Return to 2001 Abstacts List

99. Mental Disorders in the Veterans Health Administration: The 1999 Health Survey of Veterans

A Spiro III, CHQOER, MAVERIC, & BU SPH; DR Miller, CHQOER, MAVERIC, & BU SPH; A Lee, CHQOER & BU SPH; L Kazis, CHQOER, 10Q, & BU SPH

Objectives: The Veterans Health Administration (VHA) operates the largest mental health care system in the US, providing services to over half a million patients who account for 20% of inpatient episodes and 25% of outpatient episodes. VHA is mandated to provide services for veterans with PTSD, serious mental illness, substance abuse disorders, or who are homeless and mentally ill. We examined the prevalence of selected DSM-IV diagnoses among enrollees in the VHA, and compared them among VISNs (Veterans Integrated Service Networks).

Methods: The 1999 Large Health Survey randomly sampled 1.5 million enrollees from the Spring 1999 enrollee file (n=3.4 million). ICD-9-CM diagnoses were obtained from administrative data from October 1996 through June 1999 and were combined to define selected DSM-IV mental disorders (depression, anxiety, PTSD, schizophrenia, bipolar, alcohol). Prevalence rates for each disorder were computed overall and by VISN for the VHA enrollee population of 3.4 million.

Results: Over one-third (36%) had at least one mental disorder diagnosis; 18% a substance use disorder. Depression was the most common diagnosis, with a prevalence of 15.1% (ranging from 12.0% in VISN 4 to 17.5% in VISN 15). Alcohol disorder had a prevalence of 11.0% (9.4% VISN 4 to 13.3% VISN 20), Anxiety 8.8% (6.5% VISN 12 to 11.2% VISN's 1, 15), PTSD 6.8% (4.0% VISN 14 to 10.1% VISN 1), Schizophrenia 4.6% (3.3% VISN 14 to 5.2% VISN 4), and bipolar disorder 4.0% (2.8%VISN 6 to 5.4% VISN 1).

Conclusions: Mental disorders are prevalent among VHA enrollees, with higher prevalences than reported in national surveys of other populations, which report prevalence of 20-23% for any mental disorder. The prevalence of the disorders varied by VISN, from 30% to 100%. Because the prevalences are based on diagnoses from administrative data, they may under-estimate the true prevalence of disorder, by ignoring undetected and untreated conditions.

Impact: The variation in disorders among VISNs may represent differences in casemix (including selection factors), or in outcomes of care, or may reflect patient characteristics, reporting differences, and selection factors. Regardless, these variations in prevalence of mental disorder have implications for service delivery and resource allocation in the VHA.

Supported by VA Office of Quality and Performance (10Q, and by Health Services Research and Development Service, Veterans Health Administration.)