T11-06 Differences in Foodborne Outbreak Risks by Preparation Setting, 1998–2012

Wednesday, August 3, 2016: 9:45 AM
241 (America's Center - St. Louis)
Michael Batz, U.S. Food and Drug Administration, Rock Island
Michael Bazaco, U.S. Food and Drug Administration, College Park, MD
R. Michael Hoekstra, Centers for Disease Control and Prevention-NCEZID-DFWED-EDEB, Atlanta, GA
Cary Parker, U.S. Food and Drug Administration, College Park, MD
LaTonia Richardson, Centers for Disease Control and Prevention-NCEZID-DFWED-EDEB, Atlanta, GA
Joanna Zablotsky-Kufel, U.S. Department of Agriculture-FSIS, Washington, DC
Introduction: National data on foodborne outbreaks can be analyzed to attribute foodborne illness risks to specific food sources. The characteristics of these outbreaks differ, however, by where the food was prepared, such as in a home, restaurant, or school. In particular, outbreak size and implicated foods can differ by preparation setting.

Purpose:  In this Interagency Food Safety Analytics Collaboration study, we compared risk characteristics of outbreaks based on where the food was prepared (and often consumed).

Methods: Summary data on outbreaks caused by four key pathogens (SalmonellaE. coli O157:H7, Campylobacter, andListeria monocytogenes) from 1998-2012 were collected from CDC’s Foodborne Disease Outbreak Surveillance System (FDOSS). We characterized outbreaks by where food was prepared, such as homes, restaurants, churches, schools, nursing homes, and prisons. We examined differences in outbreak size, whether outbreaks occurred in a single state or multiple states implicated food sources, and other variables.

Results: Of 952 outbreaks included in our study, 27% implicated foods prepared in restaurants and 23% in private homes, with much less than 10% associated with fairs, picnics, churches, schools, nursing homes, or prisons. Over 8% had unknown preparation locations and 13% had multiple preparation locations. Food preparation location was a significant predictor of outbreak size (p<.05), and implicated food vehicles differed by preparation location. Furthermore, differences in preparation setting were observed for outbreaks occurring in a single state versus multiple states.

Significance: This exploratory analysis informed the development of a statistical model for outbreak-derived foodborne illness source attribution, in which food preparation location is one of four variables used to predict outbreak size. Understanding how risks differ by location of food preparation is important for interpreting attribution estimates and for targeting risk mitigation efforts to different audiences.