Better Prediction of Clinical Outcome with Estimated Glomerular Filtration Rate by CKD-EPI 2021

Abstract

Background

While the real-world impact of estimated glomerular filtration rate (eGFR) equation change on clinical outcome in a longitudinal cohort setting is limited, external valuation of equation performance should be performed in different population cohorts. This study aimed to compare differential impacts of eGFR values, calculated by 5 equations in a Korean patient population, on clinical outcomes.

Methods

This retrospective longitudinal follow-up cohort study analyzed 23 246 participants with standardized creatinine/cystatin C assay-based laboratory results. The primary exposure was baseline eGFR calculated by 5 different equations including the recently developed 2021 race-free Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations. Clinical outcomes including all-cause mortality, renal replacement therapy, and albuminuria were analyzed to estimate the hazard ratio of the eGFR on clinical outcomes.

Results

Among the 5 equations, CKD-EPI 2021 with creatinine and cystatin C (CKD-EPI 2021-CrCys) showed an earlier increase in hazard ratios for all clinical outcomes, while CKD-EPI 2012 with cystatin C showed a higher hazard ratio for all-cause mortality at low eGFR. Replacing CKD-EPI 2012 with CKD-EPI 2021-CrCys, 5.4% of patients with mortality and 3.3% of patients who received renal replacement therapy were reclassified to a lower risk stage.

Conclusions

The 2021 CKD-EPI equations were acceptable in a Korean population, with better predictive power for clinical outcomes when compared to previous equations. The updated race-free factors for eGFR calculation improved identification of patients at risk for clinical outcomes.