Supplementary Material for: The superiority of European Kidney Function Consortium Cystatin C-based formula for risk stratification of all-cause and cardiovascular deaths in US adults

Introduction We intended to compare the predictive value for all-cause and cardiovascular deaths between estimated glomerular filtration rate (eGFR) derived from the European Kidney Function Consortium (EKFC) cystatin C-based formula, the EKFC creatinine-based formula, and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) cystatin C- or creatinine-based formulas.
Methods 4,132 participants from the National Health and Nutrition Examination Survey between 1999 and 2002 were included, and death information was obtained through the National Death Index. To compare predictive accuracy between EKFC eGFRcys (EKFC cystatin C-based formula), CKD-EPI eGFRcys (CKD-EPI cystatin C-based formula), EKFC eGFRcr (EKFC creatinine-based formula), and CKD-EPI eGFRcr (CKD-EPI creatinine-based formula), we conducted time-dependent receiver operator characteristics (ROC) curves and reclassification analysis.
Results During a median follow-up of 17.4 years, a total of 1,987 all-cause and 530 cardiovascular deaths were confirmed. Restricted cubic splines (RCS) analyses showed that reduced EKFC eGFRcys was linearly related to higher risks of all-cause and cardiovascular deaths (P-nonlinearity > 0.05). Time-dependent ROC curves suggested that EKFC eGFRcys exhibited higher predictive ability than CKD-EPI eGFRcys, EKFC eGFRcr, and CKD-EPI eGFRcr at 5-year and 10-year follow-ups. For 10-year all-cause deaths, EKFC eGFRcys yielded significant improvement over CKD-EPI eGFRcr (integrated discrimination improvement [IDI], 9.4%; net reclassification improvement [NRI], 39.7%). Similar improvement was observed in 10-year cardiovascular deaths when comparing EKFC eGFRcys to CKD-EPI eGFRcr (IDI, 6.7%; NRI, 45.1%).
Conclusion The EKFC eGFRcys outperformed CKD-EPI eGFRcys, EKFC eGFRcr, and CKD-EPI eGFRcr in predicting all-cause and cardiovascular deaths, providing the possibility to utilize EKFC eGFRcys in the stratification of death risk among the general US population.