Does managed care affect the diffusion of psychotropic medications?

Abstract

Newer technologies to treat many mental illnesses have shown substantial heterogeneity in diffusion rates across states. In this paper, I investigate whether variation in the level of managed care penetration is associated with changes in state-level diffusion of three newer classes of psychotropic medications in fee-for-service Medicaid programs from 1991 to 2005. Three different types of managed care programs are examined: capitated managed care, any type of managed care and behavioral health carve-outs. A fourth-order polynomial fixed effect regression model is used to model the diffusion path of newer antidepressant and antipsychotic medications controlling for time-varying state characteristics. Substantial differences are found in the diffusion paths by the degree of managed care use in each state Medicaid program. The largest effect is seen through spillover effects of capitated managed care programs; states with greater capitated managed care have greater initial shares of newer psychotropic medications. The influence of carve-outs and of all types of managed care combined on the diffusion path was modest.

Figure 4. The Predicted Retraction Path of Older Psychotropic by Level of Capitated Managed Care Enrollment in State

Figure 4a: Non-SSRI or SNRI Antidepressants Figure 4b: First generation antipsychotics

Figure 1. The Predicted Diffusion Path of New Psychotropic by Level of Capitated Managed Care Enrollment in State

Figure 1a: SSRI diffusion Figure 1b: SNRI Diffusion Figure 1c: SGA Diffusion

Figure 2. The Predicted Diffusion Path of Second Generation antipsychotics by enrollment in any type of Managed Care

Figure 3. The Predicted Diffusion Path of New Psychotropic Medications by Carve-out status

Figure 3a: SSRI diffusion Figure 3b: SNRI Diffusion Figure 3c: SGA Diffusion