[Studies on the accuracy and precision of total serum cholesterol in regional interlaboratory trials (author's transl)].

The between-run precision of the Liebermann-Burchard reaction modified by Watson was, in our laboratory, 2-3%, the within-run coefficient of variation was 1-2%. The between-run precision of the enzymatic test was 3-4%, the within-run coefficient of variation was 3%. The regression analysis of 92 serum specimens from patients was y = -17.31 + 1.04 chi, the coefficient of regression was r = 0.996. Interlaboratory trials of serum cholesterol were studied in the normal and pathological range. Lyophilized samples of serum prepared commercially and from fresh specimens from patients were analysed by the method of Liebermann-Burchard as well as by the enzymatic procedure. Acceptable results estimated by Liebermann-Burchard were obtained in the different laboratories after using a common standard of cholesterol. The coefficient of variation of the enzymatic test in the interlaboratory trial was higher in comparison to the Liebermann-Burchard reaction. Methodological difficulties of the Liebermann-Burchard reaction are discussed and compared with the specific, enzymatic assay.