409-P: Evaluating the Agreement between CKD-EPI Calculated eGFR and mGFR in People with Diabetes—An International Individual Participant Data Meta-analysis

Introduction: Diabetic kidney disease (DKD) is the leading cause of kidney failure worldwide. We and others have shown that CKD-EPI equation is less accurate in estimating glomerular filtration rate (eGFR) in people with diabetes. This study aimed to assess the agreement between directly measured glomerular filtration rate (mGFR) and eGFR (2021 CKD-EPI) in individuals with type 1 or type 2 diabetes, especially in clinically relevant ranges of GFR.

Methods: Datapoints (n=16,182) were collected from people with type 1 and type 2 diabetes from six international cohorts (Diabetes Control and Complication Trial [DCCT, USA], Colorado [USA], Germany, Korea, Spain, and Australia), with mGFR assessed using exogenous filtration markers. The intra-class correlation coefficient between mGFR and eGFR was estimated using (a) a one-step intra-individual participant data meta-analysis approach, and (b) a four-level mixed-effect linear model which included studies and individual participants as first and second level random effects, method of measurement as third level fixed effect and time as fourth level random effect.

Results: Whilst overall mGFR and eGFR agreement was good (ICC: 0.82, 95%CI: 0.79, 0.84), subgroup analyses based on clinically relevant GFR ranges(<30, 30-59, 60-89, ≥90 ml/min per 1.73m²) revealed poor agreement between mGFR and eGFR measurements (ICC: 0.38, 0.40, 0.29, 0.34 respectively).

Conclusion: Our data suggest the need for alternative methods to calculate eGFR in people with diabetes, particularly across clinically relevant GFR ranges, to improve early detection and management of kidney impairment and optimize intervention strategies for DKD.