Racial Differences in 1-Year Mortality after Hospitalization for Chronic Obstructive Pulmonary Disease in the United States

Rationale: One-quarter of Medicare beneficiaries hospitalized for chronic obstructive pulmonary disease die within a year. While overall mortality is higher among chronic obstructive pulmonary disease patients of White race, racial and ethnic differences in the vulnerable period following hospitalization are unknown. Objectives: To determine the association between race and ethnicity and mortality following chronic obstructive pulmonary disease hospitalization and to evaluate the extent to which differences are explained by clinical, geographic, socioeconomic, and post-acute care factors among Medicare beneficiaries in the United States. Methods: In this retrospective cohort study of Medicare beneficiaries hospitalized for chronic obstructive pulmonary disease exacerbation, we constructed cox regression models for 1-year mortality accounting for hospital-level clustering, sequentially adjusting for clinical, geographic, neighborhood socioeconomic, and post-acute care characteristics, and stratifying by sex and individual socioeconomic status. Results: Among 244,624 hospitalizations, Medicare beneficiaries of racial and ethnic minority groups had a lower risk of dying within 1-year of hospitalization compared to those of White race [Hazard Ratio (95% CI):0.78(0.75,0.80) for Black patients, 0.79(0.76,0.82) for Hispanic patients, and 0.82(0.77,0.86) for others]. Differences in visits to physicians, attendance of pulmonary rehabilitation, and discharge disposition explained some of the mortality gap among dual-eligible but not among non-dual-eligible beneficiaries. Conclusions: Medicare beneficiaries of White race are at greater risk for mortality following chronic obstructive pulmonary disease hospitalization compared to beneficiaries of minority race and ethnicity groups. Our findings should be interpreted in the context of selection of a hospitalized population and potentially incomplete assessment of illness severity in administrative data and warrant further investigation