Equation for kidney failure risk that excludes race shows better sensitivity, specificity than eGFR alone

A four-variable kidney failure risk equation (KFRE) that includes age, sex, estimated glomerular filtration rate (eGFR), and urinary albumin-creatinine ratio but does not consider race showed better sensitivity and specificity than eGFR alone, an observational cohort study found.

Researchers used data from 3,873 participants with chronic kidney disease, 1,631 who self-reported as Black, to compare prediction of end-stage kidney disease (ESKD) among different eGFR equations. They defined ESKD as initiation of dialysis or kidney transplantation. The researchers used five Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations to calculate eGFR based on serum creatinine and/or cystatin C levels, with or without race adjustment, and calculated predicted two-year risk for ESKD using the four-variable KFRE. The study results were published Jan. 11 by Annals of Internal Medicine.

During a mean follow-up of 9.3 years (16 years maximum), 856 patients developed ESKD. Black participants had higher observed two-year risk for ESKD at baseline (6.5% [95% CI, 5.3% to 7.7%] vs. 5.5% [95% CI, 4.6% to 6.5%]) compared with non-Black participants. Across all eGFR equations, the KFRE score had superior prediction of two-year incidence of ESKD versus eGFR alone. Prediction performance of KFRE scores was similar using different eGFR equations, but calibration was better among Black participants with the creatinine equation without race adjustment. A KFRE score greater than 20% had similar specificity for predicting two-year ESKD risk (ranges, 0.94 to 0.97 vs. 0.95 to 0.98) but higher sensitivity (ranges, 0.68 to 0.78 vs. 0.42 to 0.66) among all participants versus an eGFR of less than 20 mL/min/1.73 m².

The authors concluded that the choice of eGFR equation in this study did not affect the performance of eGFR and the KFRE score in predicting two-year ESKD risk for ESKD. “All 5 eGFR equations used within the KFRE score had excellent discrimination for ESKD events, and each model was well-calibrated in both Black and non-Black participants, although the KFRE scores underestimated risk, particularly among Black participants,” they wrote. They noted that the KFRE score was superior to eGFR alone for predicting ESKD and that for all eGFR equations, in both Black and non-Black participants, a KFRE score greater than 20% had specificity similar to that of an eGFR cutoff of less than 20 mL/min/1.73 m² for two-year ESKD risk and consistently improved sensitivity regardless of the eGFR equation used.

“The National Kidney Foundation and American Society of Nephrology task force recommends using the CKD-EPI creatinine equation refit without race, which can be readily implemented in U.S. laboratories. In addition, the task force recommends increased measurement of cystatin C to facilitate use of the CKD-EPI creatinine and cystatin C equation refit without race, which may improve GFR estimation,” they wrote. “Our results, which show that the choice of eGFR equation does not affect 2-year ESKD prediction, support these recommendations.”

An accompanying editorial noted that the health inequities experienced by Black patients with kidney disease are longstanding and multifaceted and cannot be solved simply by tweaking the equations used to estimate kidney function.

“The solutions and the productive target of efforts to achieve racial equity in kidney health lie in improving our healthcare and public health systems, not in the fine tuning of the estimating equations for glomerular filtration rate nor in validating KFRE scoring with or without refit CKD-EPI equation,” the editorialist wrote.