Serious mental illness, HIV, substance abuse, and antipsychotic use

Objectives: For Medicaid recipients with serious mental illness (SMI) in 45 states, estimate rates of HIV and substance abuse (SA), and describe differences in antipsychotic use (AP). Methods: Naturalistic observational study using Medicaid Analytic Extract files for 2003-04 from all states except AZ, DE, NV, OR, and RI. Subjects were age 18-64, with fee-for-service Medicaid, and no comprehensive HMO coverage. SMI was defined as schizophrenia (SZ), bipolar illness (BP), or major depressive disorder (MDD). Classification was based on one inpatient or two outpatient encounters with a qualifying ICD-9-CM diagnosis Results: Treated prevalence of HIV in 2003-04 was 1.4% for SZ, 1.6% for BP, 2.2% for MDD, and 0.9% for no-SMI. In all groups those with HIV had higher rates than those with no-HIV for: (a) drug abuse/dependence: SZ (50.4% vs. 13.7%); BP (44.1% vs.17.4% ); MDD (37% vs. 11.2% ); and no-SMI (20.1% vs. 2.2% ) and (b) alcohol abuse/dependence: SZ (31.4% vs. 10.7% ); BP (25.8% vs. 12%); MDD (19.7% vs. 8.3%); and no-SMI (9.2% vs. 1.7%). HIV/no-HIV differences in rates of antipsychotic use were generally small (<3 percentage points). For patients with SZ, atypical AP rates were lower for those with HIV (27.4% vs. 33.7%) and for patients with MDD conventional AP rates were higher for those with HIV (41.8% vs. 37.4%). Discussion: Prevalence rates of SMI, HIV and SA were broadly consistent with smaller studies. Efforts to determine how HIV may influence AP use patterns will need to consider high rates of substance abuse.