Dropping Race From eGFR Doubles Estimated Prevalence of CKD in Black Populations

Larry Hanover

The study results were released as nephrologists and others are awaiting the findings of a joint task force of 2 national kidney organizations looking at alternative approaches to estimating glomerular filtration rate.

Last summer, the American Society of Nephrology (ASN) and the National Kidney Foundation (NKF) formed a joint task force to evaluate dropping the long-standing use of Black race in an equation estimating kidney function—the estimated glomerular filtration rate (eGFR).

On Thursday, a study published in the Journal of the American Society of Nephrology (JASN) showed that eliminating this race adjustment would double the estimated prevalence of chronic kidney disease (CKD) among Black adults.

Such a shift would have wide-ranging effects on clinical decisions, including medication dosing, referrals for specialty care, and eligibility for kidney transplantation and kidney donation, the researchers said.

GFR is the total volume of filtrate passing through the glomeruli each minute, but that is challenging to assess in real time in a physician’s office. Using serum creatine, factoring in age, height, weight, and gender, gives an estimate. Including Black race in the eGFR tacks on a multiplier, which then assigns higher values to Black patients. As higher values indicate better kidney function, there has been increasing recognition that this may lead to inequitable and delayed care in treating CKD.

Some medical centers have already moved to stop using the multiplier, as inclusion of race, a social and not a biological construct, has care implications.

In the current study, researchers used 2 national samples, the National Health and Nutrition Examination Survey (NHANES) from 2015-2016 and the Veterans Affairs (VA) Health Care System from 2015, and estimated the prevalence of CKD if race adjustments were removed from eGFR equations.

The estimated prevalence of CKD (defined as GFR < 60mL/min/1.73m2) among Black adults would double—from 5.2% to 10.6% in the NHANES sample and from 12.4% to 21.6% in the VA sample—if race adjustments were eliminated, the study found.

There were also reclassifications on severity of disease. Among all Black individuals in the NHANES database, 5.6% (95% CI­, 4.0-7.2%) were reclassified to CKD stage 3. Of those previously classified as CKD stage 3, 6.3% (95% CI, 2.7%-15.4%) were reclassified to CKD stage 4.

In the veteran population, the study examined the 6 most prescribed medications with dosing recommendations dependent on kidney function, including the antibiotic ciprofloxacin. With a new CKD classification, 5.8% to 41.0% of Black veterans might have doses adjusted or halted, including 41.0% taking gabapentin, 33.5% on ciprofloxacin, 24.1% taking metformin, 6.9% using atenolol, 6.6% taking rosuvastatin, and 5.8% taking tramadol.

One concern among clinicians was that the prediction of kidney failure would rise by removing the race-based algorithm. However, accuracy was marginally improved overall, with better calibration at higher levels of predicted risk. Among Black patients, the mean estimated 2-year risk of kidney failure increased from 2.7% to 3.4%. Analysis indicated both models overestimated kidney risk slightly in the overall population and underestimated it among the Black population at levels of predicted risk of at least 5%.

The inclusion of an adjustment for Black race may have improved precision in study populations, the researchers said. However, there is no definitive way to determine the race of individuals who, for example, consider themselves multiracial. The lack of biologic underpinnings raises concerns for “misuse of race as a biological variable,” they wrote.

The 1999 landmark study that proposed the eGFR and included a coefficient for Black patients did so on the basis of 3 flawed, small, poor-quality studies, according to a presentation at last year’s Kidney Week.

Estimating GFR is challenging and cumbersome. Use of a confirmatory test such as direct measurement of GFR or cystatin-c based estimation methods have been proposed to help make drug dosing more precise, but their lack of widespread availability and long turnaround times limit practicality.

Dropping the race adjustment in calculating the eGFR might reduce health disparities in treating kidney disease, but increase health disparities in other ways, said one of the authors in a statement, who nonetheless called the conversation one that is “long overdue.”

“The focus has rightly been on the ways in which race adjustment might disadvantage Black patients, but on the other hand, simply dropping race adjustment and leaving it at that may result in unintended consequences, such as less use of first-line treatments for diabetes and diabetic kidney disease,” cautioned Manjula Kurella Tamura, MD, MPH, of the VA Palo Alto Health Care System and Stanford University. “The fact that there would be potentially large changes in medication prescribing is not an argument for keeping the race adjustment. Rather, it's a reminder that small changes in GFR estimates can have large effects at the population level. Estimates of GFR are less precise than we often assume them to be.”

The ASN and the NKF have endorsed dropping race modifiers from creatinine-based GFR equations to estimate kidney function, and the NKF-ASN task force just published an interim report about the first phase of its work in the May issue of JASN.

The report states that the task force is working on recommendations on an alternative to the eGFR, taking into account, among other things, “biomarker choice, inputs and their availability for estimation and reporting, representation of diversity in participants in research foundational to equation development, and equation bias and accuracy compared with measured GFR for different race and ethnic groups.”

Allison Inserro contributed to this report.

Reference

Duggal V, Thomas I-C, Montez-Rath M, et al. National estimates of chronic kidney disease prevalence and potential impact of estimating glomerular filtration rate without race. J Am Soc Nephrol. Published online May 6, 2021. doi:10.1681/ASN.2020121780