Interlaboratory Difference in GHb Measurement in Japan

The Third Report of the GHb Standardization Committee, the Japan Diabetes Society

Abstract

The Glycohemoglobin (GHb) Standardization Comittee conducted the third external quality assessment to evaluate the interlaboratory variation of GHb values measured by various methods including HPLC, turbidimetric inhibition immunoassay (TINIA) and affinity chromatographic methods in 1, 424 institutes throughout the country using four samples (2 concentrated whole blood and 2 lyophilized hemolysate). In addition, we compared our assay values to Diabetes Control and Complications Trial (DCCT) values by measuring the same 48 fresh blood samples in the laboratory of Tokushima Univ. and in theirs.
We got the following results, 1) The CVs as indices of the interlaboratory variation ranged from 7.1% to 12.8% when all data were combined. It was different depending on an analytical method used: HPLC, 3.7-6.5%; T INIA, 7.6%-11.7%; Affinity 5.6%-6.4%.
2) The means of the measured values were different depending on a method used. The value measured by HPLC (4.7±0.31%) was similar to that by TINIA (4.9±0.56%), but the value measured by an affinity chromatographic method (5.5±0.31%) was signficantly higher than the others.
3) The CVs among the GHb value corrected with the assigned values of Japan Diabetes Society (JDS) calibrators ranging from 4.6% to 7.8%, were much smaller than those corrected with other calibrators or without correction, 8.3% to 15.9%.
4) Our measured values were quite similar to DCCTs': The values measured by Tosoh=0.984 X DCCT-0.228; r=0.998, Ones by Kyoto Daiichi =0.972×DCCT-0.052, r=0.997) In conclusion, the correction of measured GHb values with the assigned values of the calibrators provided by JDS is useful to reduce the interlaboratory variation to a clinically permissible level and to get a good agreement between our values and DCCT's.