RESUMEN
BACKGROUND: The high negative predictive value (NPV) of a negative nasal methicillin-resistant Staphylococcus aureus (MRSA) result in suspected MRSA pneumonia is well established; however, data are limited on the NPV of samples collected prior to hospital admission for critically ill patients. OBJECTIVE: To evaluate the predictive characteristics of MRSA nares screening performed prior to hospital admission in critically ill adult patients diagnosed with pneumonia. METHODS: A retrospective analysis was conducted in critically ill patients with pneumonia and MRSA nares screening within 60 days of respiratory culture. The primary outcome was NPV of MRSA nares for MRSA pneumonia using samples within 60 days compared to in-hospital respiratory cultures. A sensitivity analysis was performed for samples within 30 days. Secondary outcomes were prevalence, positive predictive value (PPV), sensitivity, specificity, and MRSA pneumonia risk factors. RESULTS: The NPV for MRSA nares screening collected prior to hospital admission was high at 98% (95% CI = 96%-99%) for samples collected within 60 days (n = 243) and 99% (95% CI: 94%-99.9%) for samples within 30 days (n = 119). Specificity for MRSA nares collected 60 days prior to admission (96%, 95% CI: 93-98) and 30 days (96%, 95% CI: 91%-99%) were both high. PPV and sensitivity were lower. Risk factors for MRSA pneumonia were similar. CONCLUSION AND RELEVANCE: MRSA nares screening within 60 days of intensive care unit admission has a high NPV and specificity for MRSA pneumonia in critically ill patients and may be a powerful stewardship tool for avoidance of empirical anti-MRSA therapy.