Evaluation of a Statewide Foodborne Illness Complaint System

Purpose: The purpose of this study was to analyze the effectiveness of using a foodborne illness complaints system to detect outbreaks, and to determine if using symptom information could help to identify particular pathogens.

Methods: Complaints data from 2001-2006 was obtained from the Minnesota State Department of Health. Complaints were linked to outbreaks to determine those that were associated with outbreaks and types of pathogens involved. Descriptive statistics were carried out comparing outbreak associated and non-outbreak associated calls. Symptom information was coded into various symptom profiles representing all combinations of the 6 individual symptoms. Profiles were examined to determine their predictive ability in identifying particular pathogens.

Results: We found that the vast majority, 75.8%, of food related outbreaks in Minnesota were detected through the complaints system. The most commonly detect pathogen was norovirus, followed by Salmonella, and Clostridium perfringens. There was also evidence that by focusing on particular symptom profiles public health officials may be better able to select out certain pathogens. We found that profiles including vomiting and diarrhea were predictive of Norovirus, and those that included cramps and diarrhea were more predictive of Salmonella.

Discussion: Complaint systems can be a very effective tool in detecting foodborne illness outbreaks especially those not targeted by pathogen specific surveillance. Further analysis of the data showed that use of symptom profiles may help in focusing complaints on certain pathogen types. Eliminating calls with vomiting in their symptom profile greatly reduces the amount of Norovirus related complaints.