Moving Away from Population-Based Case-Control Studies during Outbreak Investigations

Abstract

Population-based case-control studies are a time- and labor-intensive component of foodborne outbreak investigations. One alternative is a binomial trial that asks the question "if the likelihood of each case's having eaten a given food is no different from that of the average person in the population, how often would we find, by chance alone, that x of n (or more) cases would have eaten this food?" Calculating a binomial trial requires background exposure data. We conducted case-control studies and binomial trials in two foodborne outbreaks and compared results. In both outbreaks, using binomial trials we found much less than a 5% probability that the number of cases eating the suspected food vehicle would have occurred by chance. These results were comparable with results of the case-control studies, but with considerably less effort. When background exposure data are available, binomial trials are an efficient way to explore hypotheses that can be further tested by traceback efforts to identify a common source.