Nosocomial infections is a major problem in intensive care units due to both growing incidence and pathogens implicated which become increasingly resistant to antibiotics. According to the Center for Disease Control (USA), Staphylococcus aureus is responsible for approximately 10% of cases. In Europe, 79% of the S.aureus strains are resistant to methicillin, a routinely used antibiotic. Recommendations for the prevention of transmission of these resistant bacteria are rarely based on controlled trials. Therefore the aim of our study is to evaluate two protocols to prevent the acquisition of methicillin-resistant S. Aureus in intensive care units: either a reinforced isolation precautions protocol or a standard precautions protocol.

Recommendations for the prevention of transmission of resistant bacteria in intensive care units (ICU) are rarely based on controlled trials. For this reason, we compared a reinforced isolation precautions protocol (RIPP) with a standard precautions protocol (SPP) for the acquisition of methicillin-resistant Staphylococcus aureus (MRSA) in 2 intensive care units.

Evaluation: the risk for MRSA carriage was defined on admission if 1 of 3 criteria were met: hospitalization in the past year, transfer with prior length of stay ≥2 days, prior history of MRSA in the 5 past years.

Intervention: Randomization 1/1 of a total of 500 patients to either protocol; MRSA screening was performed at the sites of carriage and colonization at inclusion, every week and at ICU discharge in all patients; the results were given to the clinicians only for the patients of the RIPP group.

Protocols: the SPP was consistent with the CDC recommendations and included transmission-based isolation precautions to patients with clinical samples involving resistant bacteria (including MRSA) or highly transmissible organisms. The RIPP included the extension of isolation precautions (1) to patients at risk for MRSA on admission until screening results proved negative and (2) to MRSA-positive patients on screening or clinical samples until further negative samples, in whom nasal mupirocin decontamination was added. Compliance with the recommendations of each protocol was controlled by an audit.

The efficacy is assessed on the proportion of patients who acquired MRSA at any site.

Inclusion Criteria:

• Adults over 18 years
• Expected length of stay > 48h in intensive care unit
• Informed written consent

Exclusion Criteria:

• Cerebral death
• Care limitation
• Neutropenia
• Documented MRSA on admission
• Patients receiving antistaphylococcal topical antibiotics on admission