Racial Equity in Living Donor Kidney Transplant Centers, 2008-2018

Abstract

Importance: It is unclear whether center-level factors are associated with racial equity in living donor kidney transplant (LDKT).

Objective: To evaluate center-level factors and racial equity in LDKT during an 11-year time period.

Design, setting, and participants: A retrospective cohort longitudinal study was completed in February 2023, of US transplant centers with at least 12 annual LDKTs from January 1, 2008, to December 31, 2018, identified in the Health Resources Services Administration database and linked to the US Renal Data System and the Scientific Registry of Transplant Recipients.

Main outcomes and measures: Observed and model-based estimated Black-White mean LDKT rate ratios (RRs), where an RR of 1 indicates racial equity and values less than 1 indicate a lower rate of LDKT of Black patients compared with White patients. Estimated yearly best-case center-specific LDKT RRs between Black and White individuals, where modifiable center characteristics were set to values that would facilitate access to LDKT.

Results: The final cohorts of patients included 394 625 waitlisted adults, of whom 33.1% were Black and 66.9% were White, and 57 222 adult LDKT recipients, of whom 14.1% were Black and 85.9% were White. Among 89 transplant centers, estimated yearly center-level RRs between Black and White individuals accounting for center and population characteristics ranged from 0.0557 in 2008 to 0.771 in 2018. The yearly median RRs ranged from 0.216 in 2016 to 0.285 in 2010. Model-based estimations for the hypothetical best-case scenario resulted in little change in the minimum RR (from 0.0557 to 0.0549), but a greater positive shift in the maximum RR from 0.771 to 0.895. Relative to the observed 582 LDKT in Black patients and 3837 in White patients, the 2018 hypothetical model estimated an increase of 423 (a 72.7% increase) LDKTs for Black patients and of 1838 (a 47.9% increase) LDKTs for White patients.

Conclusions and relevance: In this cohort study of patients with kidney failure, no substantial improvement occurred over time either in the observed or the covariate-adjusted estimated RRs. Under the best-case hypothetical estimations, modifying centers' participation in the paired exchange and voucher programs and increased access to public insurance may contribute to improved racial equity in LDKT. Additional work is needed to identify center-level and program-specific strategies to improve racial equity in access to LDKT.

Figure 1.. Observed Association Between Live Donor Kidney Transplant (LDKT) Black-White Patient Rate Ratio and Transplant Center Volume

Total transplant volume (A), LDKT volume (B), and total percentage LDKT volume of total transplant volume (C). Blue color indicates the population-weighted median of percentage Black population among all census tracts or block groups in each center’s transplant referral region (TRR). Each data point represents a specific transplant center for a single year in the study period. Only centers that performed at least 12 LDKTs in all years of the study period are included. Data points with a y-axis value of 0 are jittered vertically to aid visualization. Although the y-axis reaches 2.5, the data points were truncated at 2.1 to aid in visualization, which excluded a single outlying point (center year) with a crude LDKT rate ratio of 4.27 from view. While the outlier data point affects the locally weighted scatterplot smoothing regression lines, no inferences or estimates are reported from this visualization, which does not inform the primary statistical analysis.

Figure 2.. Estimated Living Donor Kidney Transplant (LDKT) Black-White Rate Ratios (RRs): Model Covariates Optimized for Equity

Blue circles correspond to observed values of covariates; orange circles correspond to estimated LDKT RRs and associated 95% prediction intervals under a hypothetical scenario in which modifiable covariates for all transplant centers are fixed at values that promote equity of LDKT access while nonmodifiable covariates remain as observed. Error lines indicate 95% CIs. Note that the order of centers on the x-axis is identical to that of Figure 3.

Figure 3.. Estimated Living Donor Kidney Transplant (LDKT) Black-White Rate Ratios (RRs): Model Covariates Fixed at National Median

Estimated LDKT RRs for all transplant centers in 2018. Blue circles correspond to observed values of covariates; orange circles to a hypothetical scenario in which nonmodifiable covariates for all transplant centers are fixed at their overall median values while modifiable covariates remain as observed. Error lines indicate 95% CIs. Note that the order of centers on the x-axis is identical to that of Figure 2.