Background:  This study aimed to:  estimate trends in incidence of clinically significant methicillin-resistant Staphylococcus aureus (MRSA) in a cohort of HIV-infected adults, characterize cases according to source of acquisition (nosocomial /community), characterize cases by antimicrobial susceptibility, and identify risk factors for clinically significant MRSA.

Methods:  A retrospective cohort was assembled for the period January 1, 2000 to December 31, 2003. The outcome was initial clinically significant MRSA during the study period. Clinical significance of each MRSA episode was validated by record review. Episodes were categorized, by primary site, as:  skin or soft tissue, blood, respiratory, and other. Incidence rates were estimated by half-year. Risk factors were identified using Cox modeling.

Results:  Of 126 potential initial events, 94were classified as clinically significant. Primary source of the 94 events were distributed as:  78 (83.%) skin or soft tissue; 9 (10%) blood;  6 (6%) respiratory; and 1 (1.0%) other. Using hospitalization in the prior year as an indicator of nosocomial acquisition, 56 (60%) and 38 (40.%) would be classified as community acquired and nosocomial, respectively. Of reported antibiotic susceptibilities, only cotrimoxazole resistance was significantly associated (p < 0.05) with nosocomial acquisition (odds ratio 12.4, p = 0.021). The 3455 cohort members contributed 7003 person-years of follow-up. The clinically significant MRSA incidence rate increased 6.2-fold from the first to the last half year. In multivariate analysis, independent predictors of clinically significant MRSA included:  heterosexual HIV transmission (HR 0.1, p = 0.012; reference: all others); CD4 < 50 (HR 2.5, p = 0.003); log₁₀ HIV plasma viral load (HR 2.1 _{3.0 to 3.99}, 3.2 _{4.0 to 4.99}, 4.2 _{≥ 5.0}; p = 0.0001; reference < 2.99); and antiretroviral therapy in half year (HR 0.6; p = 0.02:reference:none). The effect of plasma viral load was confined to those with CD4 ≥ 50.

Conclusions:  Incidence of initial clinically significant MRSA events increased over 6-fold over a 4-year period. The association between clinically significant MRSA and HIV viral load, and its effect modification by CD4, is previously unreported and suggests a direct effect of HIV on anti-staphylococcal defenses.

Keywords: MRSA; Risk Factors; Incidence