The study will examine the impact of Medicare Part D (implemented January 1, 2006) on changes in medications, medications continuity, utilization, cost, and patient-reported outcomes based on an analysis of administrative data (Medicaid and Medicare) for persons with severe mental illnesses (SMI), ages 18-60, disabled, and dually eligible. Their experience and outcomes will be compared with a propensity-score matched group of SMI persons remaining on Medicaid only, in Maryland.
The specific aims are to: (1) Develop a set of individual-level access barrier measures that take account of Part D plan provisions (formulary structure, stepped therapy, copayment levels, quantity limits, and prior authorization requirements) that can restrict access to medications, and that can be used as predictors of Part D plan switching for persons with SMI;(2) Examine the relationship between these measures of access barriers and plan switching by individual beneficiaries;(3) Assess the impact of access barrier measures on medications continuity using data on variations among dual-eligibles as well as pre vs. post Part D implementation experience and the matched comparison group of SMI who retained Medicaid drug coverage;and (4) Assess the impact of these access barrier measures on outcomes of care including utilization-based indicators (mental health hospitalization, crisis care), self-report (symptom control, quality of life, satisfaction), mortality, and costs of care by analyzing variations among dual eligibles, and by comparing the dually eligible SMI with the matched group of Medicaid only beneficiaries. To assess short term and longer term effects, a four and a half year post implementation period will be analyzed. Analyses based on administrative data will be supplemented by analyses that incorporate data from in-person interviews of a sample of 450 dually eligible and a propensity-score matched group of 150 Medicaid-only SMI. The interviews will be conducted in early 2010 and will provide self-reported information on outcomes of care (medication continuity, symptom levels, and quality of life). Interview data will also be obtained on covariates not available in administrative data (e.g., living situation) and on factors and processes relating to individuals'decisions to switch their Part D plan enrollment. The findings from the proposed study, being undertaken in collaboration with the Maryland Mental Hygiene Administration, will provide information specific to dually eligible SMI populations that is relevant to policy-makers, Medicare and Medicaid administrators, Part D pharmacy plans, providers, and advocates for persons with mental illness.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Research Project (R01)
Project #
Application #
Study Section
Mental Health Services in MH Specialty Settings (SRSP)
Program Officer
Rupp, Agnes
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Johns Hopkins University
Public Health & Prev Medicine
Schools of Public Health
United States
Zip Code
Stuart, Elizabeth A; DuGoff, Eva; Abrams, Michael et al. (2013) Estimating causal effects in observational studies using Electronic Health Data: Challenges and (some) solutions. EGEMS (Wash DC) 1: