Objective:To investigate a mixed Salmonella heidelberg and Campylobacter jejuni foodborne outbreak in a nursing home.Design:Retrospective cohort study with a nested case-control design. Cases were defined by positive stool-culture results. Controls needed to be both asymptomatic and culture-negative.Setting and Patients:Residents of a 580-bed nursing home in Brooklyn, New York.Results:Of the 580 residents, 119 (21%) developed illness. Of the 93 symptomatic patients who submitted specimens, cultures were positive for S heidelberg in 24 (26%), C jejuni in 14 (15%), and both microorganisms in 25 (27%).Only the pureed diet was associated highly with infection by either Salmonella (odds ratio [OR], 17.6; 95% confidence interval [CI95], 4.8-68.7; P<.001), Campylobacter (OR, 13.3; CI95, 3.2-59.2; P<.001), or both organisms (OR, 8.9; CI95, 2.7-30.3; P<.001). Among the 52 pureed foods served during the 5 days before the outbreak, five meat or poultry items were associated most strongly with culture positivity.Of these five meat items, only a chopped-liver salad was implicated by the two employees reporting illness. A reported food-handling error occurred when ground, cooked chicken livers were placed in a bowl containing raw chicken-liver juices.Intervention:Recommendations for proper cleaning and sanitizing of kitchen equipment to prevent cross-contamination between raw and cooked foods.Conclusions:Mixed foodborne outbreaks occur rarely. During this outbreak, contamination of a single food item with multiple bacterial pathogens was the likely source of transmission. Improper food-handling techniques that promote growth of one microorganism also allow growth of other pathogens that may be present. Because different sources and routes of transmission may be implicated for different pathogens, specific preventive measures may vary depending on the organisms involved.

To investigate a mixed Salmonella heidelberg and Campylobacter jejuni foodborne outbreak in a nursing home. DESIGN: Retrospective cohort study with a nested case-control design. Cases were defined by positive stoolculture results. Controls needed to be both asymptomatic and culture-negative. SETTING AND PATIENTS: Residents of a 580-bed nursing home in Brooklyn, New York. RESULTS: Of the 580 residents, 119 (21%) developed illness. Of the 93 symptomatic patients who submitted specimens, cultures were positive for S heidelberg in 24 (26%), C jejuni in 14 (15%), and both microorganisms in 25 (27%). Only the pureed diet was associated highly with infection by either Salmonella