Verification of Analytical Performance of the Jaffé Kinetic Creatinine Assay on the Architect ci-2800 According to ISO 15189 Guidelines

Sabrina  Belmahi; Oumaima  Nassiri; bdessamad  Amrani; Amjad  Idrissi; El-houcine  Sebbar; Mohammed  Choukri

doi:10.18502/ijml.v11i4.20092

Sabrina Belmahi Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco
Oumaima Nassiri Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco
bdessamad Amrani Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco
Amjad Idrissi Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco
El-houcine Sebbar Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco
Mohammed Choukri Faculty of Medicine and Pharmacy of Oujda, Mohammed First University, Morocco

DOI: https://doi.org/10.18502/ijml.v11i4.20092

Keywords: Architect ci 2800, Creatinine, ISO 15189, Jaffé kinetic method, Repeatability, Reproducibility

Abstract

Introduction: Creatinine is a key parameter for evaluating renal function. This study aims to assess the repeatability, reproducibility, measurement uncertainty, and method comparison for creatinine measurement by the Jaffé kinetic method on the Architect ci-2800, relying on three levels of controls (low, medium, high).

Materials and Methods: Thirty fresh serum 7.17–74, 6 mg/L range measurements per level were carried out to determine repeatability, and a monitoring period of several days made it possible to evaluate reproducibility. Measurement uncertainty was estimated following the internal quality control  + external evaluation of quality. The method comparison in order to estimate bias and the correlation coefficient was assessed by Bland–Altman analysis.. The results were interpreted according to limits set by the reference values (SFBC and RICOS), as well as the manufacturer.

Results: Repeatability coefficient of variations (CVs) were 1.13 % (low), 1.05 % (medium), and 0.50 % (high), all below RICOS targets. Intermediate CVs over 30 days were 2.91 %, 2.02 % and 1.75 %. Expanded uncertainty (k = 2) ranged from 6.4 % to 7.2 % (RICOS ≤ 8.2 %). Regression gave y = 1.005 x – 0.365, r = 0.999, with a mean bias of 2.15 %.

Conclusion: Creatinine measurement by the Jaffé kinetic method on the Architect ci-2800 shows excellent performance in terms of precision (repeatability, reproducibility, measurement uncertainty) and inter-instrument correlation. Quality criteria are generally satisfied, ensuring reliable use for routine clinical monitoring of renal function.