Background: Clinical intuition suggests that methicillin-resistant Staphylococcus aureus (MRSA) infections should be less likely on anti-MRSA antibiotics (anti-MRSA) than off of them. If so, the common use of empiric anti-MRSA could decrease MRSA infections among carriers, and may impact acquisitions among non-carriers. We investigated the relationship between anti-MRSA and these outcomes.
Methods: We used data from inpatients in the Veterans Health Administration from October 2007 through September 2010 who had no MRSA-positive cultures at admission. Cumulative risk regression was used to model MRSA acquisition (a negative screening test followed by a positive) and MRSA nosocomial clinical culture (first positive MRSA culture greater than 2 days after admission), while accounting for the competing risk of death. Immortal time was excluded and the time of MRSA culture was lagged 2 days to account for incubation. To verify that the assigned lag correctly assigned exposure status to outcome, we performed chart review. Because incubation time is variable, a subanalysis of coagulase-negative Staphylococcus (CONS) culture-positive admissions was used as a population in which anti-MRSA was likely started for purposes other than MRSA. Anti-MRSA and ICU status were allowed to change over time. Models were adjusted for hospital and admission characteristics. The model for MRSA culture was restricted to screen-positive admissions to focus on progression.
Results: The subhazards ratio of MRSA acquisition given anti-MRSA was 0.7 (95%CI 0.6-0.7). The same for MRSA culture was 2.5 (95%CI 2.2-2.8). Chart review revealed that MRSA cultures were often collected well after anti-MRSA was started, thereby misattributing events to anti-MRSA exposure. An analysis of CONS admissions yielded a non-significant subhazards ratio of 0.7 (95%CI 0.4-1.3).
Conclusion: There is support for a protective effect of anti-MRSA on MRSA acquisition and possibly on MRSA infection as well, although the latter remains inconclusive. Further investigation is necessary given the possibility that recent decreases in MRSA may not be due to improvement in infection control alone. Use of anti-MRSA for this purpose is not recommended.
Figure 1. MRSA acquisition
Figure 2. MRSA culture
Figure 3. MRSA culture, restricted to admissions positive for coagulase-negative Staphylococcus
V. Stevens, None
R. E. Nelson, None
K. Khader, None
B. Sauer, None
C. Nielson, None
M. Rubin, None
M. Samore, None