Social determinants of health and late mortality among allogeneic blood or marrow transplant (BMT) survivors: A BMTSS report

Abstract Introduction: Conditional on surviving the first 2y after BMT, allogeneic BMT recipients remain at a 9-fold greater risk of late mortality compared to the general population (Bhatia, JAMA Onc 2021). Both demographic (e.g., increasing age at BMT, male sex) and clinical factors (e.g., chronic graft vs. host disease [cGvHD]) are associated with late mortality. Exposure to adverse area-level social determinants of health (SDoH) increases mortality risk in the general population but remains unaddressed in BMT recipients, especially in the context of individual-level sociodemographics. We leveraged the BMTSS to address this gap in knowledge. Methods: BMTSS includes individuals who received BMT at one of three transplant centers between 1974 and 2014 and survived ≥2y after BMT. For the current report, we included allogeneic BMT recipients who had completed the BMTSS survey (with a valid residential address) at age >21y. Clinical data (primary diagnosis, age at BMT, stem cell and donor source, and conditioning regimens and intensity) were abstracted from medical records. Survivors reported demographic characteristics (sex, insurance, annual household income), history of cGvHD and presence of chronic health conditions (CTCAE grade 1-4). Vital status after survey completion (and cause of death among those deceased) was ascertained via linkage with National Death Index. Residential addresses were linked with the following area-level SDoH domains: socioeconomic vulnerability (Social Vulnerability Index [SVI] theme 1), health vulnerability (Health Vulnerability Module [HVM] of Environmental Justice Index [EJI], identifying neighborhoods with >66% of population living with severe/life-threatening chronic health conditions), and environmental pollution (Environmental Burden Module of EJI). Primary outcome of interest was all-cause mortality, and the primary exposures were annual household income (individual-level) and area-level SDoH. Among survivors stratified by individual level income (low and high), we examined the association between cumulative burden of SDoH and all-cause mortality. Results: The cohort consisted of 2,233 allogeneic BMT recipients; 55% male; median age at BMT: 36y; BMT between 2005-2014: 46%. The study participants had completed the BMTSS survey a median of 10y after BMT (range, 2-46) and 483 died a median of 4.9y (range, 0-22.1) after completing the survey. The 10y cumulative incidence of all-cause mortality following survey completion was 21.5% (95% confidence interval [95%CI] =19.4%-23.5%). After adjusting for age at BMT, sex, year of BMT, insurance, primary diagnosis, conditioning intensity and pre-existing chronic health conditions, the following SDoH domains were independently associated with all-cause mortality: individual-level (annual household income: ≤$50,000: hazard ratio [HR]=1.4, 95%CI=1.1-1.7, P=0.003, ref: >$50,000) and area-level (socioeconomic vulnerability: most vulnerable: HR=1.2, 95%CI=1.0-1.5, P=0.05, ref: least vulnerable; health vulnerability: most vulnerable: HR=1.4, 95%CI=1.2-1.8, P=0.001, ref: least vulnerable, and environmental pollution: most polluted: HR=1.5, 95%CI=1.2-1.9, P=0.002, ref: least polluted). Exposure to increasing number of adverse SDoH domains was associated with greater hazard of all-cause mortality among survivors who reported annual income ≤$50,000 (2 SDoH domains: HR=1.6, 95%CI=1.1-2.4, P=0.02; 3 SDoH domains: HR=1.8, 95%CI=1.2-2.6, P=0.006; ref: ≤1 SDoH domain) as well as among those who reported annual income >$50,000 (2 SDoH domains: HR=1.7, 95%CI=1.3-2.3, P<0.001; 3 SDoH domains: HR=2.2, 95%CI=1.5-3.3, P<0.001; ref: ≤1 SDoH domain). Conclusions: We show that both individual-level socioeconomic status, and area level factors including residence in poor neighborhoods with high levels of disease and environmental pollution contribute to late mortality in allogeneic BMT recipients above and beyond the known clinical predictors of mortality. Identifying the vulnerable populations and instituting targeted interventions are needed to reduce the risk of late mortality.