Characteristics of Foodborne Disease Outbreak Investigations Conducted by Foodborne Diseases Active Surveillance Network (FoodNet) Sites, 2003-2008

Background. A mean of >= 1000 foodborne disease outbreaks (FBDOs) causing >= 20 000 illnesses are reported to the Centers for Disease Control and Prevention (CDC) annually. We evaluated characteristics of successful outbreak investigations (ie, those that identified an etiologic agent or food vehicle) in the Foodborne Diseases Active Surveillance Network (FoodNet). Methods. FBDOs were defined as the occurrence of >= 2 cases of a similar illness resulting from ingestion of a common food. FBDOs reported to CDC Foodborne Disease Outbreak Surveillance System during 2003-2008 with FoodNet supplemental data available were included in the analyses. Results. Data regarding 1200 FBDOs were available. An etiologic agent was confirmed in 715 (60%); a food vehicle was identified in 387 (32%). At least 4 fecal specimens were collected in 425 of 639 outbreaks (67%) with a confirmed etiologic agent and 48 of 232 (21%) without a confirmed etiologic agent (odds ratio [OR], 7.6; 95% confidence interval [CI], 5.3-10.9). A food vehicle was identified in 314 (47%) of 671 outbreaks investigated using a case-control or cohort study, compared with only 73 (14%) of 529 outbreaks investigated by using other methods (OR, 5.5; 95% CI, 4.1-7.3). At least 1 barrier affecting the success of the investigation was reported for 655 outbreaks, including too few patients (n = 172; 26%), too few stool specimens (n = 167; 25%), and too few control subjects (n = 152; 23%). Conclusions. Etiologic agent and vehicle are frequently undetermined in FBDOs. Greater emphasis on fecal specimen collection and overcoming barriers to pursuing analytic epidemiologic studies can improve ascertainment of these factors.