Integrating evidence and clinical application of the Kidney Failure Risk Equation (KFRE) in multimorbidity and frailty: a mixed-methods triangulation study

The Kidney Failure Risk Equation (KFRE) is recommended for guiding management and referral decision in chronic kidney disease (CKD). However, many individuals with CKD live with multimorbidity and/or frailty, which may influence both the predictive performance of KFRE and the ways in which clinical decisions are made in the context of other long-term conditions or an individual’s health status.

We conducted a mixed-methods triangulation of four complementary studies: 1) a systematic review of kidney failure prognostic models to identify consideration of multimorbidity and frailty in existing development or validation studies; 2) a quantitative validation study of KFRE in individuals with and without multimorbidity, using data from the UK Biobank and the Stockholm Creatinine Measurements project (SCREAM) assessing model performance and the impact of accounting for competing mortality risk; 3) a quantitative validation of KFRE in frailty subgroups using UK Biobank data, evaluating model performance and impact of cystatin C-based eGFR (estimated glomerular filtration rate); and 4) a qualitative study (MULTIPOINT) exploring patients’ and healthcare professionals’ perspectives on CKD, kidney failure risk and KFRE use in the context of multimorbidity and/or frailty. Findings were synthesised to identify areas of convergence (agreement), complementarity (added insight), and dissonance (contradiction) across the four analyses. 

The systematic review confirmed that multimorbidity and frailty were rarely considered or reported in existing kidney failure prognostic model development or validation. Across the quantitative studies, KFRE demonstrated good discrimination but variable calibration in individuals with multimorbidity and frailty, with underestimation in frailty groups and overestimation in those with increasing numbers of long-term conditions. Qualitative findings highlighted that CKD was often deprioritised by both patients and healthcare professionals, with limited awareness of the diagnosis and uncertainty surrounding kidney failure risk and symptom attribution. Although healthcare professionals recognised the potential value of KFRE to support shared decision making, barriers included workload pressures, system integration challenges and difficulties communicating risk in the context of competing health priorities and potential emotional burden of these discussions. Triangulation revealed convergence that frailty and multimorbidity affect both KFRE accuracy and its clinical applicability, complementarity where qualitative data highlighted barriers to implementation, and dissonance where statistical adequacy contrasted with limited real-world use.

KFRE remains a valuable prognostic tool but requires contextual adaption for individuals with multimorbidity and frailty. Effective use depends on supportive systems, continuity of care, and communication strategies that acknowledge uncertainty and competing priorities. Future research should focus on approaches to integrate KFRE into person-centred care, explore communication strategies/frameworks for discussing risk in complex CKD populations and in settings of uncertainty, and comparative studies with models that consider the competing risk of death.