Reviewed by: Cynthia T. Nguyen, PharmD, University of Chicago Medicine and Rebekah Moehring, MD, MPH, Duke University

Two recent observational studies from the depths of the Veterans Affairs (VA) robust data warehouses bring insights on where antibiotic stewards can target efforts to avoid anti-MRSA therapy.

VA Stewies Mergenhagen and colleagues sought to determine if the absence of MRSA nasal carriage predicted the absence of MRSA in cultures of >8,000 diabetic foot infections during 2007-2018. VA inpatient facilities performed MRSA nasal swabs on all patients on admission, transfers between units, and at discharge (72% via PCR, 28% chromogenic agar). Below the ankle cultures obtained after, but within 7 days of the MRSA nasal swab were included in the evaluation. Cultures obtained from abscesses, fluid, surgical samples, aspirates, and bone cultures were considered deep cultures. A positive MRSA nares screen occurred in 17.8% and there were >17,000 isolates cultured. Most were S. aureus, with MRSA comprising 7.5% of isolates and MSSA comprising 24.8% of isolates. Negative MRSA nares colonization within 7 days of clinical culture consistently had a NPV of approximately 90% for the absence of clinical MRSA infection (NPV 89.2% in those with deep cultures, 90.3% in superficial). Importantly, the data are limited by quality of the documentation, particularly related to culture labeling, and the quality of the cultures themselves, since deep cultures are less frequently obtained for diabetic foot infections. 

Another group of VA Stewies Jones et al. performed a retrospective cohort study of >88,0000 admissions in which patients received standard CAP therapy (fluoroquinolone or macrolide+beta-lactam) or CAP + anti-MRSA therapy within hospital day 1. The study aimed to estimate the association between additional empiric anti-MRSA therapy on 30-day mortality. Using propensity scores and inverse probability weighting, the authors found that all the negative outcomes were HIGHER among patients empirically started on anti-MRSA therapy. Using instrumental variable analyses, the authors explored key subgroups. Again, empiric anti-MRSA coverage was associated with mortality -- even among those with severe pneumonia admitted to ICUs, those with MRSA risk factors, and those with positive surveillance screens for MRSA. No difference in mortality was seen among the 2% of the population with a positive clinical culture for MRSA, which were mostly from respiratory specimens. These findings bring into question the widely held belief that patients with pneumonia and traditional MRSA risk factors benefit from anti-MRSA therapy in the empiric setting. This study has limitations given this is observational data. Even with advanced statistical methods, there could be residual confounding. Importantly, the study only considered Day 1 therapy, and thus outcomes attributed to a single day’s decisions may not make a difference in the long-term. Subjects had very few non-vancomycin regimens; the negative effects might not be seen with other anti-MRSA agents. Finally, the study may have limited generalizability outside of VA populations.

In conclusion, these anti-MRSA research questions investigated in large retrospective, VA databases are provocative for antibiotic stewards to consider as we aim to develop strategies to avoid unnecessary anti-MRSA agents.


  1. Mergenhagen KA, Croix M, Starr KE et al. The utility of methicillin-resistant Staphylococcus aureus nares screening for patients with a diabetic foot infection. Antimicrob Agents Chemother. 2020 Jan 27. pii: AAC.02213-19. 
  2. Jones BE, Ying J, Stevens V, et al. Empirical Anti-MRSA vs Standard Antibiotic Therapy and Risk of 30-Day Mortality in Patients Hospitalized for Pneumonia. JAMA Intern Med. 2020 Feb 17.