It Must Be the Water
Indiana Epidemiology Newsletter
Jennifer Wyatt, MPH
District 4 Field Epidemiologist
On February 17, 2006, a representative of a local health department (LHD) notified the Indiana State Department of Health (ISDH) that the infection control nurse of the local hospital had reported three confirmed cases of campylobacteriosis. The three confirmed cases were residents of a small community of approximately 600-800 persons. That same day, the LHD received three telephone calls from citizens of the same town indicating they were experiencing gastrointestinal symptoms.
The ISDH and the LHD initiated a collaborative investigation of this outbreak. A case-control study was conducted in order to determine whether the source may have been food- or water-related. The ISDH developed a questionnaire that documented illness and exposure history and forwarded it to the LHD. The ISDH and the LHD conducted telephone interviews with residents who reported illness and unmatched controls identified through interviews and random digit dialing. Completed questionnaires were forwarded to the ISDH Epidemiology Resource Center for analysis. A case was defined as any previously healthy person who resided in the town and became ill on or after February 4, 2006, with diarrhea, abdominal cramps, nausea, fatigue, fever, and/or vomiting. Any person who did not become ill was eligible to be included as a control. Any person who was ill with symptoms that did not include diarrhea, abdominal cramps, nausea, fatigue, fever, and/or vomiting was excluded from the study.
Residents who called the LHD on February 17, 2006, indicated that the municipal tap water was discolored and had an unusual odor. The LHD contacted the water operator of the town to ask if the water system was experiencing any problems. The water operator indicated there had been recent problems with the water system in the town. The LHD issued a boil-order advisory effective February 17, 2006, recommending that all residents using municipal water boil their water before use. The ISDH and LHD continued surveillance for additional cases of gastrointestinal symptoms.
Thirty-four residents reported illness, and 32 met the case definition. Thirty-seven controls were identified. Symptoms reported by the 32 cases included: diarrhea (50.7%), abdominal cramps (38.0%), fatigue (38.0%), nausea (33.8%), fever (26.8%), and vomiting (18.3%). Other symptoms included headache, joint pain, and loss of appetite. Illness onset dates ranged from February 4 until February 21 (Figure 1). Based on the issues regarding the water system, these dates represent the time period when a majority of the residents became ill. The mean duration of illness was 6.3 days (range: 1 day to 14 days). Nine people submitted stool samples for laboratory diagnosis (see Laboratory Results). At least 19 people did seek medical attention, and 3 were hospitalized. No statistical comparisons of different water sources (i.e., bottled water, well water, or alternative water source) could be analyzed, because the majority of those interviewed received and consumed water from the municipal supply.
Information about possible sources of the outbreak was collected during the investigation. Exact logistic regression (SAS 9.1) was used to evaluate the association between illness and specific risk factors. The following factors were analyzed: drinking municipal tap water, contact with pets, contact with ill people, and eating at restaurants. Table 1 shows the odds ratios, confidence intervals, and p-values for each event. Drinking tap water was found to have a statistically significant association to illness (odds ratio = 18.30, p-value = .002)
|Drinking municipal tap water||18.292 (2.590 ->100)||0.0018|
|Cantact with pets||1.849 (0.275 - 13.17)||0.7095|
|Contact with ill people||0.564 (0.103 -2.819)||0.6446|
|Eating at restaurants||0.184 (0.004 - 1.689)||0.1974|
A representative of the LHD contacted the water operator of the town on February 17, 2006, to inquire about any problems with the water system. The water supply to the town comes from a groundwater source. The water operator informed the LHD and Indiana Department of Environmental Management (IDEM) that the chlorine feed possibly did not work properly several weeks prior to the outbreak. During this time, the water supply did not meet the recommended chlorination concentration level of 1.00 ppm. The representative from the LHD issued a boil-water advisory, and the ISDH informed IDEM of the situation.
On February 20, 2006, an inspector from IDEM visited the water treatment facility to investigate the water system. Prior to the outbreak, a new water main was installed without a valid permit. The water main was pressure tested and was left under pressure with nonpotable water, resulting in a cross-contamination hazard. Prior to the outbreak, the chlorine feed did not function properly; therefore, the water supply was not receiving the proper amount of chlorination. This equipment malfunction caused the chlorine levels in the water to fall below the recommended concentration levels of 1.00 ppm. When routine samples of the water were collected on February 16, the samples tested positive for total coliform. On February 17, water samples tested positive for total coliform and E. coli, an indication of fecal contamination. IDEM advised that the boil-water advisory could be lifted on March 3, after two sets of two absent total coliform water samples were collected and tested. IDEM further required the water operator to submit weekly coliform samples for five weeks and to maintain total chlorine levels of at least 1.00 ppm in the distribution system.
Nine cases submitted stool specimens for analysis. Seven specimens were positive for Campylobacter species. All tested negative for E. coli O157:H7, Salmonella, and Shigella.
This investigation confirms that an outbreak of campylobacteriosis occurred among residents between February 4-21, 2006. The only consistent and statistically significant exposure among the cases was consuming municipal tap water. The causative agent was Campylobacter. A confirmed case was defined as a resident who tested positive for Campylobacter and experienced gastrointestinal illness on or after February 4, 2006. A probable case was defined as any person who experienced all symptoms but had no laboratory confirmation. A suspect case was defined as any person who exhibited some symptoms without laboratory confirmation.
Campylobacteriosis is an infectious disease caused by bacteria of the genus Campylobacter1. Campylobacter is one of the most common bacterial causes of diarrheal illness in the United States1. Symptoms include diarrhea (frequently with bloody stools), abdominal pain, malaise, fever, nausea and/or vomiting2. The incubation period is usually 2-5 days, but can range from 1-10 days depending on the dose ingested2. Individuals can shed the bacteria throughout the course of infection, lasting several days to several weeks2. Transmission occurs by ingestion of undercooked or contaminated food, including unpasteurized milk and untreated water, by direct contact with fecal material from infected animals (particularly puppies and cats) or contact with infected individuals3. Antibiotics can shorten the duration of illness if taken early in the illness3. Although rare, some long-term consequences such as reactive arthritis and Guillain-Barré syndrome can result from Campylobacter infection1.
According to the epidemiologic and environmental investigations, illness most likely resulted from consuming contaminated municipal water shortly before February 4, 2006 through February 11, 2006. Positive total coliform tests indicate the presence of various types of potential disease-causing microorganisms in the water from environmental sources. Some microorganisms, including Campylobacter, can cause acute infection. If water samples test positive for total coliform, the water is then tested for E. coli to determine if fecal contamination is present. Water samples collected from the municipal water system tested positive for both total coliform and E. coli, indicating contamination of the municipal water system.
Several factors may have led to the contamination of the water supply to the residents of the town. It is possible that installation of the new water main could have led to a drop in pressure and a cross-contamination hazard. Muddy conditions existed when the pipes were installed. It is essential to sanitize and super-chlorinate when work on water mains is conducted. Disinfection of water ensures that dangerous microbial contaminants are killed. When the chlorine feed in the town did not function prior to the outbreak, the chlorine levels did not meet recommended concentration levels. Thus, contaminants such as E. coli and Campylobacter could have infiltrated the water distribution system and been ingested by residents. Once the boil-water advisory was issued, the number of cases greatly decreased, and the outbreak ended.
This study included some limitations. First, exposure bias by the town’s residents that the water was the source of their infection may have occurred due to significant local media attention. Second, recall bias also may have occurred, because some individuals who were interviewed could not remember various exposures, including foods eaten. Another limitation of this study was missing data, including possible exposures and symptoms, were not collected on initial interviews. However, statistical associations between exposures and illness were not affected significantly.
The LHD and the ISDH issued several news releases to the media and surrounding LHDs describing the outbreak. The collaborative effort between the LHD, ISDH, and IDEM ensured the rapid identification and prevention of additional cases of illness.
- Centers for Disease Control and Prevention. Campylobacter infections. [online
posting] 2005 October 6.
www.cdc.gov/ncidod/dbmd/diseaseinfo/campylobacter_g.htm. Accessed 2006 June 20.
- Heymann, DL, ed. Control of Communicable Diseases Manual. 18th ed. Baltimore:
United Book Pr; 2004. 81-4.
- Pickering, LK, ed. Red book. 26th ed. Elk Grove Village: Am Academy of Ped; 2003.