Planning of Kidney Replacement Therapy in Advanced CKD using the KFRE formula in a Spanish Multicenter Cohort

Abstract Background The Kidney Failure Risk Equation (KFRE) is widely used to estimate the risk of kidney replacement therapy (KRT), but evidence in real-world multicentre advanced CKD (ACKD) cohorts—particularly among older patients with substantial competing mortality—remains limited. Methods We conducted a multicentre retrospective cohort study including adults with ACKD (baseline eGFR <20 mL/min/1.73 m²) followed at specialised multidisciplinary clinics in Barcelona and Seville (2017–2022). The 2-year 4-variable KFRE (age, sex, eGFR, UACR) was externally validated for predicting KRT initiation, treating death as a competing event. Discrimination was assessed using time-dependent AUC; calibration was evaluated using decile-based plots and quantitative metrics (calibration-in-the-large, Brier score) using Fine–Gray competing risk methods. Clinical utility was explored using decision curve analysis (DCA) and a 20% KFRE threshold in patients aged ≥65 years. Results A total of 503 patients were included. During follow-up, 248 patients initiated KRT and 94 died before KRT. The 2-year 4-variable KFRE showed good discrimination (time-dependent AUC 0.82, 95% CI 0.73–0.91), preserved across age strata. Calibration demonstrated appropriate risk ordering across deciles with mild underestimation (CITL −0.21 by Fine–Gray; Brier score 0.23). In patients aged ≥65 years, a KFRE ≥20% was strongly associated with KRT initiation in competing risk models (SHR 5.92, 95% CI 4.08–8.60). DCA showed positive net benefit across clinically relevant thresholds. Conclusions In a multicentre Spanish ACKD cohort, the 2-year KFRE demonstrated robust discrimination and acceptable calibration. These findings highlight theclinical utility of the KFRE for risk stratification and KRT planning in older patients with advanced CKD.