Positive nasal polymerase chain reaction assay is not sensitive in predicting concurrent or subsequent infection in critically ill patients.

Staphylococcal infection is associated with significant morbidity and mortality in critically ill patients. Using data from 16,681 patients who had a nasal polymerase chain reaction (PCR) assay on admission to the intensive care unit (ICU) of Royal Perth Hospital between March 2006 and September 2016, this retrospective cohort study assessed whether nasal colonisation on admission to an ICU was predictive of concurrent or subsequent infections. Culture-proven infections were identified using the hospital microbiology database. Of the 16,681 patients included, 565 (3.4%) had a positive methicillin-resistant (MRSA) assay, 146 (0.9%) had a positive methicillin-sensitive (MSSA) assay and eight (0.05%) had both positive MRSA and MSSA assays. Of those 565 patients with a positive MRSA PCR assay, 79 (13.8%) had concurrent or subsequent MRSA infections. Of those 146 patients with a positive MSSA PCR assay, only 5 (3.4%) had MSSA infection. The sensitivity and specificity for the MRSA PCR assay in predicting concurrent or subsequent MRSA infection were 72.7% (95% confidence intervals (CI) 63.4%-80.8%) and 97.0% (95% CI 96.8%-97.3%), respectively. The sensitivity and specificity for the MSSA PCR assay in predicting concurrent or subsequent MSSA infection were 3.3% (95% CI 1.1%-7.6%) and 99.1% (95% CI 98.9%-99.2%), respectively. Both nasal MRSA and MSSA PCR assays had a high specificity and negative predictive value in predicting MRSA and MSSA infections, respectively, suggesting that in centres without endemic infections, a negative nasal MRSA or MSSA PCR assay may be useful to reduce unnecessary empirical antibiotic therapy against