Racial/Ethnic Disparities in Medicare Part D Experiences

Abstract

OBJECTIVE: To examine racial/ethnic differences in Medicare beneficiary experiences with Medicare Part D prescription drug (PD) coverage. DATA SOURCES/STUDY SETTING: 2008 Consumer Assessment of Health Care Providers and Systems survey of U.S. Medicare beneficiaries. STUDY DESIGN: Surveys were administered by mail with phone follow-up to a nationally representative sample (61% response rate). This study examines 201,496 beneficiaries of age 65 and older with PD coverage [6% Hispanic, 7% non-Hispanic Black, 3% non-Hispanic Asian or Pacific Islander (API)]. Key variables are self-reported race/ethnicity and Consumer Assessment of Health Care Providers and Systems getting information and needed PDs measures. DATA COLLECTION/EXTRACTION METHODS: We generated weighted case-mix adjusted means for 4 racial/ethnic groups and for Hispanics separately by English-language or Spanish-language preference status. We calculated within-plan disparities using a linear mixed-effect model, with fixed effects for race/ethnicity, coverage type and case-mix variables, and random effects for contract and contract by race/ethnicity interactions. PRINCIPAL FINDINGS: Disparities for Hispanic, Black, and API beneficiaries on obtaining needed PDs and information regarding coverage range from -2 to -11 points (0-100 scale) relative to non-Hispanic Whites, with the greatest disparities observed for Spanish-preferring Hispanics and API beneficiaries, especially those with low income. There is wide variation in disparities across contracts, and contracts with the largest disparities for Hispanics have higher proportions of beneficiaries with lower education and income. CONCLUSIONS: Quality improvement efforts may be needed to reduce racial/ethnic disparities in beneficiary experience with PD coverage. Cultural, language, and health literacy barriers in navigating Medicare's Part D program may partially explain the observed disparities.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 46 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.