Validity of self-reported substance use in men who have sex with men: comparisons with a general population sample.

PURPOSE: To understand the validity of self-reported recent drug use in men who have sex with men (MSM).

METHODS: We obtained a probability sample of Chicago men who have sex with men (MSM; n=216) and administered urine and saliva drug testing after a self-administered interview. Analyses examined participation in drug testing, the agreement between self-reported past month drug use and drug test results, correlates of underreporting, and the relative utility of drug testing versus self-reports in identifying recent marijuana and cocaine use. For marijuana and cocaine, findings were compared with those obtained from a general population sample of men (n=241).

RESULTS: More than three quarters of the participants in both samples provided at least one specimen for drug testing. Self reports in both samples showed a high degree of correspondence with drug tests for marijuana but not for cocaine. Sensitivity for cocaine use reporting was 60% for the MSM sample and 40% for the general-population men. Conditional kappa and sensitivity statistics for marijuana, cocaine, 3,4 methylenedioxymethamphetamine (i.e., MDMA, "ecstacy"), and methamphetamine suggested that self reports among MSM are provided with a high degree of validity. Underreporting was a correlate of social class (education, income, and employment) in the general population but not in the MSM sample. The utility of drug testing was dependent on social class in the general population sample.

CONCLUSIONS: Drug testing is feasible in epidemiological surveys of drug use. Self reports among MSM are at least as valid as those provided by a general population sample of men. In some instances (e.g., cocaine use), they may actually be of higher quality. Although the findings support the merit of epidemiological studies of MSM drug use that have relied completely on self-reporting, drug tests may be useful for clarifying club drug ingestion patterns.