Background: A creatinine clearance (CrCl) is still often requested to estimate the glomerular filtration rate (GFR) in clinical practice. However, the diversity of serum and urine creatinine (Scr, Ucr) assays leads to different CrCl-results which are here compared with each other and with the CKD-EPI eGFR-formula.
Methods: We collected information on urine volume, Ucr and Scr using Roche's enzymatic (E), compensated Jaffe (CJ) and Jaffe (J) assay for 589 patients. To allow comparison with the CKD-EPI prediction results, CrCl was normalized for body surface area.
Results: Differences between CrCl-E and CrCl-CJ are rather small as opposed to the large differences with CrCl-J. However, two compensating errors in the CrCl-J calculation result in a closer agreement with CKD-EPI eGFR, than between CrCl-CJ or CrCl-E and CKD-EPI eGFR. The explained variance R(2) in all three cases is smaller than 0.66, demonstrating the very large scatter of the data around the regression line.
Conclusions: CrCl determination is very assay-dependent. Although many clinical labs have switched to ID-GC/MS-standardized assays (E and CJ) for the determination of Scr and Ucr to improve analytical accuracy, the increased deviation of the normalized CrCl from the CKD-EPI prediction illustrates that the use of CrCl remains questionable for clinical practice. When a CrCl is requested, we would even recommend clinical labs who work with compensated Jaffe assays not to compensate the Scr-J value.
Keywords: Creatinine assay; creatinine clearance; estimation of glomerular filtration rate; kidney disease; kidney function test.