Understanding which foods are most commonly responsible for certain foodborne illnesses helps in developing effective prevention measures. Today, the three federal agencies involved in the Interagency Food Safety Analytics Collaboration (IFSAC) released details of their new method for estimating the sources of Salmonella, E. coli O157, Listeria monocytogenes, and Campylobacter. IFSAC is a partnership between the Department of Agriculture’s Food Safety and Inspection Service, the Food and Drug Administration and the Centers for Disease Control and Prevention. It was formed in 2011 in order to improve foodborne illness source attribution and provide estimates for the four high priority pathogens. “The different agencies have looked at different ways of answering this question about how many illnesses are we seeing from different food commodities, but this collaboration lets us share resources, expertise and the data to produce a coordinated approach to making the U.S. food supply even safer,” said Chris Braden, director of CDC’s Division of Foodborne, Waterborne and Environmental Diseases. For the new model, the team analyzed outbreaks that occurred between 1998 and 2012, excluding those that involved multiple pathogens, those for which no food vehicle was identified, and those attributed to food containing multiple ingredients. The new model using data from the resulting 952 outbreaks differs from previous methods by using a categorization of foods updated to align with the regulatory framework of FDA and FSIS, emphasizing more recent outbreaks by giving less weight to data from 1998 through 2007, and decreasing the bias that potentially results from very large outbreaks. Eighty-two percent of E. coli O157 illnesses were attributed to beef and vegetable row crops such as leafy greens, and 81 percent of Listeria illnesses were attributed to fruit and dairy. Nearly 75 percent of Campylobacter illnesses were attributed to dairy and chicken. Most dairy outbreaks used in the analysis were related to raw milk or cheese produced from raw milk. Salmonella was distributed across a range of foods, with 77 percent linked to seeded vegetables, eggs, fruits, chicken, beef, sprouts and pork. “This suggests interventions designed to reduce foodborne salmonellosis need to include a variety of approaches,” IFSAC’s report states. The updated estimates also included “credibility intervals,” which are a way of showing how sure the team is that the percentages are accurate. For example, there were only 24 outbreaks of Listeria in the 15 years of data, which led to less confidence in the pathogen’s estimates. “For [Listeria monocytogenes], the limited number of outbreaks and wide credibility intervals dictate caution in interpreting the attribution percentages for fruit and dairy,” the report notes. “Nonetheless, Lm outbreaks have been frequently linked to the Dairy category, specifically with the consumption of soft cheeses by pregnant women and persons with weakened immune systems. Although the wide credibility interval for the Fruit category substantially limits interpretation, the analysis does suggest vigilance in seeking unrecognized sources of outbreaks and illnesses in this food category.” Braden emphasized that the data relate to overall population risk. The information is “an aggregate over the population and an aggregate of the foods in these categories and shouldn’t really be used to determine individual person or individual food risk,” he said. Members of IFSAC said that the new estimates may influence the regulatory decisions that FSIS and FDA have to make, help determine where resources are best used, and show where progress has been made in prevention efforts. Chris Alvares, director of FSIS’ Data Analysis and Integration Staff, said that one of the findings the team found particularly interesting concerned Salmonella in pork products. The agency’s recently initiated sampling program to explore Salmonella in pork was “[partially] informed by some of the attribution findings in this kind of analysis.” These data don’t examine trends, but Goldman said the collaboration will be interested to see if they can use updated estimates to measure progress over time. IFSAC members have not yet decided how often they’ll update the estimates, and one of the team’s next projects will be “to look at the appropriate frequency,” Goldman said. This is one aspect of the work on which IFSAC will be soliciting comments. Some limitations on the data and analysis outlined in the report include a lack of representation of foods responsible for sporadic disease and an unintentional emphasis on some foods such as unpasteurized milk that are not regularly consumed by the general population. In addition, the team only analyzed 36 percent of foodborne outbreaks reported over the 15-year timeframe, and 10 percent of the illnesses occurred with institutions such as prisons, hospitals and schools.