Medicaid Formulary Restrictions and Outcomes for Patients with Mental Illness

States have increasingly adopted restrictive formulary policies targeting drugs that are specifically designed to treat patients with mental illness, including schizophrenia, bipolar disorder and major depressive disorder (MDD).  These policies, such as step therapy and prior authorization, are designed to cut costs by limiting patient access to newer and more expensive branded pharmaceuticals. These policies are designed to reduce states’ overall spending on prescription drugs by Medicaid patients, but they fail to account for the possibility that these policies can have negative outcomes for patients.  Restrictions on pharmaceutical availability could lead to delays in finding the most appropriate treatment, disruptions in therapy, or declines in medication adherence.  In the case of mental illness, treatment must often be individualized through a trial-and-error approach with different drugs, suggesting that these patients could be particularly sensitive to treatment delays or disruptions. If the reactions are strong enough, this could offset part or all of the cost savings that result from the policies. 

This study uses a unique database of information on formulary restrictions on antidepressants and antipsychotics in Medicaid combined with claims data to understand the impact of formulary restrictions on the healthcare utilization and spending for patients with mental illness.  The data include the medical and pharmacy claims for all patients with severe mental illness—defined as schizophrenia, bipolar disorder, or MDD—from 24 state Medicaid programs from 2001-2008.  Key outcome variables include mental health-related and all-cause hospitalizations, ER visits, initiation of antidepressant therapy and inpatient, pharmacy and total spending.  These data are linked to information on prior authorization and step therapy for antidepressants and antipsychotics in Medicaid by state and year.  In addition to studying the effect of the policies on utilization and spending, we link the data on formulary restrictions to the American Community Survey to consider other effects on productivity by examining the impact of pharmacy restrictions on employment, productivity and the need for public assistance (e.g., SSDI or TANF).  We use difference-in-differences to identify the effect of the policies on key outcomes.  Preliminary findings suggest that formulary restrictions have adverse effects on patient outcomes, and appear to generate little in the way of aggregate cost savings.