ABSTRACT
BACKGROUND: In December 2006, we investigated an outbreak of Salmonella serotype Tennessee in a neonatal intensive care unit (NICU) that coincided with a nationwide Salmonella Tennessee outbreak associated with contaminated peanut butter. METHODS: Salmonellosis was defined as isolation of Salmonella Tennessee from any clinical specimen or more than 1 episode of bloody stool within a 24-hour period. We conducted a cohort study among 13 NICU infants, reviewed medical records, cultured stool from infants and staff, collected environmental samples, and examined infection control practices. RESULTS: Ten of the 13 infants had salmonellosis (77%). No medical or dietary risk factors were identified. The proportion of days in which the NICU census exceeded its 11-bed design capacity was higher in December compared with the previous 11 months (41.9% vs 0.3%; P < .001). Hand sinks did not meet operational standards. Salmonella Tennessee was isolated from 9 of the 13 infants, 2 of 40 staff members, and 6 of 42 environmental samples; all isolates matched the pulsed-field gel electrophoresis pattern of the nationwide Salmonella Tennessee outbreak. CONCLUSIONS: Although the source of Salmonella Tennessee was not identified, the high census and limited access to sinks likely facilitated transmission to the NICU infants. Infection control interventions, including halting new NICU admissions, interrupted further transmission.
