On the basis of the previous interim report of this committee, it became apparent that there was a big variation in the measured values of HbA₁c among the participants and this large interlaboratory variation was mainly due to differences in evaluation of a labile HbA₁c component and in values measured with an instrument of one manufacture and with the other. In the present study we investigated whether standardizing to measure only stable component using common calibrators was useful to reduce the large interlaboratory variation. If it is so, the reference intervals of HbA₁c value will be established using the standardized procedure. The committee sent a set of materials (two lyophilized erythrocyte hemolysates, two similarly processed calibrators and each one of fresh blood samples obtained from a healthy person and a diabetic patient) to 110 institutes and asked to measure stable HbA₁c percentage in the materials. To assign HbA₁c percentages to calibrators, we used HbA₁c results from 100 participants who sent us the data good enough to evaluate judgingc from their chromatograms. Two-point calibration with assigned values improved interlaboratory variation as follows; CV with and without calibration, lyophilized hemolysate 1. 1.98% and 5.33%, lyophilized hemolysate 2, 1.43% and 3.24%, normal fresh blood, 3.88% and 7.76%, and diabetic fresh blood, 2.82% and 3.53%.
We concluded that this standardized procedure i.e. measuring only stable HbA₁c component and correcting the HbA₁c percentages thus obtained with the assigned value of the calibrators, is useful to reduce the interlaboratory variation to the clinically desirable level of 4.0% which was calculated from data of the questionnaire to about 100 qualified diabetologists in Japan.
The reference intervals of HbA₁c percentage calculated from HbA₁c results of 725 healthy subjects measured using the standardized procedure was from 4.3% to 5.8%.