Background: Plasma creatinine is a key analyte for assessing renal function and staging chronic kidney disease (CKD) through the estimated glomerular filtration rate (eGFR). In many laboratories this test is performed in parallel by the kinetic Jaffe method and the enzymatic method on different systems, so the comparability of results must be assessed to ensure consistency in diagnosis and monitoring. This study evaluated the comparability of plasma creatinine results between the enzymatic method (Cobas c702) and the kinetic Jaffe method (AU5800) in CKD patients.
Methods: A descriptive experimental study following CLSI EP09-A3 was performed on 200 plasma samples from CKD patients, stratified by stages 2 to 5 (40 samples per stage), with concentrations spanning the measuring range (65 – 1130 µmol/L). Each sample was measured in duplicate on each system. Comparability was assessed using Pearson correlation, Passing–Bablok regression and Bland – Altman analysis; bias at clinical decision levels was estimated and compared with the allowable total error (TEa) derived from biological variation (8.87%).
Results: The two methods were strongly correlated (r = 0.998; r² = 0.996). Passing–Bablok regression: Enzymatic = −19.21 + 1.144 × Jaffe; the 95% CI of the slope (1.133 – 1.156) excluded 1 and that of the intercept (−21.64 to −17.29) excluded 0, indicating both proportional and constant systematic error. The mean difference (Enzymatic − Jaffe) was +15.0 µmol/L (+1.96%; 95% limits of agreement −12.4% to +16.4%). The bias was concentration-dependent: at low concentrations (stage 2) Jaffe read higher than enzymatic (−4.4%), whereas at high concentrations (stages 4–5) enzymatic read higher than Jaffe (+7.6% to +9.7%); 28.0% of samples exceeded the 8.87% TEa.
Conclusions: The two methods were strongly correlated but not comparable owing to a concentration-dependent systematic error, and therefore cannot be used interchangeably, particularly for serial monitoring within a patient and for CKD staging at high concentrations.