Medical-Surgical Hospitalization Among Veterans With Psychiatric and Substance Use Disorders

Abstract

Mental illness is associated with an increased risk for medical hospitalizations. This study investigates the degree to which nonpsychiatric factors account for these hospitalizations. Using National Veterans Health Administration (VHA) fiscal year 2012 data for 2 million veterans under the age of 60 years, hospitalization risks were compared for veterans with and without mental illnesses. Bivariate analyses identified factors associated with mental illnesses. Multiple logistic regression was used to calculate adjusted psychiatric risk for medical hospitalization, controlling for these factors. Veterans carrying mental health diagnoses were at increased risk for hospitalizations (odds ratio [OR]=2.52, 2.48-2.55). Among individual diagnoses, alcohol use disorder (AUD) (OR=3.84, 3.78-3.91) and drug use disorders (OR=4.58, 4.50-4.66) were associated with the highest risk. After adjusting for nonpsychiatric medical, addiction-related, and care utilization factors and the useof outpatient medical services, veterans with mental illnesses were at increased risk for medical hospitalization (OR=1.43, 1.41-1.45). After further adjustment for AUD and drug use, hospitalization riskdecreased further (OR=1.23, 1.21-1.26) while the association of AUD and hospitalizations remained high (OR=1.77, 1.73-1.81). Medical comorbidities and service use accounted for most, but not all, of the increased risk of medical hospitalizations associated with mental illness. Even afteraccounting for poor health, AUDremained strongly associated with medical hospitalization.