Incident hyperkalaemia risk model in chronic kidney disease and diabetes: the FIDELITY programme

Abstract

Background and Aims No tools are available for identifying patients with chronic kidney disease and Type 2 diabetes at high risk of hyperkalaemia. Using the FIDELITY pooled patient data set, a risk model for incident hyperkalaemia was developed and validated. Methods The primary outcome was new-onset hyperkalaemia (serum potassium >5.5 mmol/L). Data from the placebo arm were used for derivation and from the finerenone arm for validation. An integer risk score was built and divided into low-, intermediate-, and high-risk hyperkalaemia categories. Assessed efficacy outcomes included cardiovascular and kidney composites. Results Seven baseline covariates (serum potassium >4.5 mmol/L, prior history of hyperkalaemia, no sodium–glucose co-transporter-2 inhibitor use, urine albumin-to-creatinine ratio >1000 mg/g, haemoglobin <12 g/dL, no thiazide-type diuretic use, and estimated glomerular filtration rate <45 mL/min/1.73 m2) were independently associated with new-onset hyperkalaemia and used for building the integer risk model. The risk scores for derivation and validation were accurate and well calibrated. The derived integer score ranged from 0 to 12 points. The risk of new-onset hyperkalaemia increased across hyperkalaemia risk categories with 2.7, 7.0, and 16.7% of patients reporting a hyperkalaemia event in the low-risk (0–3 points), intermediate-risk (4–6 points), and high-risk (7–12 points) groups with placebo, respectively. Irrespective of hyperkalaemia risk, finerenone reduced cardiovascular and kidney events vs placebo. Conclusions This integer risk score for new-onset hyperkalaemia in patients with chronic kidney disease and Type 2 diabetes could facilitate tailored treatment strategies and mitigate hyperkalaemia in high-risk patients.