Has Methicillin‐ResistantStaphylococcus aureusStopped Spreading in Intensive Care Units?

SIR—Nijssen et al. [1] found no spread of methicillin-resistant Staphylococcus aureus (MRSA) over a 10-week period from 9 MRSA-colonized patients admitted to their intensive care unit (ICU), but they did not specify what infection-control precautions were being followed in the ICU, making reported compliance with gloving and hand hygiene difficult to interpret. Compliance with standard precautions? Universal barrier precautions? A combination of both? Moreover, they reported such compliance for ICU nurses but not for other health care workers, whom they acknowledged to be more prone to spread MRSA because of more-frequent movement from patient to patient. Study methods specified that health care workers were not provided surveillance culture results but did not specify whether health care workers knew that a study was being conducted or were otherwise encouraged to maintain higher levels of compliance than were previously documented in the same ICU [2]. Even though the results of surveillance cultures were not provided to health care workers, data on the proportion of colonized patients who were known to health care workers and over what proportion of the relatively brief study period they were known to be positive for MRSA (because of previously positive or newly positive clinical cultures) would have been of interest, because such knowledge could have modified behavior for colonized patients (and possibly for all patients). The proportion of patients treated during the relatively brief study period with specific and/or empirical therapy (and the duration of treatment) with drugs active against the MRSA strain (e.g., trimethoprim-sulfamethoxazole, vancomycin, quinupristin-dalfopristin, and linezolid) also could be of interest, because such therapy may have suppressed colonization and thus suppressed transmission. Nijssen et al. [1] conclude that MRSA does not spread in hospitals such as theirs despite high rates of MRSA endemicity and that active surveillance cultures and contact precautions for colonized patients, as recommended by a Society for Healthcare Epidemiology of America (SHEA) guideline [3], are thus unneeded. Small studies, such as the one performed by Nijssen et al. [1], are susceptible to statistical imprecision, and all epidemiologic studies (including randomized, controlled trials) require confirmation by studies performed by other authors in other populations. A contemporaneous, larger study by Marshall et al. [4] found that a similar percentage of patients (6.8%) were colonized with MRSA at admission to the ICU, but another 11.4% acquired MRSA in the ICU, which was using only standard precautions. Many other studies support the finding of Marshall et al. [4] that MRSA spread and infection have continued in recent years, despite the use of standard precautions [3]. Another contemporaneous ICU study by Cepeda et al. [5] reported high MRSA transmission rates whether contact precautions or modified “standard precautions” (which were similar to contact precautions) were used and concluded that new infection-control measures are needed. MRSA caused steadily increasing proportions of S. aureus infection in National Nosocomial Infection Surveillance hospitals during the 9 years since standard precautions began to be required in US hospitals [6, 7], raising questions regarding both the validity and the generalizability of the results reported in Nijssen et al. [1]. The SHEA guideline cited 45 studies reporting significantly improved control of MRSA and/or vancomycin-resistant enterococci with active surveillance cultures and contact precautions [3], raising questions regarding the validity and generalizability of the results reported by Cepeda et al. [5]. Scores of studies with data supporting the SHEA guideline have been published or presented at national infection-control meetings since the SHEA guideline was published. Multiple studies also have shown impressive reductions in MRSA infection rates despite years of previously high MRSA rates [8–11]. Potential conflicts of interest. All authors: no conflicts.