Abstract
BACKGROUND: A race-free glomerular filtration rate (GFR) estimation equation has recently been developed. However, the performance of the new Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations needs to be evaluated in Asian populations. METHODS: We performed a cross-sectional study at a single center in South Korea. The measured GFR (mGFR) was determined based on systemic inulin clearance. The GFR was estimated using the five CKD-EPI equations: 2009 CKD-EPIcr, 2012 CKD-EPIcr-cys, 2012 CKD-EPIcys, 2021 CKD-EPIcr, and 2021 CKD-EPIcr-cys. The performances of five estimated GFR (eGFR) equations were assessed by bias, precision, and accuracy (percentage of estimates within 30% of mGFR). RESULTS: The median mGFR and interquartile range (IQR) was 53.5 (32.4-80.0) mL/min/1.73 m2. The mGFR better correlated with 2009 CKD-EPIcr (ρ = 0.628) and 2021 CKD-EPIcr-cys (ρ = 0.806) than with 2021 CKD-EPIcr (ρ = 0.623) and 2012 CKD-EPIcr-cys (ρ = 0.801). The median bias of 2009 CKD-EPIcr and 2012 CKD-EPIcr-cys were lower than those of 2021 CKD-EPI equations (2009 CKD-EPIcr, 2.24 [IQR, -8.83 to 17.39] vs. 2021 CKD-EPIcr, 5.40 [IQR, -6.04 to 20.40]; 2012 CKD-EPIcr-cys, 6.74 [IQR, -2.81 to 20.80] vs. 2021 CKD-EPIcr-cys, 10.54 [IQR, 0.30-24.37]; all in mL/min/1.73 m2). The percentage of eGFR values within 30% of mGFR was higher in 2009 CKD-EPIcr and 2012 CKD-EPIcr-cys equations than 2021 CKD-EPI equations. The CKD prevalence in 2009 CKD-EPIcr, 2021 CKD-EPIcr, 2012 CKD-EPIcr-cys, and 2021 CKD-EPIcr-cys was 54.8%, 51.0%, 47.7%, and 44.8%, respectively. CONCLUSION: Our study demonstrated better performance of the original CKD-EPIcr and CKD-EPIcr-cys equations than the 2021 new CKD-EPI equations. We do not recommend the adoption of the new CKD-EPI equations in Korea.