ABSTRACT
Objective To investigate the accuracy and comparability of ?-glutamyltransferase (?- GT) measurement results on human serum samples and controI materials before and after calibration with a common human serum calibrator.Methods A human serum calibrator was prepared by pooling fresh serum aliquots and assigning a value for ?-GT with the IFCC reference method.The calibrator together with 5 human serum samples and 10 control samples were sent to 15 clinical laboratories and the sermn and control samples were measured with different analytical systems before and after a calibration with the calibrator.The results were analyzed for biases and interlaboratory variations.Results For the serum samples,the calibration resulted in reductions in biases from -9.0%～-14.2% to -0.8%～-7.9%,and in interlaboratory variations from 6.9%～11.6% to 2.8%～4.4%.No improvement was observed on the control samples.Conclusions Accuracy and comparability of serum ?-GT measurements can be improved by using a common human serum calibrator.Some control materials may not be commutable for human serum in ?-GT measurements.