Validity of intravenous drug abusers' self-reported changes in HIV high-risk drug use behaviors

The HIV-1 high-risk drug use behavior of intravenous drug abusers was assessed both retrospectively (for 6 months) and prospectively (for 6 months) via structured interview and urinalysis testing. Subjects were 281 intravenous drug abusers, 146 enrolled in outpatient methadone treatment (Treatment group) and 135 not in treatment (Community group). The Treatment group reported fewer drug injections and less needle sharing and had fewer positive urinalyses for opiates and cocaine than did the Community group. Reported drug injection and needle sharing declined over time, and an increasing proportion of subjects reported abstinence from these behaviors. In contrast to the behavioral reports of subjects, positive urinalyses indicating opiate and/or cocaine use did not decline over time. Almost half (45.8%) of the reported increase in injection abstinence from intake to month six was disconfirmed by urinalysis. In contrast to this large discrepancy regarding reported behavior change, there was good agreement between reported injection abstinence and urinalysis results at single points in time. These data indicate that the validity of the reported HIV-1 risk behavior change of drug abusers may be less than that of reported risk behavior occurrence. The data raise important questions about the validity of reported reductions in high-risk drug use behaviors, and indicate the importance of using biological indicators of HIV-1 risk behavior (such as urinalysis) whenever possible.