Our analysis goes further than prior studies to investigate how the specific race-treatment interactions between race impact overall survival in Black, White, and Latino patients by examining a more inclusive study population, but also by the statistical methodology used. During our initial exploration of this dataset, we found that one’s vital status (whether a person was alive or dead at the end of the observation period) was closely correlated with missing data on the length of follow-up. This meant that deceased patients were much more likely to have a missing data on the length of follow-up (92% of patients with missing data died vs 66% of patients with observed values died) with no significant differences with respect to other characteristics. Many studies would simply eliminate cases with missing data, despite the fact that this could significantly bias the findings significantly. To avoid this, we utilized a multiple imputation strategy in which missing values are identified and replaced by a random sample of plausible values imputations. In this study, 10 imputed data sets were used and then averaged for the estimates presented in this paper, strengthening the validity of our findings compared to studies that simply remove cases with missing data.
Our work builds upon prior publications to disaggregate general racial groupings (i.e. Black/White/Latino patients who received treatment vs counterparts that did not) by guideline-based treatment (GBT) received and to estimate the effect size of GBT on the survival benefit conferred by GBT within specific racial/ethnic subgroups. We used broad definitions to provide generous estimates of its utilization, meaning that the ‘true’ utilization of GBT may in fact be lower and subsequent disparities therefore greater than observations from our dataset. Ultimately, GBT was not received by those who needed it most- Black individuals. We found that the application of GBT alone did not completely eliminate the observed racial/ethnic disparities in individuals with MIBC, pointing to other systemic issues and non-clinical factors which may impact care but may not be adequately captured in larger national cancer registries. In order to inform effective interventions to identify structural and individual barriers to GBT, more nuanced research that includes diverse populations incorporates qualitative explorations into patient experiences and considers the social and structural determinants that impact health is required to mitigate race-based survival differences observed in MIBC.
Written by: Samuel L Washington, MD, MAS, Steven E Gregorich, PhD, Maxwell V Meng, MD, Anne M Suskind, MD, MS, Sima P Porten, MD, MPH
Department of Urology, University of California, San Francisco, CA, USA., School of Medicine, University of California, San Francisco, CA, USA.
Read the Abstract