Agents, vehicles, and causal inference in bacterial foodborne disease outbreaks: 82 reports (1986-1995).

OBJECTIVE: To examine the study design of, and the practice of causal inference in, investigations of bacterial foodborne disease outbreaks occurring in the United States and to summarize agents and vehicles identified.
DESIGN: Retrospective study. PROCEDURE: An online medical reference database was searched for reports of bacterial foodborne disease outbreak investigations published between 1986 and 1995. Reports were retrieved and reviewed for use of 9 causal criteria in investigations. Information on etiologic agents, vehicles, seasonality, and primary study design from each outbreak was also retrieved.
RESULTS: 82 reports were retrieved and reviewed. Coherence, consistency, temporality, and strength of association were the causal criteria most commonly used in foodborne disease outbreak investigations. Coherence was used in all investigations. The number of criteria used ranged from 3 to 7. Meat (n = 20) and eggs (12) were the most commonly implicated vehicles. Salmonella sp and Escherichia coli O157:H7 accounted for 55% of agents reportedly isolated. Cohort and case-control methods were the most common study designs. CLINICAL RELEVANCE: Patterns were found in the use of causal criteria in foodborne disease outbreak investigations. These criteria can provide veterinarians and other public health practitioners with a means to effectively conceptualize, communicate, and summarize causal conclusions. The 4 most commonly used criteria may represent core criteria that investigators consider most useful in explaining food-borne disease outbreaks.