SERRATIA MARCESCENS (SM) OUTBREAK IN A NEONATAL INTENSIVE CARE UNIT (NICU) RELATED TO THE EXIT PORT OF HIGH FREQUENCY OSCILLATOR (HFO)

Background: SM is an opportunistic pathogen known for its role in NICU nosocomial infection.

Objective: We report an outbreak thought to be related to aerosols from the HFO exit port of a colonized infant.

Methods: The level 3 NICU admits 900 infants/yr who are housed in 3 units; the outbreak unit is a 3067 feet ² space divided into three connected pods. The index case was a <1500 gm premature infant with respiratory tract colonization whose mother had a surgical site infection caused by SM. Because of the morbidity associated with SM Infection Prevention and Control Services (ICPS) placed the patient on contact isolation (CI). Three weeks later a 2^nd colonized infant who had spent 48 hours in the NICU was found. ICPS met with NICU staff to review all respiratory care policies and procedures. Only sterile water was used for respiratory, enteric or skin care. ICPS implemented a comprehensive staff and family education plan. Over the ensuing months 5 more affected infants were identified and placed on CI. A line listing showed the only common risk factor was residency in the same pod. The unit was closed to further admissions. All children were placed on CI. A multidisciplinary team (ICPS, medicine, nursing, microbiology, biomedical engineering, respiratory therapy) met in the pod to review all possible sources of the outbreak. Multiple (n = 16) environmental samples were taken. A small puddle of water directly under the HFO exit port was noted and sampled, as well as the port and the incubator. Respiratory cultures were taken on all inpatients. Samples positive for SM were processed according to the CDC 1 day E. coli pulsed field gel electrophoresis protocol at the National Microbiology Laboratory, Winnipeg.

Results: Seven patients were colonized (n=5) or infected (n=2 ). Of all environmental samples, only the exit port of the index case s HFO grew SM. This isolate and that of the index case were genotypically identical. Two of the clinical specimens were identical (A1); three more were related (A2, A3). Upon notification the HFO manufacturer recommended a containment device be retroactively fitted to the exit port and indicated that subsequent models will have a containment device to scavenge for droplets/aerosols. Following retrofitting no further nosocomial spread occurred.

Conclusions: A S. marcesans outbreak was terminated through a combination of CI, education, increased hand hygiene, closing the unit to admissions, single room placement and retrofitting of the exit port of HFO.  We hypothesize that small particle aerosols generated by the HFO may have dispersed the organism and facilitated patient-patient and HCW-patient spread.