KFRE Is Superior to eGFR Alone for ESKD Risk Prediction

Full access

The four-variable kidney failure risk equation (KFRE) is a better predictor of end stage kidney disease (ESKD) risk compared with the estimated glomerular filtration rate (eGFR) alone, with or without adjustment for race, reports a study in the Annals of Internal Medicine.

The researchers used data from the Chronic Renal Insufficiency Cohort to evaluate different eGFR equations for prediction of ESKD, defined as dialysis initiation or transplantation. The analysis included data on 3873 participants with chronic kidney disease (CKD), with a total of 13,902 2-year risk periods.

For each participant, eGFR was calculated using the CKD Epidemiology Collaboration

The four-variable kidney failure risk equation (KFRE) is a better predictor of end stage kidney disease (ESKD) risk compared with the estimated glomerular filtration rate (eGFR) alone, with or without adjustment for race, reports a study in the Annals of Internal Medicine.

The researchers used data from the Chronic Renal Insufficiency Cohort to evaluate different eGFR equations for prediction of ESKD, defined as dialysis initiation or transplantation. The analysis included data on 3873 participants with chronic kidney disease (CKD), with a total of 13,902 2-year risk periods.

For each participant, eGFR was calculated using the CKD Epidemiology Collaboration (CKD-EPI) equation, based on serum creatinine and cystatin C, with or without adjustment for race.

A 2-year risk of ESKD was estimated using the validated KFRE, which includes age, sex, eGFR, and urinary albumin-creatinine ratio. The old and new eGFR equations, alone and as part of the KFRE, were evaluated as predictors of ESKD risk.

At up to 15 years’ follow-up, 856 participants developed ESKD. With or without race-adjusted eGFR, the KFRE was superior in predicting ESKD risk: area under the curve ranged from 0.945 to 0.954 compared with 0.900 to 0.927 with eGFR alone. Although the KFRE had a similar predictive performance with different eGFR equations, the creatinine equation without race adjustment had better calibration among participants of Black race.

A KFRE score greater than 20% had 94%–97% specificity in predicting 2-year ESKD risk, similar to the 95%–98% range with eGFR less than 20 mL/min/1.73 m2. However, KFRE over 20% had higher specificity: 68%–78% compared with 42%–66% with eGFR under 20 mL/min/1.73 m2. Prediction was consistently better with KFRE, regardless of the eGFR estimating equation used.

Previous eGFR equations included adjustment for race, reflecting evidence that individuals of Black race have higher average serum creatinine. Newer equations have removed adjustment for race, but the impact on ESKD risk prediction remains unclear. Because the KFRE includes more information than eGFR alone, it may improve risk prediction and clinical decision-making.

The new analysis finds that KFRE scores are a better predictor of 2-year ESKD risk compared with eGFR alone. “[A] KFRE score greater than 20% could be used for preparing for kidney replacement therapy,” the researchers write [Bundy JD, et al. Prediction of end-stage kidney disease using estimated glomerular filtration rate with and without race: A prospective cohort study. Ann Intern Med, published online ahead of print January 11, 2022. doi: 10.7326/M21-2928; https://www.acpjournals.org/doi/10.7326/M21-2928].

Save