1528 Background: Access to allogeneic hematopoietic cell transplantation (alloHCT) is limited by many factors, with only 1 in 3 patients who may need a transplant receiving one. This study examined associations between social vulnerability index (SVI), physician density (PD), and alloHCT unmet need across the United States (U.S.) to identify geographic areas that may be most at-risk of access challenges. Methods: A retrospective analysis aggregated public data across 3,141 U.S. counties, including: SVI [Centers for Disease Control and Prevention/Agency for Toxic Substances and Disease Registry]; National Provider Identifier [Center for Medicare and Medicaid Services] data for all registered hematology-oncology and HCT physicians; and 2017 – 2021 alloHCT unmet need [Center for International Blood and Marrow Transplant Research (CIBMTR), NMDP]. PD was defined as providers per 10,000 population. All variables were collapsed into ordinal categories (very low, low, moderate, high) based on quartiles. Counties with moderate/high SVI and unmet need and low/very low PD were defined as at-risk. By state, linear regression and mediation analysis assessed exploratory relationships. Results: Among 396 counties defined as at-risk, top states are shown (Table). State (n=51) percent unmet need was correlated with SVI (r=0.494, p<0.001) and PD (r=-0.295, p=0.036). Univariate analysis revealed SVI (b=-0.223, p<0.001) and PD (b=-0.187, p=0.036) as predictors of unmet need. However, multivariate regression showed only SVI as significant when included with PD. Race/ethnicity as 1 of 4 SVI subthemes (socioeconomic status [SES], household characteristics [including age], race/ethnicity, and housing/transportation) was not correlated with PD, but all other themes were. SES and household characteristics were found to fully mediate the relationship between PD and unmet need. Conclusions: This study can help prioritize targeted initiatives through a better understanding of how SVI, PD, and unmet need relate to alloHCT access. This study cannot draw causal conclusions and uses multiple population-level data sources. However, data can examine trends on a regional level within at-risk states to allocate resources to overcome barriers to alloHCT, with special attention on SES and household factors like age. Relationships between variables should be further examined. [Table: see text]
Read full abstract