Prevalence of DSM-IV-Defined Mood, Anxiety, and Substance Use Disorders in an HIV Clinic in the Southeastern United States

BACKGROUND: Mood and anxiety disorders, particularly depression, and substance abuse (SA) commonly co-occur with HIV infection. Appropriate policy and program planning require accurate prevalence estimates. Yet most estimates are based on screening instruments, which are likely to overstate true prevalence. SETTING: Large academic medical center in Southeast. PARTICIPANTS: A total of 1,125 patients, representing 80% of HIV-positive patients seen over a 2.5-year period, completed the Substance Abuse-Mental Illness Symptoms Screener, a brief screening instrument for probable mood, anxiety, and SA disorders. Separately, 148 participants in a validation study completed the Substance Abuse-Mental Illness Symptoms Screener and a reference standard diagnostic tool, the Structured Clinical Interview for DSM-IV. METHODS: Using the validation study sample, we developed logistic regression models to predict any Structured Clinical Interview for DSM-IV mood/anxiety disorder, any SA, and certain specific diagnoses. Explanatory variables included sociodemographic and clinical information and responses to Substance Abuse-Mental Illness Symptoms Screener questions. We applied coefficients from these models to the full clinic sample to obtain 12-month clinic-wide diagnosis prevalence estimates. RESULTS: We estimate that in the preceding year, 39% of clinic patients had a mood/anxiety diagnosis and 21% had an SA diagnosis, including 8% with both. Of patients with a mood/anxiety diagnosis, 76% had clinically relevant depression and 11% had posttraumatic stress disorder. CONCLUSIONS: The burden of psychiatric disorders in this mixed urban and rural clinic population in the southeastern United States is comparable to that reported from other HIV-positive populations and significantly exceeds general population estimates. Because psychiatric disorders have important implications for clinical management of HIV/AIDS, these results suggest the potential benefit of routine integration of mental health identification and treatment into HIV service sites.