Vulnerability and Health Care Service Use Among Homeless Persons with Psychiatric and Co-Occurring Disorders

Abstract

This study examined the use of medical services by homeless people with psychiatric and co-occurring disorders who enrolled in the federal ACCESS (Access to Community Care and Effective Services and Supports) demonstration project. In a secondary analysis of the ACCESS database, the largest and most geographically representative database that exists on homeless persons with psychiatric disorders in the United States, the use of outpatient and emergency medical services by project participants was studied. This study used the Gelberg-Andersen Behavioral Model for Vulnerable Populations as a conceptual model to examine the impact that traditional (e.g., gender, education, social supports) and vulnerable (e.g., psychiatric symptoms, residential status, criminal behavior) domain variables have on use of outpatient and emergency medical services, the dependent variables. Separate composite scores were created for traditional and vulnerable domain variables and used as independent variables. Analysis of variance and logistic regression analyses were used to examine the relationship between traditional and vulnerable domain composite scores and type of medical services used, either outpatient or emergency. The study hypothesized that traditional domain composite scores would be positively associated with use of outpatient, or discretionary, medical services and negatively associated with use of emergency, or non-discretionary, medical services. Conversely, it was expected that vulnerable domain composite scores would be negatively associated with use of outpatient medical services and positively associated with use of emergency medical services. The results of the study were mixed, as it was found that both traditional and vulnerable composite scores were important predictors in use of outpatient and emergency medical services. Additional analyses of the data revealed several notable findings, including that higher vulnerable predisposing composite scores predicted higher use of emergency medical services when compared to use of outpatient medical services. This study provides empirical evidence that vulnerable domain variables are important factors in understanding use of outpatient and emergency medical care services by homeless people with psychiatric and co-occurring disorders. Given the high rate of medical co-morbidity among this population, innovative efforts to promote their use of outpatient and preventative medical services and reduce their use of emergency medical care are of critical importance.