engleski

Comparison of HbA1c results by two immunoturbidimetric methods

Background. The measurement of HbA1c as a biomarker for long-term monitoring of glycemic control in diabetic patients demands high analytical precision. The aim was to evaluate analytical performance of immunoturbidimetric method for measurement of HbA1c (Tina-quant Hemoglobin A1c Gen. 2)on Cobas Integra 400 plus analyzer (Roche) before introduction into routine use. Methods: Within-run and day-to-day imprecision was determined using two commercial control samples (PreciNorm and PreciPath, Roche) and one patient sample with value near the target point for good glycemic control (7% ; 54 mmol/L). Method comparison was obtained with parallel measurement of patients samples (N=75) on Dimension Xpand (Siemens) analyzer (immunoturbidimetrical method traceable to DCCT) used routinely in our laboratory. Statistical analysis was performed using MedCalc 12.2.1. software. Results: Coefficients of variation for within-run imprecision ranged (N=25): PreciNorm 0.95%, PreciPath 0.82%, and patient sample 1.16%. Coefficients of variation for day-to-day imprecision ranged (N=20): PreciNorm 0.95%, PreciPath 0.54%, patient sample 0.72%. Cumulative intralaboratory imprecision for low and high commercial controls and patient sample was 1.4%, 1.0% and 1.3% respectively. The correlation coefficient (r) between two analyzers was r=0.98. The parameters of Passing-Bablok regresion showed systematic differences among analytical systems (95% CI, a=-1.343 t0-0.400 ; b=1.000to1.143) ; Systematic differences evaluated by Bland-Altman method were Cobas-Dimension (95%CI=-0.93 t0 0.28). Conclusion: Method demonstrated high reproducibility meaning that is suitable for long-term follow-up of diabetic patients. If method is changed, constant bias should be corrected for improved results agreement.

methods; comparison HbA1c

nije evidentirano