To the Editor.—
The accuracy of the Du Pont cholesterol-level determination methodology for the aca discrete clinical analyzer was recently challenged by Blank and coworkers1 in the Nov 28, 1986, issue of JAMA. These authors claim that the Du Pont methodology shows a bias of 19% when compared with the Lipid Research Clinics reference methodology. Their calculation of bias was based on a complex series of correlation studies that involved different sample types analyzed on a number of different instruments at different times. The authors did not conduct a direct comparison of the Du Pont and reference methods.
We have conducted a thorough evaluation of the Du Pont cholesterol-level determination method as part of our participation in the national effort to improve the standardization of cholesterol-level determination methodologies. This study involved the analysis in triplicate of 45 fresh serum pools prepared from patient samples having similar cholesterol concentrations
Schifreen RS, Dobler GL, Delaria DM. Measuring Serum Cholesterol. JAMA. 1987;257(12):1602-1603. doi:10.1001/jama.1987.03390120064022