Potential for selection and misclassification bias when sampling men who have sex with men (MSM) in gay bars

Background/Objectives: HIV research and prevention activities often target gay bar patrons. Gay bars are typically analyzed as homogeneous venues. This study describes variation between gay bars and the impact of enrollment in different bars on the estimated prevalence of unprotected anal intercourse (UAI) among MSM.

Methods: MSM were interviewed in 12 traditional gay bars (frequented mainly for socialization), 8 mixed bars (with heterosexual and gay/bisexual patrons), 7 cruising/hustler bars (where MSM seek sex partners) and one drag bar (known for male/female impersonators) in Upstate New York (n=594) (2000-2004). MSM's characteristics and the prevalence of UAI in different bars were compared using chi-square, Fisher's exact tests and poisson regression.

Results: Cruising/hustler bar patrons were more likely to be MSM of color (p<0.0001) and HIV positive (p=0.0034). Mixed bar patrons were more likely to be younger (p=0.0022) and bisexual (p=0.0088). The prevalence of UAI was highest in cruising/hustler bars (71.8%) and lowest in traditional gay bars (51.4%, p=0.0114). However, 4 mixed and 6 traditional gay bars were high risk (>65% had UAI). In the multivariable regression analysis, cruising/hustler bar patrons were slightly more likely to have UAI than traditional gay bar patrons (adjusted prevalence risk ratio: 1.3, 95% CI: 0.9-1.9).

Discussion: Enrolling MSM in very few bars or one type of bar does not produce a representative sample. Characteristics of MSM in one bar are not generalizable to patrons of other bars of the same type. Collecting contextual information about each bar may minimize misclassification and selection bias and guide prevention efforts.