Developing the Evidence Base for Mental Health Policy and Services: Inquiries into Epidemiology, Cost-Benefits, and Utilization

The overarching aim of this dissertation is to use health services research methods to address three problems in behavioral health services. This dissertation seeks to address the knowledge gaps in behavioral health services through the generation of evidence intended to support evidence-based practices (EBP). Previous work has examined epidemiology of behavioral health disorders in the ED, but they have not attempted to examine disorders by the cause of injury. Chapter 2 examines the epidemiology of psychiatric disorders among adults who seek care in the emergency department (ED) by cause of injury. Data from a national hospital discharge survey was analyzed using logistic and multinomial regression. Estimates are given as average marginal effects (AME) to simplify the interpretation and application. Intentionally-caused injury and undetermined cause of injury are significantly associated with psychiatric disorders. Patients with undetermined cause of injury were more likely to be diagnosed with anxiety disorders, depressed mood, and psychoses relative to patients with unintentional injuries Since there are several treatment options for obsessive-compulsive disorder (OCD), including cognitive behavioral therapy (CBT), serotonin reuptake inhibitors (SRIs), and combinations of these, a comparison of treatment effects denominated in dollars is helpful when comparing risks and benefits. Chapter 3 builds on previous randomized control trials of treatments for OCD in children and adolescents by ranks the cost-benefits of first-line treatments. The analysis aggregates treatment effects from published trials in meta-analytic framework and a Monte Carlo simulation of 100,000 hypothetic children and adolescents to derive ranked cost-benefit. Treatments strategies starting with CBT, but not CBT and SRIs concurrently, were the most cost-beneficial. The relationship between cost-sharing and utilization of behavioral health services has been studied in the aggregate, but there has been little work examining the relationship by disorder and treatment modality. The aim of Chapter 4 is to examine the association between cost-sharing and utilization of psychotherapy and adherence to pharmacotherapy among insured adults with OCD. This chapter utilizes the Truven MarketScan Commercial Claims and Encounters dataset to perform zero-inflated negative binomial regression and logistic regression analyses. Increased cost-sharing was significantly, negatively associated with psychotherapy intensity and dose, but not associated with SRI adherence. This dissertation examined three different research questions to address gaps in the behavioral health services research. The findings of these chapters have implications for patients, clinicians, insurers, and policymakers. The results can be used to improve aspects of cost, quality, access, and efficiency of behavioral health services