Stephen Leff1, Benjamin Cichocki1, Clifton M Chow1, Chuck Lupton1. 1. Dr. Leff is with the Harvard Medical School Department of Psychiatry at the Cambridge Health Alliance and with the Human Services Research Institute, both in Cambridge, Massachusetts (e-mail: sleff@hsri.org). Dr. Cichocki is also with the Human Services Research Institute. Dr. Chow is with the Department of Economics, Southern New Hampshire University, Manchester. At the time of this research, Mr. Lupton, who is deceased, was with Manila Consulting, McLean, Virginia.
Abstract
OBJECTIVE: The authors evaluated the Substance Abuse and Mental Health Services Administration's mental health transformation state incentive grant program, which provided more than $100 million to nine states to make infrastructure changes designed to improve services and outcomes. METHODS: The authors measured infrastructure changes, service changes, and consumer outcomes in the nine programs. Although the federal program had no logic model, the authors adopted a model that hypothesized positive, but small, correlations between the program elements. RESULTS: There were few statistically significant correlations and a number of negative correlations between infrastructure changes, service changes, and consumer outcomes. CONCLUSIONS: Federal investments should take into account evidence that infrastructure changes alone do not necessarily contribute to better consumer outcomes, support operationally defined infrastructure improvements, require that service improvements accompany infrastructure changes, and provide sufficient resources to oversee grantee behaviors. In addition, future evaluation should support evaluation best practices.
OBJECTIVE: The authors evaluated the Substance Abuse and Mental Health Services Administration's mental health transformation state incentive grant program, which provided more than $100 million to nine states to make infrastructure changes designed to improve services and outcomes. METHODS: The authors measured infrastructure changes, service changes, and consumer outcomes in the nine programs. Although the federal program had no logic model, the authors adopted a model that hypothesized positive, but small, correlations between the program elements. RESULTS: There were few statistically significant correlations and a number of negative correlations between infrastructure changes, service changes, and consumer outcomes. CONCLUSIONS: Federal investments should take into account evidence that infrastructure changes alone do not necessarily contribute to better consumer outcomes, support operationally defined infrastructure improvements, require that service improvements accompany infrastructure changes, and provide sufficient resources to oversee grantee behaviors. In addition, future evaluation should support evaluation best practices.