Clinically relevant lot-to-lot reagent difference in a commercial immunoturbidimetric assay for glycated hemoglobin A1c

Abstract

ObjectivesHemoglobin A1c (HbA1c) is employed for diagnosis and therapy monitoring of diabetes mellitus. The effect of a change of reagent lot on the measured values of a commercial immunoturbidimetric HbA1c assay (A1C3) was investigated. Design and methodsComparison measurements of A1C3 and an automated affinity chromatography method (VIIT) were performed in 15 samples for the initial and in 20 samples for the subsequent A1C3 lot. The results of 27 and 19 measurements of a normal and of 28 and 20 of a pathological control (before and after the switch of the A1C3 reagent lot, respectively) were evaluated. Finally, the results of 6463 patient samples that had been measured with the initial and 434 that had been measured with the subsequent A1C3 lot were investigated. ResultsVIIT yielded significantly higher results than the initial A1C3 lot (bias: 0.41% HbA1c, 4.5mmol/mol) but agreed well with the subsequent lot (bias: −0.01% HbA1c, −0.1mmol/mol). Changing to the subsequent reagent lot resulted in significant increases of the mean of the normal control of 0.316% HbA1c (3.5mmol/mol) and of the pathological control of 0.749% HbA1c (8.2mmol/mol). The median of patient samples measured with the subsequent lot was significantly higher by 0.40% HbA1c (4.4mmol/mol). ConclusionsThe subsequent A1C3 reagent lot yields significantly higher measurement results than the initial by approximately 0.5% HbA1c (5.5mmol/mol). This difference is considered as clinically relevant. A combined effort of manufacturers and notified bodies is necessary to minimize lot-to-lot variation.