Policy-makers assume that there is a direct relationship between the way that health care is delivered and the cost of that care. More recently, there have been questions about whether differences in delivery systems also affect patient outcome. We propose to compare three systems of care that reflect naturally occurring variations in the public mental health system in Massachusetts. Three different service system models have evolved: one system has a well developed community-based service model, a neighboring system has retained the traditional state hospital model, and in Boston which has no state hospital, there are five community-mental health centers. The three systems offer an opportunity to explore two central themes that pervade public mental health policy: community-based care, and continuity of care. Each is widely believed to improve clinic outcomes and lower client costs. The framework for the proposed study is the observation that no agreed-upon methodology for defining service systems exists. The question of what constitutes a system is at the heart of the study. For that reason we do not simply assume the ontology of the previously described systems. Instead out analyses are all aimed at shedding light on the meaning of the concept service system. We will collect service use data from a random sample of 425 Medicaid recipients, disabled by virtue of their psychiatric illness to test the effect of different systems on service use and costs. The purpose of our research is to explore four questions. 1) are the opinions of administrators about their system accessibility, quality, adequacy and co-ordination consistent with our a prior descriptions? 2) are the patterns of service use by the clients in our study congruent with the configurations of system-level service expenditures? 3) do client outcomes vary systematically with respect to systems when we have controlled for case-mix? 4) do the treatment costs of clients in the study vary with respect to systems when client case-mix differences are controlled? Studies without random assignment to the conditions under study, must rely on both sampling and statistical methods for controlling the noise in these systems. To reduce the unmeasured variance, we will stratify the sampling frame, and case-mix co variates, and include medical as well as psychiatric service use to capture medical co-morbidity, especially alcohol and drug use. Service use data will come from Medicaid, DMH and from case-manager logs; cost data from Medicaid paid claims, DMH resource inventories, and from DMH vendor contracts.
| Norton, Edward C; Van Houtven, Courtney Harold; Lindrooth, Richard C et al. (2002) Does prospective payment reduce inpatient length of stay? Health Econ 11:377-87 |
| Dickey, B; Normand, S L; Norton, E C et al. (2001) Managed care and children's behavioral health services in Massachusetts. Psychiatr Serv 52:183-8 |
| Dickey, B; Fisher, W; Siegel, C et al. (1997) The cost and outcomes of community-based care for the seriously mentally ill. Health Serv Res 32:599-614 |
| Dickey, B; Azeni, H (1996) Persons with dual diagnoses of substance abuse and major mental illness: their excess costs of psychiatric care. Am J Public Health 86:973-7 |
| Fisher, W H; Dickey, B (1995) Regional variation in service system performance: comparing the perceptions of key stakeholders. J Ment Health Adm 22:68-76 |
| Dickey, B; Azeni, H (1993) Cost-containment in mental health care: the role of utilization review. Adv Health Econ Health Serv Res 14:197-207 |
