There are two main objectives of this proposal: 1) to develop a mental health care disparities index to be used in state mental health care report cards;and 2) to disseminate the results of the index using web-enabled interactive maps and receive initial feedback on the presentation and results from state policymakers. There is a need for developing a state-based mental health care disparities index to improve translation from research to policy with a concise, readily useable tool for assessing and tracking disparities. However, difficulties arise in reducing multiple disparities measures into a single number, including identifying the optimal weighting of disparities indicators and the choice of thresholds determining states'success in reducing disparities. We first use information from qualitative interviews of policymakers, advocates and clinicians to elicit the most salient constructs and variables to be included in the index. We then develop an index using modern test theory methods (confirmatory factor analysis and latent variable modeling) that combines six indicators of mental health care initiation, expenditure, and quality disparities while optimizing the internal and external validity of the index. We use Bayesian small area estimation and hierarchical modeling to overcome small sample sizes and nested data using restricted and public use data from the 2004-2009 Medical Expenditure Panel Survey. To ensure wide dissemination and use, we present results using interactive maps that allow for the presentation and manipulation of state-level disparities index component data. Feedback from state policymakers will be elicited to better understand their reactions to the presentation and the results of the index and their interpretation of the mechanisms underlying the mental health care disparities climate in their state.
In this project, we develop a disparities index to create a state-by-state mental health care disparities report card.
We aim to develop an index/report card that is salient and has face validity for policymakers and advocates, while being backed by rigorous quantitative and qualitative methods. The methods used to develop the report card are small area estimation and modern test theory methods enhanced by key informant interviews. For the second part of the proposed project, we focus on the dissemination of the index/report card, using a web-enabled geographic information systems setup to present information in an easily accessible and interactive manner. The results and presentation of the index/report card will be shown to state mental health directors for feedback and to inform future directions in the development of the index and in understanding policy mechanisms underlying mental health care disparities.
Showing the most recent 10 out of 20 publications
