Value of Screening and Isolation for Control of Methicillin‐ResistantStaphylococcus aureus

Sir—Nijssen et al. [1] screened patients admitted to their medical intensive care unit (ICU) for Staphylococcus aureus carriage over a period of 10 weeks. This ICU has endemic levels of methicillin-resistant S. aureus (MRSA) (5 [36%] of 14 clinical isolates of S. aureus were MRSA). The authors claim that there was not a single episode of transmission of S. aureus, including MRSA and methicillin-susceptible S. aureus (MSSA), between their patients in this period. They conclude that active surveillance and isolation of patients with MRSA may not be warranted, and they caution against the introduction of such measures in guidelines. Although we appreciate the systematic daily sampling of their ICU patients, we question the validity and interpretation of their data. First, surveillance cultures were based on direct plating of nasal swabs and en-dotracheal aspirate samples only. They did not sample the perineum nor did they sample other body sites with lesions (e.g., wounds or intravascular access sites) for S. aureus colonization. Also, they did not use enrichment broths needed for sensitivity when screening for staphylococcal carriage [2]. Thus, they probably missed a sizable proportion of the S. aureus col-onization events. A second point is the arbitrary choice of a 48-h period for discrimination of importation from acquisition. It is not clear why the authors did not use their initial microbiological screening, taken within 12 h after admission, for identifying staphylococci as either imported or acquired. Thus, the authors may have misclassified many patients as importers who were actually contaminated during the first 48 h after admission to the ICU. Even brief encounters with a staphylococcal source or reservoir may suffice to transmit S. aureus [3]. We suspect that the use of a 12-h period as the criterion would yield many potential cases of transmission in the ICU discussed in Nijssen et al. [1]. The fact that the authors did not find strains that had identical DNA PFGE profiles does not rule out cross-transmission, because only patients were screened; other reservoirs of staphylococci (e.g., health care workers and the environment [3]) were not. Also, the article by Nijssen et al. [1] lacks a graphic presentation of the expected clustering of common clones of MRSA and MSSA [4]. Of interest, a similar but methodologically better study (involving more sample sites and using enrichment broth) was performed at the same time in 2 ICUs in the United Kingdom. In this study [5], the authors …