Serratia marcescens (SM) pulmonary infections in a cardiothoracic intensive care unit (CTICU)

Abstract

BACKGROUND: The CTICU is a 22-bed unit that provides the immediate post operative care of cardiac surgery (CT) patients. In March 2004, a cluster of four patients with serious pneumonias and one patient with bacteremia due to SM were identified in the CTICU. SM has sporadically been the cause of infection in the CTICU, but no clustering of cases had ever occurred. All five patients had CT in late March 2004, two on the same day in the same operating room (OR). All patients were admitted post-op to the CTICU. The cases occurred with increasing severity and an overall mortality rate of 60%. OBJECTIVE: To investigate a cluster of SM infections in a CTICU. METHODS: The investigation centered on environmental inspections of potential sources of SM in both the CTICU and the OR. In the CTICU the investigation focused on respiratory therapy and nursing practices. In the OR investigation centered on the cleaning of equipment, OR rooms, common supplies and OR personnel practices. Environmental cultures were obtained. All patient isolates were sent for molecular typing. RESULTS: Molecular typing revealed identical genetic patterns supporting a common source for the SM. No direct sources were identified in the CTICU although environmental culturing found one culture positive for SM from a handwashing sink. All patients involved in the outbreak were found to have had a trans-esophageal echocardiogram (TEE) as part of their CT procedure. To lubricate the TEE probe, ultrasound gel (UG) was applied directly into the patient's mouth prior to passing the probe down the esophagus. The UG is a reusable nonsterile gel kept on the TEE cart in the OR room. The UG was not to be used internally, as per manufacturer's instruction. All five of the cases involved had the UG used for the TEE. The particular TEE probe used in each case was not identifiable as the probes were not labeled and were stored in a central location after reprocessing. The UG was not available for culturing. Following removal of the gel there have been no additional SM cases. CONCLUSIONS: 1) Common source of the SM may have been the UG used for the TEE procedures. 2) Epidemiologic investigations using molecular typing help to improve current practice in the OR. 3) TEE probes should be labeled and logged to allow for future tracking. 4) Only sterile products should be used to lubricate probes.