Social determinants of health and outcomes of bone marrow transplantation in patients with aplastic anemia: A 23-year retrospective cohort study

Abstract Introduction: Bone marrow transplantation (BMT) outcomes in patients with aplastic anemia (AA) have significantly improved during the last few decades. Whether these outcomes are impacted by patients' socio-economic variables is unknown. Objective: To evaluate the impact of zip code–level social determinants of health (SDOH) on BMT outcomes in patients with AA, including chronic graft-versus-host disease (cGVHD) incidence and mortality. Methods: We conducted a retrospective chart review of patients diagnosed with AA who underwent either matched related or unrelated BMT at a tertiary cancer center in the Northwestern United States between 2001 and 2024. Residential zip codes documented at the time of BMT were used to link each patient with zip code-level SDOH data, extracted from the U.S. Census Bureau's American Community Survey. The SDOH variables included median household income, proportion of individuals below the poverty line, food stamp usage, proportion of individuals receiving Supplemental Security Income (SSI), proportion without a high school education, housing cost to income ratio, and access to transportation and communication (e.g., vehicle and landline or cellphone availability). We used multivariable logistic regression to evaluate the association of SDOH variables and binary outcomes, including mortality and development of cGVHD. Cox proportional hazards models were used to account for the timing of events. Odds ratios (ORs) or hazard ratios (HRs) were reported with their 95% confidence intervals, adjusted for donor type, recipient age, sex, and total nucleated cell dose. Statistical analyses were conducted using R (version 4.2.3). Results: The study included 154 BMT recipients. The median age was significantly lower among survivors (25 years, SD: 18) compared to non-survivors (44 years, SD: 22; p < 0.001). In time-to-event analysis, higher proportions of SSI use (HR: 2.39; 95% CI: 1.23–4.67) were associated with a more than two-fold increase in the hazard of cGVHD, while lack of telephone access (HR: 1.61; 95% CI: 0.82-3.14) was associated with 60% increase in the hazard of cGVHD. Similarly, proportions of SSI use (HR: 1.32, 95% CI: 0.94-1.87) and proportions without a telephone (HR: 1.40, 95% CI: 1.02-1.93) were associated with greater chance of all-cause mortality. In logistic regression, lack of telephone access was associated with a 56% increase in the odds of death (OR: 1.56; 95% CI: 1.06–2.29), as did higher proportions of SSI use associate with increased mortality (OR: 1.38 95% CI: 0.86-2.20). Several other SDOH did not show a strong relationship with BMT outcomes, including percent below the poverty line (HR for cGVHD: 1.00; 95% CI: 0.89–1.14; HR for mortality: 0.99; 95% CI: 0.90–1.08), median income (HR for cGVHD: 1.00; 95% CI: 1.00–1.00; HR for mortality: 1.00; 95% CI: 1.00–1.00), housing cost to income ratio (HR for cGVHD: 0.99; 95% CI: 0.91–1.09; HR for mortality: 1.02; 95% CI: 0.94–1.11), proportion with less than a high school education (HR for cGVHD: 0.95; 95% CI: 0.84-1.07; HR for mortality: 1.01; 95% CI: 0.96–1.07), and proportion without vehicle access (HR for cGVHD: 0.89; 95% CI: 0.68-1.19; HR for mortality: 1.02; 95% CI: 0.95–1.10). Conclusion: This is the first study to demonstrate that percentages of lack of telephone access and SSI usage were associated with increased risks of both cGVHD and mortality after BMT for AA. These findings suggest that structural barriers linked to place of residence may impact transplant outcomes for AA. Further studies and methods to reverse these negative impacts are needed.