KFRE as a criterion for referral to advanced chronic kidney disease consultations

The KDIGO guidelines recommend as a criterion for referral to ACKD consultations a risk >10% of requiring renal replacement therapy before 2 years. This risk should be assessed with externally validated prediction models. The prediction models (KFRE) and the Grams model meet these requirements. In an ACKD unit with a remission criterion of eGFR <30 mL/min, we proposed a retrospective cohort study to evaluate whether the application of a KFRE > 10% as a remission criterion allows differentiation of patients according to their clinical and analytical characteristics and their evolution.

We studied 573 patients followed in the ACKD clinic for at least 4 years. In the first consultation we classified them into two groups according to their 2-year KFRE: < or ≥10%. We compared their clinical and analytical characteristics and whether the prognoses made by Grams' model at baseline matched the observed evolution. Both groups were analyzed according to two age ranges: older and younger than 75 years.

Patients with KFRE ≥ 10% (53.4%), with respect to those with KFRE < 10%, were significantly younger and their annual eGFR loss was greater. They had different evolution: at 2 years, 38.9% vs 3% (p < 0.05) started RRT and 45.8% vs 79.8% (p < 0.05) remained in the clinic; at 4 years, 60.7% vs 15.6% (p < 0.05) started RRT and 16.7% vs 52.3% (p < 0.05) remained in the clinic. In the group with KFRE < 10% those ≥75 years vs <75 years initiated RRT and died previously in a significantly different proportion at 2 years: 1.2 vs 6.1 % and 23.7 vs 6.1 % respectively (P < 0.05). In the group with KFRE ≥ 10% those ≥75 years and those <75 years initiated RRT and previously died in a significantly different proportion at 2 years: 25.9 vs 46.4 % and 32.1 vs 5.7 % respectively (P < 0.05). Prediction models predicted all these differences quite accurately.

A KFRE ≥ 10% at 2 years would be an appropriate criterion for referral to ACKD consultations, since it would select a group of patients who are really going to require RRT in the medium term, regardless of their age, and would avoid the referral of patients at low risk of requiring RRT, mainly elderly patients.