Methicillin-Resistant Staphylococcus aureus (MRSA) Nasal Screening: Clinical Use for Excluding Diabetic Foot Infection with MRSA

LETTER

I read with great interest the recently published manuscript by Mergenhagen and colleagues (1) who retrospectively studied the use of methicillin-resistant Staphylococcus aureus (MRSA) nasal colonization in predicting the absence of MRSA in clinical cultures collected from adults with diabetic foot infections. A total of 8,163 of cultures obtained from below the ankle of unique inpatients at the Veterans Affairs medical centers between 1 January 2007 and 1 January 2018 were included. MRSA was isolated in 7.5% of cultures, and a MRSA nasal screening test taken within 7 days of culture was positive in 17.8% of cultures. The negative predictive value (NPV) of MRSA nasal screening for MRSA diabetic foot infections was 89.6%.

I agree with Mergenhagen et al. that a negative MRSA nares swab may be predictive of the absence of MRSA in a subsequent culture from a diabetic foot infection. This would be helpful for de-escalation of empirical anti-MRSA antibiotic treatment. However, clinicians should know the local prevalence of MRSA before making clinical decisions. As the prevalence decreases, the NPV increases (2). Note that Mergenhagen et al. found low MRSA infection rates of 7.5% in this cohort, which were lower than the rates of 15% to 30% described in previous studies (3, 4). To that end, one could expect a lower NPV for MRSA nasal screening for MRSA diabetic foot infection in a population where MRSA is more prevalent. For instance, for a population in which the prevalence of MRSA is 30% (as described in previous studies), the new calculated NPV would decrease to 80.4%. That infers a probability of 19.6% for MRSA diabetic foot infection despite a negative MRSA nasal screen. I further agree with Mergenhagen et al. (1) that MRSA infections are associated with significant morbidity and mortality. Thus, failure to empirically treat MRSA while awaiting final tissue cultures would result in a high risk of treatment failure.

The Infectious Disease Society of America (IDSA) guidelines for management of diabetic foot infections published in 2012 recommend empirical MRSA coverage for all severe diabetic foot infections (5). For mild and moderate diabetic foot infections and in the absence of history of MRSA infection, IDSA recommends empirical coverage for MRSA if the local prevalence of MRSA is high enough, i.e., 50% and 30%, respectively (5). This again emphasizes the importance of knowledge of local epidemiology and the risk factors for MRSA infections in treating empirically diabetic foot infections.

Yet, the threshold of an acceptable NPV for optimal therapeutic decisions has not been standardized, and it often differs by clinician. Data supporting MRSA nasal screening for empirical MRSA coverage in patients with pneumonia are robust (5, 6). The NPV of MRSA nasal screening was high, at 98.6%, in a population where the prevalence of MRSA was 10.5% (6). In this case, one may feel very confident to de-escalate MRSA coverage in treating pneumonia if the MRSA nasal screen is negative. Moreover, the NPV of MRSA nasal screening for MRSA wound infections was shown to be far lower than the NPV for MRSA pneumonia (7). Rioux et al. (7) studied nasal swab screening for MRSA in predicting MRSA bloodstream, wound, and respiratory infections in 273 inpatients where the prevalence of MRSA infection in the study population was a low 4.4%. The NPV of the MRSA nasal screen for all MRSA infections was 98%. However, the NPV of the MRSA nasal screen specific for MRSA wound infections was only 71%. Consequently, MRSA nasal screening seems to be more valuable in regard to allowing for de-escalation or confirmation of need for continued MRSA therapy in patients with pneumonia than in those with wound infections (5, 7).

In conclusion, I commend Mergenhagen and colleagues for the article and further advise clinicians to have a good understanding of the limitations of a test and their local epidemiology before making a final decision about the management of patients with diabetic foot infection. I also support, if local prevalence matches the data set of this cohort with no recent history of MRSA infection, that nasal MRSA screening can provide optimal value to antimicrobial stewardship in treating diabetic foot infections.