Racial differences in survival and healthcare resource utilization among Medicaid-insured adults with metastatic castration-resistant prostate cancer.

29 Background: Prostate cancer (PC) disproportionately affects Black or African American (BAA) men in the United States (US), but racial disparities in outcomes are not well understood. This study analyzed racial disparities in OS, healthcare resource utilization (HRU), and PC treatments in Medicaid-insured metastatic castration resistant PC (mCRPC) patients. Methods: This retrospective longitudinal cohort study evaluated de-identified data from the Centers for Medicare and Medicaid Services 100% Medicaid data files from 01/01/2010 – 12/31/2018. The study included adult patients with a diagnosis of PC and metastasis, and a claim for at least one of the following drugs prior to 9/30/2018, which is specific to mCRPC (date defined index): ≥1 claim for cabazitaxel, mitoxantrone, enzalutamide, radium-223, or sipuleucel-T, ≥1 claim for abiraterone acetate before June 2017, ≥1 claim for docetaxel at least 90 days after initiation of hormone therapy, or evidence of castration resistance. Patients were required to be continuously enrolled for ≥6 months pre-index (i.e., baseline period) and ≥3 months post-index. Outcomes were assessed from the index date to the earliest of end of continuous enrollment, data availability, or death. A multivariable Cox proportional hazards model of OS, and a multivariable Poisson model of HRU were implemented and controlled for age, plan type, region, median state income, residence in a state with Medicaid expansion, index year, Charlson comorbidity index, baseline HRU, baseline PC treatments, and clinical characteristics. Results: The study included 1,095 mCRPC patients (320 [29%] White [W], 278 [25%] BAA, 190 [17%] Hispanic [H], 307 [28%] Other races [O]). H had the highest mean age of 69 years, followed by 68 years for O, and 63 years for both W and BAA. Median unadjusted OS was 46.7 months in H, 40.5 months in BAA, 35.3 months in O, and 32.3 months in W. After adjustment, H had significantly lower risk of death vs. W (hazard ratio [95% confidence interval (CI)]: 0.62 [0.42, 0.90]) and BAA had comparable survival to W (0.85 [0.62, 1.16]). BAA had significantly fewer PC-related outpatient (OP) visits vs. W (adjusted incidence rate ratios [IRR] [95% CI]: 0.67 [0.48, 0.93]), but significantly more PC-related emergency room (ER) visits (5.03 [2.03, 12.47]) per patient per year. Race cohorts differed in the proportions of patients treated with novel hormonal therapy (70% W, 63% BAA, 58% H, 67% O) and chemotherapy use (29% BAA, 26% H, 24% W, 19% O). Conclusions: Among Medicaid-insured adult mCRPC patients, H were more likely to live longer than W patients, while BAA and W patients had similar OS. BAA patients had a higher rate of PC-related ER visits but fewer PC-related OP visits as compared to W, showing differential use of PC-related healthcare resources.