Methodological evaluation of a turbidimetric method for the analysis of serum cystatin c and comparison with a nephelometric method

Objective: In reliable clinical laboratory practice, to use test results for the maximum benefit of patients, the new method in a laboratory should be validated. In this study our purpose is to evaluate performance characteristics of cystatin C immunoturbidimetric method in Roche Cobas Integra 800 analyser, and compare this method with immunonephelometric method in Dade Behring BNII analyser. Methods: Precision, linearity, recovery, interference experiments in Roche Cobas Integra 800 analyser and method comparison experiments were performed for cystatin C. Results: For Roche Cobas Integra 800 instrument, low and high concentrations for withinrun and day to day CV values were 3.97%, 1.32% and 7.24%, 4.16% respectively. In linearity experiment, regression graph equation was found to be y = 0.9817x 0.0149 and r2 = 0.99. In recovery experiment, % recovery was found to be 81. In hemolysis interference experiment, interference was detected for the hemoglobin concetrations above 200 mg/dL, Method comparison experiment was performed by analyzing 100 patients serum in both Dade Behring BNII nephelometry and Roche Cobas Integra 800 analyser. Correlation factor was r2=0.95, Deming regression equation was y = 0,98x + 0,22. Conclusion: It was found for turbidimetric method, CV values were higher, % recovery was lower and hemolysis interference was detected at lower concentrations than what was stated in package insert of cystatin C kits. However, Deming regression results indicated turbidimetric and nephelometric methods were compatible.