You are here

Characterizing the Potential Use of Consumer Complaints Surveillance for the Early Detection of Foodborne Illness Events

Award No.: 
Principal Investigator: 
Craig Hedberg
PI Organization: 
University of Minnesota-Twin Cities
Rapid detection of foodborne illness events remains a challenging activity that has important implications for response and recovery efforts. The recent large scale outbreaks involving a myriad of food items such as peanuts, peanut butter, tomatoes, and peppers, illustrate of the challenges of outbreak detection. The cornerstone of foodborne illness surveillance has been pathogen-specific surveillance conducted by local and state health departments. However, this type of surveillance has some serious limitations. Only a small percentage of those suffering from a foodborne illness will ever seek medical attention. Additionally the lag time between onset of illness and report to a food department can be rather long. The CDC estimated that in the recent Salmonella outbreak in 2008 caused by peppers, the time to report to a health department was 2 to 3 weeks. This often is too long to effectively intervene on a contaminated product. A recent New England Journal of Medicine report suggested that surveillance of Google searches for illness-related keywords could be a rapid method of detecting the occurrence of an outbreak. Unfortunately, signals generated by Google searches lack specificity for both illness and exposure- the two critical elements for defining the event and mitigating its effects. Direct consumer complaints offer an alternative approach to rapidly linking specific illness and exposure information. States such as Minnesota have a robust complaints systems that facilitate outbreak detection, regardless of agent or source. Over the years 2000-2006, 77% of all confirmed foodborne outbreaks were identified through the complaints surveillance system. Importantly, the median lag time between onset of disease and report to the health department in the consumer complaints system has a median lag time of just 2 days. Use of complaints systems in the United States is not well characterized. Each jurisdiction maintains control of their public health surveillance and there is no standardization across the nation. The aim of this proposal is to characterize the use and barriers to complaints surveillance in the United States. Results will improve event modeling efforts, and foster development of a more uniform complaints surveillance system in the United States that will greatly help in detecting more cases of illness in a more timely manner. This would allow faster mitigation of contamination and public health impact of the event.
Analytic Methods Utilized: