Nearly half of HIV-infected persons in the U. S. have and 70% need mental health care. However, no one has comprehensively examined mental health service utilization, service intensity received, or satisfaction with care among HIV-positive persons nationally. National studies of HIV-positive persons have only simply examined predictors of mental health utilization and of type of service provider utilized. This study applies Andersen?s Behavioral Model of Health Services Use (BMHS) to examine individual, facility/provider, and environmental predictors among a cross-sectional, nationally representative sample of HIV-positive adults. These adults were in regular medical care for HIV and either perceived a need for mental health care or met criteria for psychiatric diagnoses (n=l046). The BMHS is a widely used multi-level framework of health service utilization with three levels of factors: environmental, provider/facility, and individual. The BMHS will be applied in a novel manner by specifying which level of factors has the greater effect on four outcomes: Individual-level factors are hypothesized to explain more variance than provider/facility-or environmental-level factors in models predicting receiving any mental health service and predicting clients? satisfaction with the provider used. Provider/facility-level factors are hypothesized to explain more variance than the other two levels of factors in models predicting receiving and intensity of: mental health visits, psychiatric hospitalizations, and psychopharmacological medications. Individual predictors will be estimated for each outcome. AHRQ data used in this dissertation are from the HIV Cost and Services Utilization Study (HCSUS) conducted in 1996-1997. This is the first survey to randomly select and interview a nationally representative sample of adults in medical care for HIV. Multivariate regressions will test hypotheses, and hierarchical linear analysis will be used to determine predictors at the three nested levels of data. If these hypotheses are correct, applying the principle demonstrated in this study will maximize the utility of Andersen?s model. In addition, this multi-level approach to the examination of service utilization will be an improvement over single-level approaches and will aid in the design of multi-level based interventions. Most of these national-level findings will be the first.
