to full conventional treatment (flocculation, sedimentation, sand filtration, and chlorination) at the surface water treatment plant. Both water treatment plants fed a common distribution system, and at the time of the outbreak, both were in operation. The town of Battleford had its own, separate source of municipal water supplied by wells and treated by a treatment plant using chlorination and filtration to remove metals. Methods Descriptive study The purpose of the descriptive study was to better define the scope, geographic distribution, and time frame of the outbreak. Data were collected using two line listing forms: one for cases within the BHSA, and the second for cases outside of the BHSA. For the purposes of the descriptive study an epidemiologically linked case was ultimately defined as a person reporting onset of diarrhea (unquantified) commencing after 20 March, 2001 and who was either a resident, or had visited, the Battlefords area. A confirmed-case was a person meeting the criteria for an epidemiologically linked case, but who also had a stool specimen positive for C. parvum oocysts. The data for individuals meeting either of the case definitions were verified to remove duplicates prior to analysis. Geographic information maps were created to describe the geographic distribution of cases based on the reported residential address. Antidiarrheal drug sales review The primary objective of this study was to help characterize the time frame of the outbreak. A convenience sample of three pharmacies in the city of North Battleford was selected and each was asked to provide unit-sales information for common over-the counter antidiarrheal medications. Each pharmacist determined the number, brands, and types of 'common' antidiarrheal medications to include in the review. Weekly sales data were requested for the January through May, 2001 period and for the January through May, 2000, period to provide comparative historical data. Cross-sectional study This study was used to determine risk factors for the outbreak; to explore whether or not a spatial trend in disease risk existed among North Battleford residents; and to estimate the number of persons with diarrheal illness in the city of North Battleford and the town of Battleford attributed to the outbreak. The cross-sectional study consisted of a telephone survey of a random sample of households within the Battlefords area. The survey was administered to the adult in each household who could best answer questions on behalf of the entire household. The household portion of the questionnaire included: household location, source of tap water, use of home water treatment systems, employment in risk settings for cryptosporidiosis (e.g., daycares, long-term care facilities), and exposure to potential sources of Cryptosporidium oocysts (e.g., municipal water, unpasteurized milk, contact with livestock, swimming, pet ownership). Information collected on individuals within each household included: age, sex, symptoms of gastroenteritis (presence, onset, duration, specific symptoms), use of healthcare services, stool sample testing and results, location of work/school, swimming in the local public pool, travel outside Canada and the United States, camping/fishing, and attendance at special events in the Battlefords. For the purposes of this study, a primary case was ultimately defined as a person with reported onset of diarrhea (>= three loose stools in a 24-hour period) between 21 March and the time of survey administration (4 to 9 May, 2001) and from a household which did not experience any diarrhea from 14 February up to and including 20 March. The latter stipulation conservatively precluded symptomatic individuals who may have been infected with a non-outbreak gastrointestinal pathogen acquired through an ill family member. A secondary case was defined as a person who met the criteria of a primary case and whose onset of symptoms was > 7 days (the average incubation period for cryptosporidiosis) after the initial primary case of diarrheal illness within the same household. A confirmed case was a person who qualified as a primary or secondary case and who had a stool specimen positive for C. parvum oocysts. A control was defined as a person from any household who reported no symptoms of gastroenteritis. Univariate and multivariate analyses of risk factors were performed on individual level variables comparing primary cases and controls using SAS analysis software (Version 8, SAS Institute). Spatial regression software (Version 2000, S-Plus) was used to explore whether or not there was a spatial trend in disease risk among North Battleford residents. This analysis involved primary cases and controls within the city of North Battleford and utilized a generalized additive model using the binomial distribution and logit link option(1). The estimates of the number of residents of the city of North Battleford and the town of Battleford who experienced diarrheal illness due to this outbreak were derived from the cross-sectional study. Previously-published estimates of background rates of gastroenteritis in Canada, the United States, and the United Kingdom(2-4), as well as an estimate derived from the cross-sectional study data were used to estimate the background rate of gastroenteritis in the Battlefords area. Using these calculated background rates, age-specific attack rates were calculated for each community. These age-specific rates were then applied to the estimated age-specific populations for each community (Statistics Canada. 1996 Census), to generate age-specific estimates of cases of diarrheal illness. The age-specific estimates were then totalled for each community to derive estimates of diarrheal illness for the city of North Battleford and the town of Battleford. Environmental study On-site visits were made to the surface water treatment plant and the ground water treatment plant in the city of North Battleford, and the ground water treatment plant in the Town of Battleford. Information obtained from officials of the city of North Battleford and Saskatchewan Environment and Resource Management included: details of treatment plant operations, a schematic drawing of the water distribution system, daily water volumes contributed by each plant, daily operational parameters associated with the solids contact unit (SCU), data on bacterial water quality, chlorine residuals from the water distribution system, and hourly values for water turbidity of finished water from the surface water treatment plant. Results The descriptive study identified 1,907 persons who met either of the case definitions. Of the 1,907 cases identified, 1,039 persons lived within the BHSA of Saskatchewan, the majority lived either in the city of North Battleford (639) or the town of Battleford (136). Most of the other 868 persons (living outside the BHSA) were from elsewhere in Saskatchewan (701), but there were also persons living in Alberta (141), Manitoba (19), and British Columbia (7). Of the 1,039 cases from within the BHSA, 55% were female and the median age of cases was 27 years of age (range: < 1 to 90 years). The peak date of onset of diarrhea was 13 April, 2001. Of the 868 cases from outside the BHSA, 53% were female, with a median age of 29 years of age (range: < 1 to 88 years). The onset of diarrhea in these cases peaked on 24 and 25 April, 2001. Of the 1,907 cases, 597 (31%) reported visiting a physician, and 50 (3%) were hospitalized. Two hundred and fifty-six cases (25%), from within the BHSA, also reported visiting an emergency department. No deaths were reported as attributable to this outbreak. One hundred and sixty (15%) of the cases from within the BHSA submitted a stool specimen, of which 110 (69%) were laboratory-confirmed for C. parvum oocysts. Two hundred and eighteen (25%) of the cases from outside the BHSA submitted a stool specimen, and 165 (76%) were laboratory-confirmed for C. parvum oocysts. The Battlefords Health District did not receive any notifications of other parasites or bacteria in stool specimens submitted from persons associated with this outbreak. In addition, no viruses associated with gastroenteritis were detected in 12 randomly selected stool specimens submitted to the Saskatchewan provincial laboratory in early May. Each of the pharmacies included in the antidiarrheal drug sales review experienced approximately a five-fold increase in sales of over-the-counter antidiarrheal medications in late March and early April, 2001 compared with sales between January and early March, 2001. For the two pharmacies reporting 2000 sales data, there was no similar increase in the sale of antidiarrheal medications for March and April, 2000. The cross-sectional study identified 196 primary and 51 secondary cases within the Battlefords area, for an overall crude attack rate (not adjusting for background gastroenteritis) of 38%. The peak onset of diarrhea in primary and secondary cases was 13 April, 2001 (Figure 1). Fifty-three percent of primary cases were female with a median age of 38 years of age (range: 1 to 90 years), which was significantly younger than the median age (44 years) of controls (p = 0.0008). Of the 196 primary cases from the cross-sectional study, 10 (5%) submitted a stool sample for laboratory testing and three were positive for C. parvum oocysts. Of the primary cases for which data were available, 22 (11%) reported visiting a doctor, two (1%) reported visiting an emergency department, and one (< 1%) reported being admitted to hospital. The multivariate analyses found that the risk of developing gastroenteritis significantly increased with increasing opportunity for exposure to North Battleford municipal water. Individuals exposed to the city of North Battleford water at home, work, or school were 1.5 times (95% confidence interval [CI]: 0.90-2.54) more likely to develop gastroenteritis than persons not exposed at all. Individuals exposed to North Battleford municipal water at home and work or school were 2.7 times (95% CI: 1.55-4.81) more likely to develop diarrhea than persons not exposed at all. It was estimated that approximately 5,800 to 7,100 persons living in the city of North Battleford and the town of Battleford developed diarrheal illness as a result of this outbreak. Spatial regression analysis of the risk of illness within the city of North Battleford found that persons residing in the southeast part of the city were two to four times more likely to develop gastroenteritis compared to those individuals living in the northwest. No abnormalities in the operational parameters of the ground water treatment plants in either the town of Battleford or the city of North Battleford were reported. However, the SCU of the surface water treatment plant in the city of North Battleford, as monitored by the 'percent settling', was found to be functioning at a sub-optimal level after maintenance on 20 March, 2001. The apparent effect of this decreased SCU efficiency was a decrease in the quality of the finished water leaving the surface water treatment plant as suggested by the increase in finished water turbidity. The reported data showed no abnormalities in the bacterial quality or chlorine residuals in the finished water in the distribution system. However, Cryptosporidium oocysts were found in the treated drinking water from the surface water treatment plant. Discussion The Battlefords area of Saskatchewan experienced an outbreak of gastroenteritis between late March and early May 2001. An estimated 5,800 to 7,100 people from the Battlefords were affected along with hundreds of visitors from other parts of Saskatchewan, Alberta, Manitoba, and British Columbia. By May 2001, C. parvum infection was confirmed in 275 people. No other pathogens were identified. No specific community event could account for the observed increase in gastroenteritis. The observations that gastroenteritis affected a high proportion of the population, was found in all age groups, and affected persons with an epidemiologic link to one community (the city of North Battleford), were characteristic of a waterborne disease outbreak. Multivariate analysis found that the only significant predictors of increased risk of diarrheal illness were exposure to city of North Battleford municipal water and age. Due to the resistance of Cryptosporidium oocysts to chemical disinfection, physical removal through filtration is the primary means to remove oocysts from water sources. The sub-optimal efficiency of the SCU, as measured by 'percent settling', after it was serviced on 20 March, 2001 could potentially have allowed oocysts to pass into the finished drinking water. The decrease in SCU efficiency was followed by an increase in finished water turbidity leaving the surface water treatment plant. The peaks in cases from both the descriptive and the cross-sectional studies, the increases in sales of antidiarrheal drugs reported by pharmacies, and the increase in finished water turbidity leaving the surface water treatment plant, temporally suggested the likely involvement of the SCU malfunction in this outbreak (Figure 2). The result of the spatial regression analysis also suggested involvement of the surface water treatment plant. The gradient in the risk of diarrheal illness increased from west to east within the city of North Battleford, and was consistent with the reported relative distribution of surface water within the city. Additionally, Cryptosporidium oocysts were found in the finished drinking water during the outbreak investigation. These findings suggest that the decreased efficiency of the SCU had a direct impact on the quality of the finished municipal drinking water from the city of North Battleford. The ultimate source of the Cryptosporidium oocysts that caused this outbreak is unknown. However, given that the outbreak was caused by surface river water, it can be assumed that the source of the oocysts originated from some point upstream. A thorough review of sewage inputs (both human and animal) into the North Saskatchewan River and water treatment systems of communities utilizing this source is warranted to minimize the risk of gastroenteritis from the consumption of contaminated drinking water.
|
|
|
Figure 1: Cross-sectional study - number of persons with diarrheal illness by date of symptom onset, Battlefords area, Saskatchewan, spring, 2001
Figure 2: Temporal comparison of disease burden and water quality data
|
||
|
References
Source: R Stirling, MD, MSc, MHSc, FRCPC, Field EpidemiologyTraining Program; J Aramini, DVM, MSc; A Ellis, DVM, MSc; G Lim, BSc; R Meyers, BSc; M Fleury, BSc, Division of Enteric, Foodborne and Waterborne Diseases; D Werker, MD, MHSc, FRCPC, former Director, Field Epidemiology Training Program, Public Health Agency of Canada, Health Canada, Ottawa, Ontario.
|
||
 |
|||
|
Last Updated: 2001-11-15 |  |
|
