Authors (including presenting author) :
Thomas W F Chung*, Rutherford C S Sin, Jonathan T C Kwan
Affiliation :
Department of Medicine, North District Hospital
Keyword 1: :
Chronic kidney disease
Keyword 2: :
24h urine collection
Keyword 3: :
Creatinine clearance
Keyword 4: :
2021 CKD-EPI equation
Keyword 6: :
Bland-Altman analysis
Introduction :
KDIGO 2024 guideline is promoting the use of the latest “2021 CKD-EPI eGFR equation” as a standard for chronic kidney disease (CKD) evaluation worldwide. To our knowledge, no comparative validation study has yet been done on Hans Chinese. Many hospitals in Hong Kong continue to utilise urine creatinine clearance (CrCl) in the estimation of glomerular infiltration rate (GFR). This legacy method is susceptible to measurement errors stemming from factors such as laborious sampling, often with missing samples, and variations in tubular creatinine secretion, influenced by age/sex differences and underlying pathologies.
Objectives :
This study aims to compare the interchangeability between urine CrCl and estimated GFR (eGFR) derived from the 2021 CKD-EPI equation among Han Chinese individuals with CKD, thereby validating its utility as a potential alternative to urine CrCl.
Methodology :
Bland-Altman analysis and Pearson correlation study were conducted using data from Han Chinese CKD patients receiving care at North District Hospital (NDH). Subgroup analyses were conducted for different CKD stages. Urine CrCl was adjusted for body surface area using the Dubois formula.
Result & Outcome :
A total of 3148 data entries were analysed. A very strong overall correlation was observed between urine CrCl and eGFR (r = 0.879, p < 0.001). Bland-Altman analysis revealed an acceptable overall bias of -2.6ml/min, indicating that urine CrCl tends to yield lower values than eGFR. We observed that this difference is more pronounced in less advanced CKD stages. However, the limits of agreement (LoA) (17.0 to -22.2ml/min) were sizeable to assume interchangeability between urine CrCl and eGFR for all data entries. Biases were similar but LoAs were favourably reduced in subgroup analyses for more advanced CKD stages, 9.5 to -16.6ml/min for CKD G4 and 4.6 to -7.8ml/min for CKD G5. This study affirms the correlation of urine CrCl and eGFR calculated by the 2021 CKD-EPI equation. While the agreement between both methods of GFR estimation is somewhat limited in less advanced CKD stages, the more advanced CKD stages demonstrated acceptable bias and agreement. The discrepancy between urine CrCl and eGFR is postulated to be due to incomplete urine collection in patients with preserved urine output. However, given the inconvenience associated with 24-hour urine collection, high laboratory cost and the satisfactory agreement between both methods in advanced CKD stages, where it matters most clinically in accurate GFR estimation, we advocate for the replacement of urine CrCl with CKD-EPI eGFR for monitoring the renal function trends in Han Chinese CKD patients. We plan to carry out future studies with bigger population to provide further reassurance with valid subgroup analyses.
