Evaluation of the reliability of MRSA screens in patients undergoing universal decolonization.

Abstract

Colonization of methicillin-resistant Staphylococcus aureus (MRSA) can be detected via nasal screens. Evidence indicates that negative MRSA nasal screens may be used to de-escalate anti-MRSA antibiotics in pulmonary infections. In the ICU, universal decolonization with intranasal mupirocin is implemented to reduce MRSA infection risk. This study aimed to determine whether mupirocin administration affects the reliability of MRSA PCR nasal screens. This retrospective study divided subjects based on timing of intranasal mupirocin administration-before and after MRSA screen. Subjects with confirmed pulmonary infection that received vancomycin, blood/respiratory cultures, and had MRSA PCR screen collected were included. Subjects with concurrent infection requiring vancomycin or MRSA infection in prior 30 days were excluded. Primary outcome of this non-inferiority study was the negative predictive value (NPV) of the screen. Secondary outcomes included the positive predictive value (PPV), sensitivity, and specificity of the screen and duration of vancomycin. Ultimately, 125 subjects were included in each group. The NPV in the group receiving mupirocin before screen was 95.2%, whereas the NPV in the group receiving mupirocin after screen was 99%. The difference between groups was -3.8% (90% CI -7.8%-0.2%; p=0.31), which failed to meet non-inferiority criteria. The secondary outcomes of PPV, sensitivity and specificity of the screen were similar in both groups. The duration of vancomycin was significantly longer in subjects receiving mupirocin before screen (3 days vs. 2 days; p<0.05). Intranasal mupirocin prior to the screen may reduce NPV in pulmonary infections. Approach de-escalation of vancomycin based on screen results with caution.