Authors:
 Sumedha Chhatre, Ruchika Talwar, Elina Medvedeva, John Cashy, Kyle Robinson, Ravi Jayadevappa
 
 
 Affiliations:
 University of Pennsylvania; Corporal Michael J. Crescenz VA Medical Center
 
 
 Abstract Body:
 "Background: In contrast to the non-Veteran US prostate cancer population, African-American (AA) and White Veterans may have similar outcomes. Clarifying mechanisms underlying improved survival for AA Veterans may provide insights for ameliorating disparities in non-Veteran populations. The objective of this analysis was to examine factors contributing to racial disparities in mortality among men with prostate cancer diagnosed within a large national equal-access health system. 
 
 Methods: In retrospective cohort study using administrative data within the Veterans Health Administration (VA) Corporate Data Warehouse, our cohort consisted of a population-based sample of AA and non-Hispanic White men between ages 21 to 99 diagnosed with high-risk localized prostate cancer between January 1st, 2004 and December 31st, 2013 within the VA. Patients were followed through December 31st, 2019. Data were analyzed from April 1st to August 1st, 2020. The exposure of interest was race (Black vs. White). Our main outcomes included all-cause mortality, adjusted for demographic, cancer-related, treatment, and socioeconomic characteristics and using inverse probability weighting (IPW). We used multivariable proportional hazards regressions to estimate the risk of death associated with race, adjusted for baseline covariates. 
 
 Results: Among 25,997 Veterans receiving continuous VA-based care (median age at diagnosis 66 years [interquartile range [IQR] 61-74]), 7,472 (28.7%) were AA and 21,338 (82.1%) followed continuously within the VA. Median follow-up for Veterans following was 8.4 years (IQR 6.1-11.1). After adjusting for baseline demographic-, cancer-, treatment-, socioeconomic-, facility-, and followup-related factors, AA Veterans (adjusted hazard ratio [aHR] 0.84, 95% confidence interval [CI] 0.80-0.88) had improved overall survival compared to White Veterans. This association was sustained after IPW (HR 0.92, 95% CI 0.88-0.96) and was not impacted by continuity of VA care nor adherence to guideline-based survivorship care. This association was strongest among Veterans living in urban domiciles (aHR 0.82, 95% CI 0.70-0.96). 
 
 Conclusions and Relevance: Improved survival among African-American Veterans does not appear to be mediated by cancer, treatment, facility, or continuity nor quality of followup care. Equal access to care may improve racial disparities in prostate cancer, although the influence of selection bias should be investigated."
 
 
 Funding Acknowledgments:
 VA Center for Health Equity, Research, and Promotion
 
 
 Corresponding Author Email:
 ravi.parikh@pennmedicine.upenn.edu
 
 
 Conflict of Interest:
 Nothing to Disclose

Understanding Racial Disparities in Survival among US Veterans with Localized high-risk Prostate Cancer, 2004-2019