[Case control study of extended-spectrum betalactamase producing Serratia marcescens outbreak in a paediatric intensive care unit].

OBJECTIVES

To identify patient-related risk factors of infection and ways of transmission of extended-spectrum betalactamase (ESBL) producing Serratia marcescens in the paediatric intensive care unit (PICU) of Amiens university hospital (France) between June and July 2002.

METHODS

Five cases (four pulmonary infected and one stool contaminated symptom-free neonates) and 35 controls, admitted in the PICU, are included. S. marcescens ESBL analysed are isolated from respiratory tract and faecal samples for cases and urine and pus samples from two non-paediatric other patients. Univariate and multivariate analysis are performed on EPI INFO 6.04 dFr and SPSS 11.0.1.

RESULTS

S. marcescens ESBL infections or colonisations rate is 12.5% [4.7-27.6]. The incidence is 8.8 [6.7-11.6] per 1000 hospital-stay days. By univariate analysis, cases and controls don't differ with respect of age, sex, and weight at admission or preterm delivery. Cases don't have more often invasive nursing care than controls. But, they were intubated (P <0.03) and hospitalised (P <0.03) for a longer time than controls. Linear regression analysis showed that duration of intubation was independent predictor of acquisition of S. marcescens ESBL (P <0.008). S. marcescens ESBL strains implicated in pulmonary infections, showed the same pattern of multidrug resistant and ERIC-PCR profile. This clone differs from others isolated from stool or other samples from other hospital wards.

CONCLUSION

As S. marcescens cross-colonization appears to be due to lake of hand hygiene and asepsis during invasive nursing care, reinforcing hygiene measures permit to contain the outbreak.