NEW findings presented at UK Kidney Week 2026 from the AIM CKD UK study suggest that commonly used creatinine-based equations to estimate glomerular filtration rate (eGFR) may substantially misclassify kidney function, potentially delaying diagnosis and treatment for chronic kidney disease (CKD), particularly in certain ethnic groups.
The multi-centre analysis compared the performance of five widely used eGFR equations against measured GFR (mGFR) in a large and ethnically diverse UK cohort.
Concerns Over Creatinine-Based eGFR Equations in Diverse Populations
Accurate kidney function assessment is essential for diagnosing and staging Chronic Kidney Disease. However, the reliability of creatinine-based equations has been questioned, particularly in younger adults and minority populations. The study was designed to address these gaps highlighted by National Institute for Health and Care Excellence guidance.
Researchers conducted a cross-sectional analysis across 11 UK centres including 15,879 adults who underwent mGFR testing between 2009 and 2022 with a paired serum creatinine measurement. Five equations were evaluated: MDRD, CKD-EPI-2009, CKD-EPI-2021, EKFC, and Lund-Malmö Revised (LMR). Performance was assessed through bias, precision, 30% accuracy (P30), and agreement with measured GFR, with results stratified by ethnicity.
Overall, all equations systematically overestimated measured kidney function. The LMR equation demonstrated the best performance, showing the lowest bias (1.4 mL/min/1.73m²) and highest accuracy (P30 86%). The EKFC equation also performed well (P30 83.6%). In contrast, accuracy was significantly lower for CKD-EPI-2009 (76.0%), MDRD (73.5%), and CKD-EPI-2021 (70.1%).
Performance also varied significantly across ethnic groups. In White participants, LMR and EKFC achieved P30 values of 87% and 85%, respectively, compared with 77% using CKD-EPI-2009 and 71% using CKD-EPI-2021. Among Black participants, EKFC and LMR again performed best, while CKD-EPI-2021 and MDRD remained below recommended thresholds. South Asian participants showed the poorest accuracy overall, with CKD-EPI-2021 overestimating kidney function by almost 18 mL/min/1.73m² and achieving only 58% accuracy.
Reduced performance was also observed in younger adults under 25 years, as well as individuals with low body mass index or hypoalbuminaemia, highlighting limitations of creatinine as a biomarker in these groups.
Improving Equity in Kidney Disease Diagnosis and Management
Importantly, equation choice affected clinical classification and treatment eligibility. CKD-EPI equations correctly staged CKD in only 55–60% of cases compared with 65% using LMR. Differences also influenced eligibility for therapies such as Dapagliflozin, potentially delaying treatment in up to half of patients depending on the equation used.
The researchers conclude that continued reliance on CKD-EPI equations may underestimate CKD severity and create inequities in care. Adoption of LMR or EKFC equations in UK clinical practice could improve diagnostic accuracy and treatment access across diverse populations.
Reference
Gama R et al. Assessing eGFR vs measured GFR in a diverse UK multi-centre cohort (The AIM CKD UK study): Time for a change? UK Kidney Week, 10-12 March, 2026.
Featured image: Hospital Man on Adobe Stock
