Accurate analysis of race/ethnicity and socioeconomic status

TO THE EDITOR:

We read with great interest the study by Farhadfar et al¹ that investigated graft-versus-host disease outcome disparities using Center for International Blood and Marrow Transplant Research (CIBMTR) data. We commend the authors for their examination of this critically important, highly complex subject, and herein share considerations relevant to this and future studies in this field.

First, it was not stated whether donor type, a confounder in the relationships between patient race/ethnicity,² socioeconomic status (SES),³ and transplant outcomes, was included in the multivariable analyses. Because patients from racial/ethnic groups other than White non-Hispanic received more than twofold the proportion of HLA-mismatched donors, the significant outcome differences reported could be in part due to differences in donor type.

Second, the authors included both neighborhood poverty and insurance status in their multivariable analyses of chronic graft-versus-host disease patient outcomes. However, because different SES measures classify different patient subsets as having low SES,³ using a composite SES variable incorporating multiple measures would be appropriate; for example, classifying patients as low SES if in the lowest bracket by neighborhood poverty, or if on Medicaid or without insurance. This would help to identify financially vulnerable patients mistakenly classified as “higher SES” by the neighborhood poverty index.

Third, additional patient and donor characteristics that would be important to incorporate into multivariable analyses include graft characteristics (e.g., donor age) and time to transplantation. These variables are relevant given that non-European ancestry patients (especially those of African or non-Black Hispanic heritage) are more likely to receive suboptimal grafts² and have delayed transplantation⁴ due to unrelated donor unavailability⁵ and/or structural racism.

Finally, we emphasize that prospectively collected, comprehensive, granular data concerning the social determinants of health are required to accurately investigate and address disparities. Studies piloting such data collection are underway (eg, ClinicalTrials.gov identifier: NCT06431347). Multicenter adoption with CIBMTR data sharing will permit large-scale outcome analyses to further interrogate health disparities in transplant and cellular therapy.

Contribution: W.B.F. designed and wrote the manuscript; A.S. and J.N.B. reviewed the manuscript and provided critical feedback; and all authors approved the final version.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Warren B. Fingrut, The University of Texas MD Anderson Cancer Center, 1515 Holcolmbe Blvd, Houston, TX, 77030; email: wbfingrut@mdanderson.org.