METHODS: We reviewed interviews of 2013–2015 case-patients with isolates matching this PFGE pattern and reinterviewed those with illness onset after June 2015 to gather details on exposures, including foods consumed from suspect Restaurant X. The restaurant was inspected and an environmental assessment was conducted by the CDPH Food and Drug Branch. Seventy-five environmental swabs were collected from food and non-food contact surfaces. All restaurant employees were interviewed regarding gastrointestinal symptoms and submitted stool for bacterial testing.
RESULTS: Between June and October, 2015, nine Salmonella Oranienburg case-patients were identified who reported exposure to Restaurant X in the week before illness onset. Four additional case-patients reported eating at Restaurant X before illness onset in April 2013, February 2014, June 2014 and August 2014. The median age of case-patients was 24 years (range 4-64 years); 62% (8/13) were male; 31% (4/13) were hospitalized and 31% (4/13) had Salmonella isolated from blood. No single food item or ingredient was associated with illness. No employees reported symptoms; all stools tested negative. Six environmental swabs tested positive for SalmonellaOranienburg matching the outbreak strain. Positive samples were collected from the dish sponge rack, the floor, floor drains, and underneath floor mats. The restaurant was cleaned and sanitized and re-opened after follow-up environmental swabs tested negative.
CONCLUSIONS: Finding the outbreak strain in the environment of Restaurant X where case-patients reported exposure identifies this restaurant as the source of this outbreak over a 3-year period. Salmonella may persist in a restaurant environment for extended periods of time contributing to human illness. Our findings emphasize the importance of routinely cleaning and sanitizing kitchen environments beyond food contact surfaces and providing education to restaurant operators on foodborne illness prevention. This investigation highlights the value of local public health surveillance for identifying PFGE-based clusters that are not identified for investigation by national PulseNet surveillance.