Representativeness of Medicare Participants in the Jackson Heart Study for African American Medicare Beneficiaries

Abstract

Background The Jackson Heart Study (JHS) assesses cardiovascular disease risk factors among African Americans in Jackson, Mississippi. Whether characteristics of JHS participants differ from those of a broader African American population is unknown. Methods In a retrospective observational analysis, we compared characteristics and outcomes of JHS participants 65 years and older and enrolled in Medicare (n = 1105) to regional (n = 57,489) and national (n = 95,494) cohorts of African American Medicare beneficiaries. We weighted the regional and national cohorts to match the age and sex distributions of the JHS–Medicare cohort for pairwise baseline comparisons. Outcomes of interest included mortality and Medicare costs. We used Cox proportional hazards models to test associations between cohorts and outcomes. Results The JHS–Medicare cohort was younger, included more women, and had fewer beneficiaries with dual Medicare-Medicaid eligibility, compared with regional and national Medicare cohorts. The cohort also had lower risks of stroke, lung disease, heart failure, diabetes, and renal disease. Mean Medicare costs were lower ($5066 [SD, $11,932]) than in the regional ($7419 [SD, $17,574]) and national ($8013 [SD, $19,378]) cohorts. The regional and national cohorts had higher mortality (adjusted hazard ratios, 1.52; 95% CI, 1.31–1.76; and 1.49; 95% CI, 1.29–1.73, respectively). Subgroup analysis for dual Medicare-Medicaid eligibility attenuated mortality differences. Conclusion JHS–Medicare participants had fewer comorbid conditions, better survival, and lower Medicare costs compared with regional and national cohorts. Observed differences may reflect healthy volunteer bias and higher socioeconomic status.