Using Propensity Score Matching Techniques to Establish Treatment/Control Groups to Assess Medicare/Medicaid Service Use

Abstract

At the AcademyHealth Annual Research Meeting (ARM), Hilltop Senior Policy Analyst Karen E. Johnson, MS, gave this poster presentation, which was also authored by Hilltop Director of Special Research and Development Anthony M. Tucker, PhD, and UMBC Assistant Professor Yi Huang, PhD. The poster described the use of a propensity score matching methodology to identify comparison groups among Medicare-Medicaid beneficiaries who received Medicaid-paid long-term services and supports (LTSS) via home and community-based (HCBS) waivers versus those who did not receive LTSS. This matching technique was used to establish comparable treatment/control pairs for subsequent analysis of cross-payer effects of providing Medicaid-paid LTSS on Medicare acute care resource use, and could be used more generally to strengthen policy analyses that are based on observational and/or administrative data.