The cost of high-fidelity supported employment programs for people with severe mental illness

OBJECTIVE: This study determined the costs of evidence-based supported employment programs in real-world settings. METHODS: A convenience sample of 12 supported employment programs known to follow closely the principles of evidence-based supported employment was asked to provide detailed information on program costs, use, and staffing. Program fidelity was assessed by using the Supported Employment Fidelity Scale. Cost and utilization data were analyzed in a comparable manner to yield direct and total costs per client served, per full-year-equivalent client, and per employment specialist. RESULTS: Usable data were obtained from seven programs in rural and urban locations in seven states: Indiana, Kansas, Massachusetts, New Hampshire, Oregon, Rhode Island, and Vermont. All programs received high fidelity ratings, ranging from 70 to the maximum value of 75. Annual direct costs per client served varied from dollars 860 in New Hampshire to dollars 2723 in Oregon, and direct costs per full-year-equivalent client varied from dollars 1423 in Massachusetts to dollars 6793 in Indiana. Direct costs per employment specialist did not show as much variation, ranging from dollars 37339 in Rhode Island to dollars 49603 in Massachusetts, with a mean of dollars 44082. Differences in cost per client arose in part from differences in rules for determining who is or is not considered to be on a program's caseload. By assuming a typical caseload of about 18 clients, it was estimated that the cost per full-year-equivalent client averaged dollars 2449 per year, ranging from dollars 2074 to dollars 2756. CONCLUSIONS: The results point to the need for greater uniformity in caseload measurement and help specify the costs of high-fidelity supported employment programs in real-world settings.