Comparison of Two Surveillance Methods for Detecting Nosocomial Infections in a Neonatal Intensive Care Unit / 병원감염관리

ABSTRACT

BACKGROUND: The aim of this study was to evaluate the sensitivity of a clinicians' self-report method for the detection of nosocomial infections (NIs) in comparison with a total surveillance method in a neonatal intensive care unit (NICU). METHODS: Two surveillance methods were concurrently performed in the NICU of a university hospital during 5 months in 2003. Clinicians' self-report surveillance (CSRS) was based on the retrospective verification of monthly reports of positive bacteriologic results by NICU clinicians. Total surveillance (TS) was done prospectively by an infection control nurse based on chart review and laboratory data. RESULTS: One hundred fifty nine patients accounting to 2759 patient-days were included in the study. Twenty-seven NIs among 26 patients were identified by TS. The sensitivity of CSRS compared to TS was 14.8% (4 of 27 NIs). The specificity was 98.5% (131 of 133 non-NIs). Kappa measures of agreement were -0.309. CONCLUSIONS: Our results confirm that the retrospective review of charts and laboratory data by clinicians lacks sensitivity and agreement for the surveillance of nosocomial infections.