Multifactorial origin of high incidence of Serratia marcescens in a cardio-thoracic ICU: analysis of risk factors and epidemiological characteristics.

OBJECTIVE

A four-fold increase in the incidence of Serratia marcescens occurred in a cardio-thoracic ICU within a 13-month period. Clinical, epidemiological and molecular characteristics were analysed to elucidate the outbreak's origin.

METHODS

Epidemiological data were analysed by mapping clustered cases; isolates were genotyped by AFLP analysis. A case-control study was performed to identify risk factors for the acquisition of S. marcescens. Data were obtained from files and electronic databases of the ICU and Department of Medical Microbiology. The adherence to hygiene protocols on the ICU was reviewed by a medical audit.

RESULTS

Genotyping showed 16 distinct S. marcescens strains. Twenty-one cases and 39 controls were enrolled in the case-control study. Significant differences found by univariate analysis included the duration of surgery, APACHE-II-score on ICU admission, length of ICU stay, duration of mechanical ventilation, tube feeding and the sum of the number of days per invasive device. In a multivariate logistic regression model, the length of ICU stay and tube feeding were independent risk factors. Outbreak strains were not more frequently resistant to gentamicin, ciprofloxacin, meropenem or trimethoprim-sulfamethoxazole as compared to a reference group. Hygiene protocols, including hand washing, were insufficiently practiced by the ICU's medical staff.

CONCLUSIONS

The heterogeneity of the strains points to transmission from various sources. This outbreak of S. marcescens was most probably caused by reduced hand washing and other breaks in infection prevention protocols in combination with the presence of the identified risk factors, which act by affecting the number and intensity of potential transmission events.