Michelle Roozeboom-Baker, Ph.D.

Michelle Roozeboom-Baker is an applied statistician and health data analytics leader with decades of experience analyzing national Medicare and Medicaid administrative data and partnering with the Centers for Medicare & Medicaid Services (CMS). She leads analytic systems, payment-policy implementation, quality measurement, program monitoring, and evaluation work across Medicare fee-for-service, Medicare Advantage, and Medicaid. Her expertise includes linking claims, enrollment, assessment, registry, and participant-submitted data; designing summary statistics and person-level analyses; and translating complex payment rules into accurate, auditable calculations. She has led multidisciplinary analytic teams and advanced reproducible approaches across major CMS data environments.

At Abt, Roozeboom-Baker directs and supports analytic systems for multiple CMS initiatives. As system and task lead for the MAHA ELEVATE Model, she oversees access to Model Space and defines requirements for recipient and beneficiary portals. She also transitions SAS analyses to Databricks notebooks in the CCW Virtual Research Data Center for home health, hospice, infusion, and inpatient rehabilitation payment-system monitoring. Across additional assignments, she calculates measures from Medicare, Medicaid, registry, and participant-submitted data in CCSQ, Model Space, and VRDC environments, helping teams produce reliable analytic outputs and maintain secure, policy-aligned workflows.

Previously, Roozeboom-Baker served as a principal data analytics leader and senior systems analyst at Mathematica, where she led CMS implementation and evaluation systems, Medicare payment standardization, and a 12-person Data Innovation Lab team using Medicare and Medicaid data. At General Dynamics Information Technology, she managed 12 statisticians supporting national CMS evaluations and analyses, including standardized payment calculations and linked Medicare, Medicaid, assessment, and mortality data. Earlier roles at Buccaneer Computer Systems and Services and the Iowa Foundation for Medical Care combined statistical leadership, database and reporting methodology, software-system development, and communication of technical findings to nontechnical audiences.

Expertise

• Medicare and Medicaid administrative data analytics
• CMS payment policy and payment standardization
• Analytic systems leadership and data environments
• Quality measurement, monitoring, and evaluation
• Statistical programming and reproducible analytics

Key Projects

• MAHA ELEVATE Model Implementation and Monitoring, CMS
• Home Health and Post-Acute Care Payment System Analysis Support and Monitoring, CMS
• Home Health Quality Reporting Program, CMS
• Cell and Gene Therapy Evaluation, CMS
• Nursing Home Compare, CMS

Publications

• Roozeboom, M.A. “Identification of Beneficiaries with Clinical and Chronic Conditions Using Linked Medicare and Medicaid Data.” Presented at the Medicaid Enterprise Systems Conference, Boston, August 2012.  
• Larson, M.D., Roozeboom, M.A, and K. Schneider. “Nonresponse Adjustment Methodology for NHIS-Medicare Linked Data.” Presented at the American Statistical Association Annual Meeting, San Diego, July 2012. 
• Roozeboom, M.A., B.E. O’Donnell, K. Schneider, and G. Lessman. “Medicare and Medicaid Cost and Utilization for New Accretion Dual Beneficiaries.” Presented at the AcademyHealth Annual Research Meeting, Orlando, FL, June 2012. 
• O’Donnell, B.E., Roozeboom, M.A, G. Lessman, G. Martens, and C. Weigel. “Medicare and Medicaid Drug Payments for Medicare Only and Medicare/Medicaid Dual Eligible Populations.” Presented at the American Public Health Association Meeting, Washington, DC, November 2011. 
• Roozeboom, M.A., B.E. O’Donnell, and K. Schneider. “Linking Medicare and MAX Data Files to Examine Health Care Utilization for Dual Eligible Medicare/Medicaid Persons.” Presented at the AcademyHealth Annual Research Meeting, Seattle, June 2011.