Clinical Utility in Adopting Race-free Kidney Donor Risk Index

How do we increase deceased donor kidney utilization and reduce discard?

PURPOSE OF REVIEW

This review aims to address the critical issue of expanding deceased donor kidney pool and reducing the discard rates of viable kidneys in the United States. It highlights advances in organ preservation techniques and explores strategies for expanding the donor pool by leveraging suboptimal and high-risk nonuse kidneys, including those affected by acute kidney injury (AKI), hepatitis C virus (HCV), and hepatitis B virus (HBV).

RECENT FINDINGS

Innovations in organ preservation, including hypothermic and normothermic machine perfusion, have demonstrated efficacy in improving outcomes for marginal and extended-criteria kidneys. The integration of normothermic regional perfusion (NRP) for donation after cardiac death (DCD) donors has enhanced organ utilization and graft viability. Additionally, research confirms that kidneys from AKI and HCV-positive donors, when managed with appropriate protocols, yield comparable long-term outcomes to standard transplants. Emerging data on HBV-positive donor kidneys further underscore their potential to safely expand transplant access with targeted antiviral prophylaxis.

SUMMARY

Optimizing deceased donor kidney utilization requires a multi-faceted approach, including advancements in preservation technologies, evidence-based decision-making for high-risk organs, and policy innovations. Leveraging these strategies can help address the growing organ shortage, enhance transplant outcomes, and ensure broader access to life-saving kidney transplants.

Multivariable Predictors of Poorer Renal Function Among 1119 Deceased Donor Kidney Transplant Recipients During the First Year Post-Transplant, With a Particular Focus on the Influence of Individual KDRI Components and Donor AKI

Given our desire to reduce kidney transplant waiting times by utilizing more difficult-to-place ("higher-risk") DD kidneys, we wanted to better understand post-transplant renal function among 1119 adult DD recipients consecutively transplanted during 2016-2019. Stepwise linear regression of eGFR (CKD-EPI formula) at 3-, 6-, and 12-months post-transplant (considered as biomarkers for longer-term outcomes), respectively, was performed to determine the significant multivariable baseline predictors, using a type I error ≤ 0.01 to avoid spurious/weak associations. Three unfavorable characteristics were selected as highly significant in all three models: Older DonorAge (yr) (p < 0.000001), Longer StaticColdStorage Time (hr) (p < 0.000001), and Higher RecipientBMI (p ≤ 0.00003). Other significantly unfavorable characteristics included: Shorter DonorHeight (cm) (p ≤ 0.00001), Higher Natural Logarithm {Initial DonorCreatinine} (p ≤ 0.001), Longer MachinePerfusion Time (p ≤ 0.003), Greater DR Mismatches (p = 0.01), DonorHypertension (p ≤ 0.004), Recipient HIV+ (p ≤ 0.006), DCD Kidney (p = 0.002), Cerebrovascular DonorDeath (p = 0.01), and DonorDiabetes (p = 0.01). Variables not selected into any model included DonorAKI Stage (p ≥ 0.24), Any DonorAKI (p ≥ 0.04), and five KDRI components: two DonorAge splines at 18 years (p ≥ 0.52) and 50 years (p ≥ 0.28), BlackDonor (p ≥ 0.08), DonorHCV+ (p ≥ 0.06), and DonorWeight spline at 80 kg (p ≥ 0.03), indicating that DonorAKI and the weaker KDRI components have little, if any, prognostic impact on renal function during the first 12 months post-transplant. Additionally, biochemical determinations with skewed distributions such as DonorCreatinine are more accurately represented by natural logarithmic transformed values. In conclusion, one practical takeaway is that donor AKI may be ignored when evaluating DD risk.

Online Database of Clinical Algorithms with Race and Ethnicity

Some clinical algorithms incorporate an individual's race, ethnicity, or both as an input variable or predictor in determining diagnoses, prognoses, treatment plans, or risk assessments. Inappropriate use of race and ethnicity in clinical algorithms at the point of care may exacerbate health disparities and promote harmful practices of race-based medicine. We identified 42 risk calculators that use race as a predictor, five laboratory test results with different reference ranges recommended for different races, one therapy recommendation based on race, 15 medications with guidelines for initiation and monitoring based on race, and four medical devices with differential racial performance. Information on these clinical algorithms are freely available at http://www.clinical-algorithms-with-race-and-ethnicity.org. This resource aims to raise awareness about the use of race in clinical algorithms and to track the progress made toward eliminating its inappropriate use. The database will be actively updated to include clinical algorithms based on race that were missed, along with additional characteristics of these algorithms.

Updating the kidney donor risk index: Removing donor race and hepatitis C virus status

This study reports the results of a recalculation of the kidney donor risk index (KDRI) formula requested by the Organ Procurement and Transplantation Network's Minority Affairs Committee to remove the donor race and hepatitis C virus (HCV) status variables. The updated KDRI model was fit on adult, deceased donor, solitary kidney, first-time transplants from 2018-2021. Deceased donors from 2018 through 2021 were included in a counterfactual analysis to evaluate how the kidney donor profile index (KDPI) would change if race and HCV seropositivity were excluded. When recalculating the original KDRI models on 2018-2021 transplants, the donor Black race coefficient was only slightly lower (β = 0.18 in the original model; β = 0.15 in the 2018-2021 cohort), while the donor HCV seropositivity coefficient was substantially lower (β = 0.24 in the original model; β = -0.04 in the 2018-2021 cohort). Among Black donors, the probability of being classified as KDPI ≤20% increased and the probability of being classified as KDPI >85% decreased notably when the Black race and HCV variables were removed from the model. Removing the donor race and donor HCV status variables in an updated KDRI model resulted in more racially equitable KDPI distributions.

Enhancing Clinical Decision Support in Nephrology: Addressing Algorithmic Bias Through Artificial Intelligence Governance

There has been a steady rise in the use of clinical decision support (CDS) tools to guide nephrology as well as general clinical care. Through guidance set by federal agencies and concerns raised by clinical investigators, there has been an equal rise in understanding whether such tools exhibit algorithmic bias leading to unfairness. This has spurred the more fundamental question of whether sensitive variables such as race should be included in CDS tools. In order to properly answer this question, it is necessary to understand how algorithmic bias arises. We break down 3 sources of bias encountered when using electronic health record data to develop CDS tools: (1) use of proxy variables, (2) observability concerns and (3) underlying heterogeneity. We discuss how answering the question of whether to include sensitive variables like race often hinges more on qualitative considerations than on quantitative analysis, dependent on the function that the sensitive variable serves. Based on our experience with our own institution's CDS governance group, we show how health system-based governance committees play a central role in guiding these difficult and important considerations. Ultimately, our goal is to foster a community practice of model development and governance teams that emphasizes consciousness about sensitive variables and prioritizes equity.

Inconsistent values and algorithmic fairness: a review of organ allocation priority systems in the United States

BACKGROUND

The Organ Procurement and Transplant Network (OPTN) Final Rule guides national organ transplantation policies, mandating equitable organ allocation and organ-specific priority stratification systems. Current allocation scores rely on mortality predictions.

METHODS

We examined the alignment between the ethical priorities across organ prioritization systems and the statistical design of the risk models in question. We searched PubMed for literature on organ allocation history, policy, and ethics in the United States.

RESULTS

We identified 127 relevant articles, covering kidney (19), liver (60), lung (24), and heart transplants (23), and transplant accessibility (1). Current risk scores emphasize model performance and overlook ethical concerns in variable selection. The inclusion of race, sex, and geographical limits as categorical variables lacks biological basis; therefore, blurring the line between evidence-based models and discrimination. Comprehensive ethical and equity evaluation of risk scores is lacking, with only limited discussion of the algorithmic fairness of the Model for End-Stage Liver Disease (MELD) and the Kidney Donor Risk Index (KDRI) in some literature. We uncovered the inconsistent ethical standards underlying organ allocation scores in the United States. Specifically, we highlighted the exception points in MELD, the inclusion of race in KDRI, the geographical limit in the Lung Allocation Score, and the inadequacy of risk stratification in the Heart Tier system, creating obstacles for medically underserved populations.

CONCLUSIONS

We encourage efforts to address statistical and ethical concerns in organ allocation models and urge standardization and transparency in policy development to ensure fairness, equitability, and evidence-based risk predictions.

Adjusting for race in metrics of organ procurement organization performance

The Scientific Registry of Transplant Recipients has previously reported the effects of adjusting for demographic variables, including race, in the Centers for Medicare & Medicaid Services (CMS) organ procurement organization (OPO) performance metrics: donation rate and transplant rate. CMS chose not to adjust for most demographic variables other than age (for the transplant rate), arguing that there is no biological reason that these variables would affect the organ donation/utilization decision. However, organ donation is a process based on altruism and trust, not a simple biological phenomenon. Focusing only on biological impacts on health ignores other pathways through which demographic factors can influence OPO outcomes. In this study, we update analyses of demographic adjustment on the OPO metrics for 2020 with a specific focus on adjusting for race. We find that adjusting for race would lead to 8 OPOs changing their CMS tier rankings, including 2 OPOs that actually overperform the national rate among non-White donors improving from a tier 3 ranking (facing decertification without possibility of recompeting) to a tier 2 ranking (allowing the possibility of recompeting). Incorporation of stratified and risk-adjusted metrics in public reporting of OPO performance could help OPOs identify areas for improvement within specific demographic categories.

Immunologic Benefits of 0-antigen Mismatched Transplants: No Added Boost for Racial and Ethnic Minorities

Background

Systemic barriers to posttransplant care, including access to immunosuppressant medications, contribute to higher rates of kidney transplant failure in racial minorities. Matching donor and recipient HLA alleles reduce allorecognition, easing reliance on immunosuppression. We hypothesize that 0-antigen mismatch transplants may provide stronger protection against graft loss in racial minorities.

Methods

We compared adult, single-organ, deceased-donor kidney transplants in the United States from 2007 to 2016 by degree of HLA mismatch (0- versus ≥1-antigen mismatch). We examined time-to-allograft failure, with death as a competing event, using multivariable Weibull models, stratified by recipient race (White versus non-White), and evaluated the interaction between mismatch and recipient race. We used Kaplan-Meier imputation to account for competing risk of death.

Results

We analyzed 102 114 transplants (median follow-up, 5.6 y; 16 862 graft losses, 18 994 deaths). Zero-antigen mismatch was associated with improved allograft survival (adjusted subdistribution hazard ratio [sHR] 0.80; 95% confidence interval [CI], 0.75-0.85). When stratified by recipient race, the effect of 0-antigen mismatch was more pronounced in White (unadjusted sHR 0.78; 95% CI, 0.72-0.83) versus non-White recipients (sHR 0.88; 95% CI, 0.79-0.99; interaction P = 0.04). The differential effect was attenuated after adjusting for covariates (sHR 0.78; 95% CI, 0.73-0.84 versus sHR 0.87; 95% CI, 0.77-0.98; interaction P = 0.10).

Conclusions

Zero-antigen mismatch transplants conferred a 20% risk reduction in allograft loss, which was similar between non-White and White recipients. This may reflect an increased degree of mismatch at other HLA alleles and non-HLA alleles in non-White recipients or because of the extent of systemic barriers to healthcare borne by minority recipients.

The Impact of Health Care Algorithms on Racial and Ethnic Disparities : A Systematic Review

BACKGROUND

There is increasing concern for the potential impact of health care algorithms on racial and ethnic disparities.

PURPOSE

To examine the evidence on how health care algorithms and associated mitigation strategies affect racial and ethnic disparities.

DATA SOURCES

Several databases were searched for relevant studies published from 1 January 2011 to 30 September 2023.

STUDY SELECTION

Using predefined criteria and dual review, studies were screened and selected to determine: 1) the effect of algorithms on racial and ethnic disparities in health and health care outcomes and 2) the effect of strategies or approaches to mitigate racial and ethnic bias in the development, validation, dissemination, and implementation of algorithms.

DATA EXTRACTION

Outcomes of interest (that is, access to health care, quality of care, and health outcomes) were extracted with risk-of-bias assessment using the ROBINS-I (Risk Of Bias In Non-randomised Studies - of Interventions) tool and adapted CARE-CPM (Critical Appraisal for Racial and Ethnic Equity in Clinical Prediction Models) equity extension.

DATA SYNTHESIS

Sixty-three studies (51 modeling, 4 retrospective, 2 prospective, 5 prepost studies, and 1 randomized controlled trial) were included. Heterogenous evidence on algorithms was found to: a) reduce disparities (for example, the revised kidney allocation system), b) perpetuate or exacerbate disparities (for example, severity-of-illness scores applied to critical care resource allocation), and/or c) have no statistically significant effect on select outcomes (for example, the HEART Pathway [history, electrocardiogram, age, risk factors, and troponin]). To mitigate disparities, 7 strategies were identified: removing an input variable, replacing a variable, adding race, adding a non-race-based variable, changing the racial and ethnic composition of the population used in model development, creating separate thresholds for subpopulations, and modifying algorithmic analytic techniques.

LIMITATION

Results are mostly based on modeling studies and may be highly context-specific.

CONCLUSION

Algorithms can mitigate, perpetuate, and exacerbate racial and ethnic disparities, regardless of the explicit use of race and ethnicity, but evidence is heterogeneous. Intentionality and implementation of the algorithm can impact the effect on disparities, and there may be tradeoffs in outcomes.

PRIMARY FUNDING SOURCE

Agency for Healthcare Quality and Research.

A translational perspective towards clinical AI fairness

Artificial intelligence (AI) has demonstrated the ability to extract insights from data, but the fairness of such data-driven insights remains a concern in high-stakes fields. Despite extensive developments, issues of AI fairness in clinical contexts have not been adequately addressed. A fair model is normally expected to perform equally across subgroups defined by sensitive variables (e.g., age, gender/sex, race/ethnicity, socio-economic status, etc.). Various fairness measurements have been developed to detect differences between subgroups as evidence of bias, and bias mitigation methods are designed to reduce the differences detected. This perspective of fairness, however, is misaligned with some key considerations in clinical contexts. The set of sensitive variables used in healthcare applications must be carefully examined for relevance and justified by clear clinical motivations. In addition, clinical AI fairness should closely investigate the ethical implications of fairness measurements (e.g., potential conflicts between group- and individual-level fairness) to select suitable and objective metrics. Generally defining AI fairness as "equality" is not necessarily reasonable in clinical settings, as differences may have clinical justifications and do not indicate biases. Instead, "equity" would be an appropriate objective of clinical AI fairness. Moreover, clinical feedback is essential to developing fair and well-performing AI models, and efforts should be made to actively involve clinicians in the process. The adaptation of AI fairness towards healthcare is not self-evident due to misalignments between technical developments and clinical considerations. Multidisciplinary collaboration between AI researchers, clinicians, and ethicists is necessary to bridge the gap and translate AI fairness into real-life benefits.

Long and Short-Term Effects of Hypothermic Machine Perfusion vs. Cold Storage on Transplanted Kidneys from Expanded Criteria Donors-A Matched Comparison Study

Hypothermic machine perfusion (HMP) has been shown to reduce delayed graft function (DGF)-rates in kidneys from expanded criteria donors (ECD) and may increase graft survival compared with static cold storage (SCS). This single-center, retrospective observational study aimed to evaluate this effect. The primary endpoint was the DGF-rate, defined as the use of dialysis in the first postoperative week, excluding the first 24 h. The main secondary endpoint was graft survival at 5 years. Recipients of ECD-kidneys between 2013 and 2021 with ≤2 grafts were included (n = 438). The SCS-kidneys were marginal-matched by propensity score to the HMP-group for donor age, cold ischemia time, and graft number. Multivariable adjusted analysis for confounders in the unmatched cohort and caliper-based ID-matching constituted sensitivity analyses. HMP showed a trend to lower DGF-rate in the marginal-matched comparison (9.2% vs. 16.1%, p = 0.063). This was strengthened by a significant benefit observed for HMP in both the sensitivity analyses: an adjusted OR of 0.45 (95% CI: 0.24; 0.84; p = 0.012) in the multivariable analysis and DGF-rate of 8.7% vs. 17.4% (p = 0.024) after ID-matching. The 5-year graft survival rate was >90% in both groups, with no benefit using HMP (HR = 0.79; 95% CI:0.39-1.16; p = 0.52). Our results suggest that HMP may be effective in decreasing DGF-rates, however, without any significant benefit in graft survival.

Impacts of removing race from the calculation of the kidney donor profile index

The kidney donor risk index (KDRI), standardized as the kidney donor profile index (KDPI), estimates graft failure risk for organ allocation and includes a coefficient for the Black donor race that could create disparities. This study used the Scientific Registry of Transplant Recipients data to recalculate KDRI coefficients with and without the Black race variable for deceased donor kidney transplants from 1995 to 2005 (n = 69 244). The recalculated coefficients were applied to deceased kidney donors from 2015 to 2021 (n = 72 926) to calculate KDPI. Removing the Black race variable had a negligible impact on the model's predictive ability. When the Black race variable was removed, the proportion of Black donors above KDPI 85%, a category with a higher risk of organ nonuse, declined from 31.09% to 17.75%, closer to the 15.68% above KDPI 85% among non-Black donors. KDPI represents percentiles relative to all other donors, so the number of Black donors moving below KDPI 86% was roughly equal to the number of non-Black donors moving above KDPI 85%. Removing the Black donor indicator from KDRI/KDPI may improve equity without substantial overall impact on the transplantation system, though further improvement may require the use of absolute measures of donor risk KDRI rather than relative measures of risk KDPI.