Background: Despite high negative predictive values (NPVs) seen with methicillin-resistant Staphylococcus aureus (MRSA) nares polymerase chain reaction (PCR) assays, utilization of both respiratory sample Gram stain and MRSA nares PCR in patients with pneumonia may contribute to overuse of laboratory resources. The purpose of this study was to evaluate if a Gram stain demonstrating no Gram-positive organisms from a respiratory sample is sufficient to allow for de-escalation of vancomycin therapy. Methods: This single center study retrospectively identified intensive care unit (ICU) patients started on vancomycin for presumed pneumonia at University of Wisconsin (UW) Health in Madison, WI between August 2022 and March 2023. Patients with respiratory sample demonstrating no Gram-positives on Gram stain met inclusion criteria if the sample was ordered within 24 h of vancomycin initiation. The primary outcome was NPV of respiratory sample Gram stain demonstrating no Gram-positive organisms with respect to MRSA detection of the respiratory culture. Secondary outcomes included the NPV of combined MRSA nares PCR plus respiratory sample Gram stain, and difference in time to event in patients that had both a respiratory sample and MRSA nares PCR ordered. Results: A total of 370 patients were screened for study eligibility; of which 99 patients met inclusion criteria. NPV of respiratory sample Gram stain was 99% for MRSA culture. The combined NPV of respiratory sample Gram stain plus MRSA nares PCR was 98.9% for MRSA culture (n = 88). Respiratory sample was ordered 2.3 h faster compared to MRSA nares PCR (4.3 vs 6.6 h, P = .036). Respiratory sample Gram stain resulted 4.5 h faster compared to MRSA nares PCR (10.7 vs 15.2 h, P = .002). Conclusion: Respiratory sample Gram stains demonstrating no Gram-positive organisms may be used to de-escalate vancomycin and deprioritize the use of MRSA nares PCR.
Keywords: Gram stain; ICU; MRSA nares PCR; pneumonia.