Effectof Calibration on Dispersion of Glycohemoglobin ValuesDeterminedby 111 Laboratories Using21 Methods

per method (from 10% to 4% arid from6% to 3%), intermethod variation (from 18% to 4% and from 16% to 3%), and overall interlaboratoryvariation (from25% to 7% and from15% to 4%). Without calibration, 71% of thelaboratories didnotmeetthe clinicallydesirableintralaboratoryCV of 3.5%; calibration reduced this proportionto 39%. We conclude that, irrespective of the analytical method used, calibration greatly reduces all sourcesof GHbvariation.