IHP 515 Module 8 Outbreak Investigation Paper Example

Sixty-Four Cases and a Food Stand: Investigating a Festival Salmonella Outbreak

[Student Name]

Southern New Hampshire University

IHP 515: Population-Based Epidemiology

Module Eight Paper

[Instructor Name]

[Date]

Sixty-Four Cases and a Food Stand: Investigating a Festival Salmonella Outbreak

On a Tuesday in May, the Harlan County health department received reports from two emergency departments of an unusual number of patients with diarrhea, fever and abdominal cramps, several mentioning the county's spring festival the previous weekend. By Wednesday, the state laboratory had confirmed Salmonella in stool samples from four patients. This paper describes how the county investigated the cluster, following established steps, and what the investigation found.

The Steps of an Investigation

Reingold (1998) described outbreak investigation as a sequence of steps that are often carried out at the same time rather than strictly in order. Investigators establish that an outbreak exists, confirm the diagnosis, create a case definition, find and count cases, describe them by time, place and person, develop hypotheses about the source, test the hypotheses with analytic studies, carry out control measures and communicate findings. He emphasized that control measures should not wait for final proof when public health is at risk, and that clear communication with the public and partners is part of the job, not an afterthought.

Confirming the Outbreak and Defining Cases

The county typically sees two to four Salmonella cases a month; four confirmed cases in two days, plus many suspected cases, clearly exceeded the expected number. The team created a case definition with three levels. A confirmed case was a person with laboratory-confirmed Salmonella of the outbreak strain who attended the festival. A probable case was a festival attendee with diarrhea and fever beginning within five days after attending, without laboratory testing. Excluded were people whose symptoms began before attending.

Finding Cases

Investigators asked hospitals and urgent care centers to report patients meeting the definition, posted a notice and online survey through the festival's social media and vendor lists and contacted people identified by early cases as companions. Within a week, they identified 64 cases: 19 confirmed and 45 probable. Six people were hospitalized, and none died.

Stool samples were requested from probable cases willing to provide them, which moved several into the confirmed category.

Time, Place and Person

The epidemic curve, plotting cases by date of symptom onset, showed a sharp rise beginning Sunday evening, a peak on Monday and a decline by Wednesday, the classic shape of a point-source outbreak in which people were exposed at roughly the same time. The median time from festival attendance to illness was about 30 hours, consistent with Salmonella's usual incubation period. Most cases attended on Saturday rather than Friday or Sunday. Cases ranged in age from 3 to 78 years, and illness occurred in families and groups who ate together.

Forming Hypotheses

Interviews with the first twenty cases about foods eaten at the festival pointed to several vendors, with chicken tamales from one stand mentioned most often, followed by a lemonade stand and a pulled pork vendor. Because most cases attended on Saturday, investigators hypothesized that a food served that day was contaminated.

Testing Hypotheses with a Case-Control Study

Schulz and Grimes (2002) explain that case-control studies compare exposures in people with and without an outcome and are especially useful in outbreaks, where cases are identified first. Investigators interviewed 58 of the 64 cases and 116 controls, festival attendees on Saturday who did not become ill, recruited through the online survey and companions of cases, using a standard questionnaire listing every food vendor. Among cases, 46 of 58 reported eating chicken tamales; among controls, 22 of 116 did. The odds of exposure among cases were 46 to 12, or 3.83; among controls, 22 to 94, or 0.23. The odds ratio is 3.83 divided by 0.23, about 16.4. Lemonade and pulled pork showed odds ratios near 1.

Table 1. Chicken Tamale Exposure Among Cases and Controls

Note. Odds ratio = (46 x 94) / (12 x 22) = 16.4. Figures are illustrative.

Considering Bias

Grimes and Schulz (2002) describe how selection bias and information bias can distort associations in observational studies. In this investigation, recall bias was possible, since ill people might remember or report foods differently from well people, especially after news reports mentioned tamales. Selection bias could arise if controls recruited online differed from typical attendees. Investigators reduced these risks by interviewing most cases before the vendor was named publicly, using the same questionnaire for everyone and recruiting some controls as companions of cases. The very strong association, the dose-response pattern in which people who ate two tamales were more often ill and laboratory findings reduced concern that bias explained the result.

Environmental and Laboratory Findings

Environmental health inspectors visited the tamale vendor, who had cooked chicken on Friday, held it overnight in coolers without adequate ice and assembled tamales Saturday morning. Leftover chicken tested positive for the same Salmonella strain found in patients, confirming the source.

Control Measures and Communication

Following Reingold's advice not to wait for final proof, the health department had already asked the vendor to stop selling on Wednesday once early interviews pointed to tamales. After confirmation, it issued a public notice describing symptoms and when to seek care, worked with the vendor on safe cooling and holding practices and required food safety training and a temporary event permit review before future festivals. It also shared findings with the state and with other counties where the vendor operated.

Lessons

The investigation showed the value of prompt reporting by emergency departments, a clear case definition and a quick case-control study. It also revealed a gap: temporary food vendors at festivals received less inspection than restaurants. The county will add pre-event inspections for vendors serving high-risk foods.

The team also found that the online survey recruited controls quickly but skewed toward younger attendees; in future outbreaks, investigators will supplement it with phone recruitment from vendor sales records and ticket lists so that controls better reflect everyone who attended.

Conclusion

Following established steps, the county confirmed the outbreak, defined and found cases, described them by time, place and person, identified chicken tamales through a case-control study with an odds ratio of about 16, confirmed the source with laboratory evidence and acted quickly to protect the public.

References

Grimes, D. A., & Schulz, K. F. (2002). Bias and causal associations in observational research. The Lancet, 359(9302), 248-252. https://doi.org/10.1016/S0140-6736(02)07451-2

Reingold, A. L. (1998). Outbreak investigations: A perspective. Emerging Infectious Diseases, 4(1), 21-27. https://doi.org/10.3201/eid0401.980104

Schulz, K. F., & Grimes, D. A. (2002). Case-control studies: Research in reverse. The Lancet, 359(9304), 431-434. https://doi.org/10.1016/S0140-6736(02)07605-5