ABSTRACT
BACKGROUND: We describe our experience using a real-time polymerase chain reaction (PCR) assay for methicillin-resistant Staphylococcus aureus (MRSA) during a period of active surveillance in the neonatal intensive care unit (NICU) from March 2007 until November 2007. OBJECTIVE: To compare PCR with bacterial culture methods and find the screening algorithm that most successfully ensures appropriate isolation of colonized patients. METHODS: Patients in the NICU were screened for MRSA on admission and weekly thereafter until discharge. Healthcare workers (HCWs) were also screened as part of an outbreak investigation. A total of 599 individuals were screened for MRSA with both a PCR assay and selective bacterial culture. Strain typing was performed on all MRSA isolates to determine clonal relatedness. RESULTS: Twenty-one of 435 infants (4.8%) screened positive for MRSA with the PCR assay. Only 11 patients (52.4%) had concomitant bacterial cultures positive for MRSA. Compared to bacterial culture, the PCR assay had a sensitivity of 100% and a specificity of 97.6%, with a positive predictive value (PPV) of 52.4%. Infants that tested positive for MRSA by both culture and PCR were more likely to have a positive PCR assay result when retested than were those who tested positive by PCR alone (80% vs 20%; P = .02). Strain typing of MRSA isolates identified a common clone in only 2 colonized infants. CONCLUSION: Our data show that, in our neonatal population, the reproducibility of PCR assay results for culture-negative patients was low compared with the reproducibility of results for culture-positive patients. Furthermore, the low PPV suggests that for nearly half of individuals who were PCR-positive, the result was falsely positive, which argues against the use of PCR assays alone for MRSA screening in the NICU.
