RESUMO
BACKGROUND: Risks for methicillin-resistant Staphylococcus aureus (MRSA) among those with HIV infection have been found to vary, and the epidemiology of USA-300 community-acquired (CA) MRSA has not been adequately described. METHODS: We conducted a retrospective review of HIV-infected out-patients from January 2002 to December 2007 and employed multivariate logistic regression (MLR) to identify risks for MRSA colonization or infection. Pulsed-field gel electrophoresis (PFGE) was used to identify USA-300 strains. RESULTS: Seventy-two (8%) of 900 HIV-infected patients were colonized or infected with MRSA. MLR identified antibiotic exposure within the past year [odds ratio (OR) 3.4; 95% confidence interval (CI) 1.5-7.7] and nadir CD4 count <200 cells/microL (OR 2.5; 95% CI 1.2-5.3) as risks for MRSA colonization or infection. Receipt of antiretroviral therapy (ART) within the past year was associated with decreased risk (OR 0.16; 95% CI 0.07-0.4). Eighty-nine percent of available strains were USA-300. MLR identified skin or soft tissue infection (SSTI) as the only predictor for infection with USA-300 (OR 5.9; 95% CI 1.4-24.3). CONCLUSION: Significant risks for MRSA among HIV-infected patients were CD4 count nadir <200 cells/microL and antibiotic exposure. Only the presence of an SSTI was associated with having USA-300, and thus the use of patient characteristics to predict those with USA-300 was limited. In addition, ART within the previous year significantly reduced the risk of MRSA colonization or infection.