# Epidemiological Assessment and Inference in Race-Based Clinical Algorithms: A Narrative Review and Health Policy Analysis Focused on Living Kidney Donation

## TL;DR

This paper reviews how race-based clinical algorithms affect living kidney donation policies and highlights issues in the evidence base that hinder health equity.

## Contribution

The paper identifies threats to scientific validity in race-based health policies and offers recommendations for improving equity-focused health policy.

## Key findings

- Ambiguity in defining race and ethnicity variables undermines the validity of clinical algorithms.
- Type III errors and racial essentialist biases are present in subgroup analyses of kidney donation policies.
- Structural barriers prevent effective use of evidence to promote health equity in kidney donation.

## Abstract

Minoritized racial and ethnic groups in the United States face long-standing disparities in a variety of health outcomes, owing to inequitable distribution of social and structural determinants of health along racial and ethnic lines. Although the existence of such disparities has long been a topic of scientific inquiry, there has been a dearth of investigations regarding their underlying mechanisms and potential remedies. This presents a challenge to those creating evidence-based and equity-focused health policy.

We conducted an evidence-based, equity-focused narrative review about living kidney donor eligibility using salient literature about donor eligibility and racial and ethnic disparities in kidney transplantation and donation in the United States. We sought to examine the rigor and reproducibility of the evidence base regarding race- and ethnicity-based living kidney donation policies.

Our review identified several threats to scientific validity in the evidence base, including ambiguity in the operationalization of race and ethnicity variables, instances of type III error and racial essentialist biases, and causal inferences made using underpowered or scientifically unsubstantial subgroup analyses. We also identified structural barriers to the interpretation of this evidence to advance health equity, including barriers to the practices of clinical equipoise and shared medical decision-making.

Threats to scientific validity and inferential errors in the evidence base about health inequities may forestall progress toward equity. We provide recommendations for addressing such barriers using standards applied in other clinical and research domains.