Evaluation of the performance of common kits for the measurement of creatinine in serum and urine: are the results of these methods suitable for clinical use?

Measuring serum and urine creatinine level is currently the most common method for assessing renal function. For this purpose, Jaffe’s method is most commonly used. The inconsistency in the results obtained from one single sample may be due to poor quality of the measurement method, inappropriate calibration, or improper use of the measurement method by the laboratory.
To evaluate the performance of creatinine measurement methods, five commonly used kits in Iran, i.e. Pars Azmon, Pishtaz Teb, Audit, Man and Bionik, were evaluated for uncertainty, detection capabilities, linearity and comparison of results. The results were statistically analyzed by using the coefficient of variation, T-test and linear regression and clinically evaluated based on the allowable total error.
The imprecision and detection capabilities of the kits were acceptable, but linearity results were not acceptable according to measurement range provided by the manufacturers, exept for Audit and Bionik kits. In addition, statistically in all cases and clinically in most cases, there was a significant difference between the results of each kit with the total mean results as the target level.
According to the results of this study, which have been performed in Iran for the first time, it is necessary to replace the Jaffe's method creatinine measurements with the enzymatic methods. In addition, the homogenization and standardization of these methods in comparison to a reference method is necessary.