Duration of colonization with methicillin-resistant Staphylococcus aureus among patients in the intensive care unit: implications for intervention.

OBJECTIVES: To determine the duration of methicillin-resistant Staphylococcus aureus (MRSA) colonization or infection before entry and during hospitalization in the intensive care unit (ICU) and the characteristics of patients who tested positive for MRSA.
DESIGN: Prospective observational cohort survey.
SETTING: A combined medical and coronary care ICU with 16 single-bed rooms in a 427-bed tertiary care Veteran Affairs Medical Center. PATIENTS: A total of 720 ICU patients associated with 845 ICU admissions were followed up for the detection of MRSA from January 13, 2003, to October 12, 2003. MRSA colonization was detected in patients by using active surveillance cultures (ASCs) of nasal swab specimens obtained within 48 hours of ICU entry and 3 times weekly thereafter. The duration of colonization during ICU stay and before ICU entry was calculated after a review of surveillance culture results, clinical culture results, and medical history.
RESULTS: Ninety-three (11.0%) of 845 ICU admissions involved patients who were colonized with MRSA at the time of ICU entry, and 21 admissions (2.5%) involved patients who acquired MRSA during ICU stay. ASCs were positive for MRSA in 84 (73.6%) of the 114 admissions associated with MRSA positivity and were the sole means of identifying MRSA in 50 cases (43.8%). More than half of the MRSA-associated admissions involved patients who were transferred from hospital wards. The total bed-days of care for 38 admissions involving patients who tested positive for MRSA before ICU entry (1131 days) was nearly 20% higher than the total bed-days of care for all admissions associated with MRSA positivity (970 days). Admissions involving MRSA-positive patients were associated with a longer length of hospitalization before ICU entry (P < .001), longer length of ICU stay (P < .001), longer overall length of hospitalization (P < .001), and greater inpatient mortality than admissions involving MRSA-negative patients (P < .001). A total of 22.8% of all bed-care days were dedicated to MRSA-positive patients in the ICU, and 55 (48.2%) of 114 admissions associated with MRSA positivity involved patients who were colonized for the duration of their ICU stay.
CONCLUSIONS: In our unit, ASCs were an effective means to identify MRSA colonization among patients admitted to the ICU. Unfortunately, the majority of identified patients had long durations of stay in our own hospital before ICU entry, with prolonged MRSA colonization. Enhanced efforts to control MRSA will have to account for the prevalence of MRSA within hospital wards and to direct control efforts at these patients in the future.