Indiana Epidemiology Newsletter
Lynae Granzow, BS
On July 11, 2006, the Marion County Health Department (MCHD) notified the Indiana State Department of Health (ISDH) of an increase of salmonellosis in recent weeks. Many of these initial cases resided on the south side of Marion County, bordering the Johnson County line. The Johnson County Health Department (JCHD) confirmed a similar increase in cases of salmonellosis. The ISDH Laboratories confirmed the disease agent in these cases as Salmonella I 4,,12:i:- monophasic. Pulse-field gel electrophoresis (PFGE) testing yielded identical band patterns among the cases, indicating a common source was likely.
The ISDH, MCHD, and JCHD initiated a collaborative investigation of the outbreak on July 11, 2006. A case was defined as any previously healthy person who developed salmonellosis on or after May 19 (the onset date of the earliest known case) and exhibited the same PFGE band pattern. The ISDH Salmonellosis Case Investigation reporting form was used to interview all cases. During the investigation, a supplemental questionnaire provided by the CDC and the ISDH was added to determine other specific exposures. Six cases did not respond or provided incomplete information, thus all statistical analysis was based on 193 cases.
A total of 199 cases were confirmed in the outbreak by PFGE testing at the ISDH Laboratories. The onset of the first case occurred on May 19, with the last case onset on October 2. The outbreak peaked in August (see Figure 1). The majority of cases were identified in central Indiana counties: Marion (89), Johnson (87), and Shelby (6). Fifteen cases were identified in 12 other Indiana counties. Cases were also identified in residents of New York and New Hampshire who reported recent travel to Indiana.
The mean age of the cases was 26 years, with a range of <1-83 years (see Figure 2); 57 percent of the cases were female. Symptoms reported included: diarrhea, fever, nausea, cramps, and some vomiting. Twenty-four persons (12%) were hospitalized; none died.
Cases were interviewed about foods consumed, supermarkets frequented, and restaurants patronized prior to illness. No particular food or restaurant was commonly reported among the cases. The most commonly reported exposure (76.2%) among most cases was Supermarket A. Of those who reported purchasing food items at Supermarket A, 78.9 percent bought items from the deli. Items consumed by cases from the Supermarket A deli included: prepackaged sandwiches, deli tray items, rotisserie chickens, fried deli items, and bulk deli meats and cheeses from multiple producers.
Geographic Information System (GIS) mapping showed clustering of cases around the southern Marion County and northern Johnson County border. Supermarket A is located adjacent to Interstate 65, a major interstate highway often used for commuting into Indianapolis. Among 23 cases residing in other Indiana counties, 30 percent reported eating items purchased at Supermarket A, and 17 percent reported exposures to ill contacts who had exposure to Supermarket A prior to onset. The cases from out of state reported consuming food items purchased at Supermarket A.
The ISDH conducted a case-control study using a reverse digit-dialing system. The study included 32 cases and 32 controls. Cases included those with onset dates in July, and controls were matched by ZIP codes. The Fisher’s exact test (SAS 9.1) was used to evaluate the association between illness and exposure. The following variables were analyzed: restaurants; supermarkets; gatherings; and consumption of beef, pork, chicken, turkey, fish, eggs, and fruit. Of the five supermarkets reported, a statistical significance (α = 0.05) was found between illness and shopping at Supermarket A (OR = 21.21, p<0.0001), see Table 2. No statistical association with illness was found with any restaurants, attendance at gatherings, or general consumption of abovementioned foods prior to illness.
Supermarket A opened to the public on April 19, 2006. A final pre-opening inspection of the facility was conducted on April 13, 2006. The JCHD conducted a routine inspection on June 16, 2006, and no critical violations were found.
On July 21, 2006, the JCHD inspected Supermarket A at the ISDH’s request based on the suspicion of an outbreak of salmonellosis. The deli and produce departments were checked for possible sources of contamination, and none was found at that time. No critical violations were noted. On July 31, 2006, representatives from the JCHD and the ISDH met with supermarket management to discuss epidemiologic data indicating the establishment as a possible source of the outbreak. Supermarket management provided a list of deli and bakery associates, their corresponding hire dates, sick days, and schedules on August 4, 2006.
On August 10, 2006, a team of representatives from the JCHD and the ISDH met with supermarket employees to discuss the outbreak. All deli and bakery workers were interviewed to assess gastrointestinal symptoms, work hours, exposures to food and uncooked meat while working, risk factors for salmonellosis, and any other information they were willing to provide about their work environment. Twenty-two of 23 employees responded to the interview. While not used for analysis, these data provided a guide for the environmental assessment and further investigation.
Due to the ongoing nature of the outbreak, deli and bakery associates were requested to provide three serial stool specimens at least 72 hours apart for Salmonella testing to identify anyone asymptomatically infected. Two employees tested positive (see Laboratory Results) and were immediately restricted from all food-handling duties. The investigation revealed that one of these individuals had a pet snake. Since reptiles are known to harbor Salmonella, the snake was tested to determine if it could be a source of the outbreak. The snake tested negative for the outbreak strain of Salmonella.
During mid to late August, Supermarket A provided training in food preparation, the foodborne illness protocol, and the sick leave policy. On August 24, 2006, when the second positive employee was identified, Supermarket A voluntarily removed all bakery and deli employees from food-handling duties and thoroughly cleaned both departments. Employees returned to their previous duties once all three stool samples tested negative. Supermarket A also thoroughly cleaned and disinfected these two departments using staff and a third-party vendor on August 24, August 25, and August 30, September 6, and September 13. On August 31, Silliker Laboratories collected and tested 55 environmental samples from the deli and bakery for Supermarket A. A knife block tested positive for the outbreak strain and was removed from the deli/bakery area on September 3 (see Laboratory Results). Silliker Laboratories collected another 111 samples from September to October; all samples were negative.
Supermarket A conducted additional cleanings of the deli and bakery departments on September 6 and September 13. Due to the identification of additional case, the JCHD and the ISDH Food Protection Program (FPP) collected food and environmental samples at Supermarket A on September 14 to identify any possible residual environmental contamination. All food and environmental samples tested at the ISDH Laboratories were negative for Salmonella. On September 14-15, the ISDH FPP conducted 24-hour surveillance of the Supermarket A deli and bakery employees regarding food-handling procedures. The ISDH FPP observed potential cross-contamination sources, including soiled aprons and seasoning packets in contact with the packaged raw chickens. These are considered critical violations.
The ISDH Laboratory identified 199 confirmed cases of Salmonella I 4,,12:i:- monophasic associated with this outbreak through culture and two-enzyme PFGE matching.
All food and environmental samples collected by the JCHD and the ISDH tested negative for Salmonella. In addition, food samples provided by patrons also tested negative. A knife block tested positive for Salmonella I 4,,12:i:- monophasic according to Silliker Laboratories. This serotype result was verified by ISDH Laboratories and matched the two-enzyme (Xba I and Bln I) PFGE outbreak pattern.
On August 23, 2006, a total of 32 isolates was sent to the CDC Epidemic Investigations and Surveillance laboratory for multi-locus variable-number tandem-repeat analysis (MLVA). All isolates were indistinguishable by MLVA, and the pattern was unique to the MLVA database (N of I 4,,12:i:- database = 450).
This investigation confirmed an outbreak of salmonellosis involving residents from 15 Indiana counties and 2 other states. A total of 199 cases were identified over a four-month period that were culture positive and matched a 2-enzyme PFGE pattern. A unique MLVA pattern was confirmed in all isolates submitted to the CDC. Therefore, a common source was extremely likely.
Illness was statistically associated with consumption of foods purchased at Supermarket A. The likely mode of transmission was environmental contamination, possibly from raw rotisserie chicken preparation within the deli. A knife block used to house multi-purpose knives tested positive for the Salmonella I 4,,12:i:- outbreak strain and, once removed on September 3, the outbreak ceased shortly after. The cases following the removal of the knife block can be explained by purchase and consumption dates of the deli items. Knives were used throughout the deli and bakery. The likely mode of transmission was contaminated knives being used to open packages of ready-to-eat foods, exposing a small surface area of the package or specific portions of food to contamination. This hypothesis explains why no deli food samples tested positive for Salmonella and how only some persons who were exposed to the deli items became ill.
Salmonellosis is characterized by diarrhea, fever, abdominal cramps, nausea, headache, and sometimes vomiting. Approximately 40,000 cases of salmonellosis are reported annually in the United States, and approximately 600 persons die each year due to acute complications.3 Persons become ill within 6-72 hours after exposure, usually within 18-36 hours. Infections usually resolve within 5-7 days. Complications such as dehydration or septicemia can occur, requiring antibiotic therapy or hospitalization. Young children, the elderly, and immunocompromised persons are more likely to have severe complications.3 Salmonella bacteria are shed in the stool of infected cases for several days to several weeks.1 Rarely, persons can shed the bacteria for months or years after convalescence.
Salmonellosis is most often transmitted by the fecal-oral route. Salmonella is commonly found in the intestines of animals including poultry, cattle, swine, fish, and reptiles. Foods of animal origin, such as meats, poultry, eggs, and dairy, or externally contaminated produce, are common sources of Salmonella. Most of these infections can be avoided by properly storing, cleaning, or cooking food before consumption. Person-to-person transmission can also occur, as can infection from contaminated fomites.1
Salmonella I 4,,12:i:- monophasic is an emerging serotype in Indiana and nationwide. This serotype has been isolated from cattle, horses, poultry, domestic animals/environment, pigs, and reptiles.2 The leading hypothesis is that exposure to poultry is the most common human risk factor for Salmonella I 4,,12:i:- monophasic. In Indiana, from 2003 through 2006 (*through November), respective sequential totals for this serotype are 7, 20, 34, and 234 (including outbreak cases). The PFGE pattern was unique during the outbreak period on the national PulseNet database. The more discriminatory MLVA testing indicated that the pattern was unique among 450 isolates of the same serotype, and that there was no genetic variability among the 32 samples analyzed. These tests confirmed that cases were exposed to a common source.
This Salmonella source was isolated to Supermarket A. This was supported by the fact that no other outbreaks or cases of the same serotype or PFGE pattern were identified in other states or locations in Indiana or the U.S. unless they were linked to Supermarket A. This outbreak was not related to a particular food item within the distribution chain. A contaminated food item would have resulted in numerous cases along the distribution line from the manufacturer or producer. Also, the outbreak lasted over four months, which would make it very unlikely that a consistently contaminated food item would be the exposure.
Two asymptomatic deli employees tested positive for the outbreak strain; however, given no date of illness onset, it is unclear whether these employees were a source of the outbreak or part of it. These two employees were removed from food-handling duties immediately after their positive Salmonella results. They were allowed to return to work only after completing antibiotic therapy and at least 48 hours later having two negative successive stool samples. However, all employees were removed from food-handling duties at Supermarket A on August 24 and were not allowed to return to work until completing three successive negative stool samples. There was no change in the epidemiologic curve of cases when the employees were removed from food handling.
Some cases did not report exposure to Supermarket A but reported exposure to ill contacts, daycare settings, and gatherings involving foods of unknown origin. Several familial clusters resulted from eating deli items purchased at Supermarket A, and many of these involved secondary cases. Of note, the last three cases reported did not report exposure to Supermarket A. Recall and interviewer bias may have played a significant role in unknown exposures.
The Indiana State Department of Health extends its deepest appreciation to the Johnson and Marion County Health Departments, the Centers for Disease Control and Prevention, and all other local and state health departments involved in this outbreak. The awareness, surveillance, and long hours of investigation provided by these agencies generated data-driven evidence in the outbreak. The active cooperation and concern of the Supermarket A corporate office led to cooperative prevention measures and the cessation of the outbreak.
More information about salmonellosis and prevention measures can be found at www.isdh.in.gov/isdh.
American Public Health Association. (2004). Control of Communicable Diseases Manual 18th ed. United Book Press, Inc., Baltimore, MD.
Centers for Disease Control and Prevention. Salmonella Annual Summary, 2004. Table 7: Non-clinical salmonella isolates from nonhuman sources.
Centers for Disease Control and Prevention. Salmonellosis: General Information. http://www.cdc.gov/nczved/divisions/dfbmd/diseases/salmonellosis/