OBJECTIVES: Hyperkalaemia (HK) is associated with increased risk of mortality. This study aimed to develop a risk prediction tool to assess the probability of mortality in patients with heart failure (HF) or chronic kidney disease (CKD) at risk of HK, based on relevant demographics and clinical risk factors.

METHODS: Poisson Generalized Estimating Equations fitted to 23,541 HF and 144,388 CKD patients in Clinical Practice Research Datalink (Jan 2006–Dec 2015) were incorporated within an Excel-based tool to predict annual probabilities of mortality using demographic, comorbidity, concomitant medication and clinical measurement data. An illustration of the tool’s output is presented relative to a baseline probability for a male/female with CKD/HF, non-smoker, aged 60, eGFR 50 ml/min/1.73m, serum potassium (K⁺) 4.5 mEq/L, without diabetes or renin-angiotensin-aldosterone system inhibitor (RAASi) prescription. Results were expressed as percentage increase/decrease from baseline.

RESULTS: The listed characteristics were all statistically significant predictors of mortality. Baseline annual mortaility probability was 0.016 and 0.068 in males and 0.008 and 0.055 in females, with CKD and HF, respectively. RAASi use was associated with decreased probability of death compared to baseline in CKD (56.6%) and HF (69.3%). Older age, increased K⁺, diabetes, smoking and reductions in eGFR all increased estimated probability of death. The influence of eGFR was greater for CKD patients (36.4%–153% for 10–30 ml/min/1.73m reduction) compared to HF (18.4%–65.2% for 10–30 ml/min/1.73m reduction). While the impact of K⁺ was greater for HF: 7.2% and 58.4% for K^{+ CONCLUSIONS: Utilising real-world UK data, this evaluation of the impact of clinical risk factors on mortality risk in patients with CKD or HF serves as the structural framework for a broader tool to enhance the assessment of risk of outcomes in patients susceptible to HK.}