A Step Toward Health Equity for Veterans: Evidence Supports Removing Race From Kidney Function Calculations.

ABSTRACT

BACKGROUND: The practice of race-based medicine fails to recognize that race cannot be used as a proxy for genetic ancestry and that racial and ethnic categories are complex sociopolitical constructs without biological basis. Clinical algorithms and equations that incorporate race modifiers and are currently considered standard for diagnosis and management of disease are appropriately being scrutinized for lack of biological plausibility and their role in exacerbating health inequities. In this paper, we review the history, evidence, and implications of using a Black race coefficient when calculating estimated glomerular filtration rate (eGFR) in the diagnosis and management of kidney disease. OBSERVATIONS Currently, the US Department of Veterans Affairs (VA) uses the Modification of Diet in Renal Disease (MDRD) equation for eGFR. This equation includes a Black race coefficient that results in an eGFR that is 21% higher for a Black patient when compared with a patient of any other race. The rationale for the inclusion of this coefficient is based on racist science that incorrectly assumes race as a proxy for genetic ancestry. Multiple studies across diverse Black populations demonstrate that the application of a race coefficient in kidney function estimation equations is inferior when compared with the race-neutral option. Furthermore, the most utilized eGFR equations are biased and imprecise. Because eGFR is the primary diagnostic method for detecting and managing kidney disease, preventing its progression, planning for dialysis, and evaluating for transplantation, it is vital that eGFR be as accurate, precise, and equitable as possible. CONCLUSIONS: The incorporation of a race coefficient in kidney estimation equations lacks biological plausibility and its use exacerbates kidney health disparities. Until a better method to estimate kidney function becomes available, a race-neutral option for current estimation equations should be applied for all patients.