Association of neighborhood-level factors with hospitalization for community-associated methicillin-resistant Staphylococcus aureus, New York City, 2006: a multilevel observational study

BACKGROUND: Hospitalizations with community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) infection have increased in New York City, with substantial geographic variation across neighborhoods. While individual-level risk factors, such as age, sex, HIV infection, and diabetes have been described, the role of neighborhood-level factors (e.g., neighborhood HIV prevalence or income) has not been examined. METHODS: To explore plausible neighborhood-level factors associated with CA-MRSA-related hospitalizations, a retrospective analysis was conducted using New York City hospital discharges from 2006 and New York City-specific survey and health department surveillance data. CA-MRSA-related hospitalizations were identified using diagnosis codes and admission information. Associations were determined by using sex-specific multilevel logistic regression. RESULTS: The CA-MRSA hospitalization rate varied by more than six-fold across New York City neighborhoods. Females hospitalized with CA-MRSA had more than twice the odds of residing in neighborhoods in the highest quintile of HIV prevalence (adjusted odds ratio [AOR](Q5 vs. Q1) 2.3, 95% CI: 1.2, 2.7). Both males and females hospitalized with CA-MRSA had nearly twice the odds of residing in neighborhoods with moderately high proportion of men who have sex with men (MSM) residing in the neighborhood (males: AOR(Q4 vs. Q1) 1.7, 95% CI: 1.1, 2.7; females: AOR(Q4 vs. Q1) 2.0, 95% CI: 1.1, 3.6); but this association did not hold for neighborhoods in the highest quintile (males: AOR(Q5 vs. Q1) 1.2, 95% CI: 0.76, 1.8; females: AOR(Q5 vs. Q1) 1.5, 95% CI: 0.82, 2.7). CONCLUSIONS: Neighborhood-level characteristics were associated with CA-MRSA hospitalization odds, independent of individual-level risk factors, and may contribute to the population-level burden of CA-MRSA infection.