Protracted, Intermittent Outbreak of Salmonella Mbandaka Linked to a Restaurant

Michigan, 2008-2019

William D. Nettleton, MD; Bethany Reimink, MPH; Katherine D. Arends, MPH; Douglas Potter, MBA; Justin J. Henderson, MPH; Stephen Dietrich, MS; Mary Franks, MPH


Morbidity and Mortality Weekly Report. 2021;70(33):1109-1113. 

In This Article

Environmental and Laboratory Investigation

As part of the 2018 investigation, restaurant A employee stool specimens and environmental samples were collected in parallel and analyzed for Salmonella. None of the 100 employees reported symptoms at the time of sample collection or in the weeks preceding collection. MDHHS Bureau of Laboratories identified five isolates from four of 100 asymptomatic employees' stool specimens that shared the outbreak subtype. Stool cultures were collected approximately every 30 days from asymptomatic employees with stool specimens that tested positive until a negative test result was received. A repeat stool culture was taken at least 48 hours after the first negative result. Asymptomatic employees who received positive Salmonella Mbandaka test results were required to have two negative stool cultures before returning to work. Because antibiotics can increase the likelihood of prolonged Salmonella shedding in stool, treatment was not recommended.[1] None of the restaurant employees received antibiotics to treat asymptomatic Salmonella Mbandaka infection. The duration of Salmonella shedding among the four asymptomatic restaurant employees with positive cultures varied (range = 31–123 days).

A team of local and state public health officials collected 80 environmental samples from food and nonfood contact surfaces to be tested for Salmonella. Salmonella was isolated from 39 (49%) environmental samples. Positive samples shared the same PFGE and WGS results as the outbreak subtype (Table). Positive samples were collected throughout the restaurant kitchen, including cooking, preparation, dishwashing, storage, and employee restroom areas. Following the identification of a new case in September 2018, a second round of environmental Salmonella sampling was conducted. In the second round, Salmonella was isolated from 11 of 81 samples (14%) and shared the outbreak subtype. Positive environmental sites were generally similar, but not identical, in the two rounds. Environmental, asymptomatic employee, and symptomatic patient isolates identified by core genome multilocus sequence typing (cgMLST) revealed three clades (A, B, and C) (Table). Isolates within each clade were highly related, differing by ≤5 alleles. The clades were also considered highly related to each other, differing by ≤15 alleles. All environmental and employee isolates were in clade A; symptomatic patient isolates were identified in all three clades.