Standardization of a commercial (Boehringer Mannheim diagnostics) enzymic method for cholesterol.

The Laboratory Standardization Panel of the National Cholesterol Education Program recommends that cholesterol method accuracy ideally be within 3% of the true value determined by the Abell-Kendall Reference Method, a component of the National Reference System for Cholesterol. As one of the Abell-Kendall network laboratories established to facilitate cholesterol standardization, the approach we recommend for determining accuracy involves a comparison analysis on patients' specimens by the method in question and by the Abell-Kendall method. Use of fresh specimens precludes matrix interactions that may influence enzymic measurement. Using this approach, we assessed an enzymic method for cholesterol with two instruments (Boehringer Mannheim/Hitachi 717 and 737), with BMD reagent, controls, and calibrator. Fresh and frozen sera were analyzed with both instruments over three days. The Abell-Kendall method was used at the Northwest Lipid Research Center on frozen aliquots of the same sera. Both instruments demonstrated good agreement with the Reference Method, as determined by linear regression; overall bias averaged less than -2% for the Hitachi 717 and -1% for the Hitachi 737 at 2000 mg/L--i.e., within the accuracy recommendation. We observed a difference in bias for fresh and frozen specimens; with the Hitachi 717, fresh specimens exhibited -3% bias at 2000 mg/L, but there was virtually no bias of determinations of frozen specimens.