ABSTRACT
Several recent studies to evaluate the performance of laboratory instruments have shown that with some instrument systems processed (lyophilized, frozen, and stabilized) materials exhibit matrix effects that cause the assay for cholesterol to respond differently for them than for patient specimens. To understand this phenomenon better the College of American Pathologists, Northfield, III, and the Centers for Disease Control, Atlanta, Ga, have conducted a collaborative study with 44 laboratories where 16 instruments manufactured by nine companies are evaluated. The purposes were to assess measurement variation on several reference materials used for standardizing total cholesterol measurements and to evaluate a new stabilized liquid serum as a potential reference material. Lypophilized, frozen, fresh-frozen, and stabilized materials at three concentrations were measured for total cholesterol. The results show that the average coefficient of variation of measured total cholesterol for all instruments, laboratories, vials, and replicates is 3.6% to 4.1% for each of the materials measured (excluding the results for one instrument). For one instrument, however, significant bias was found on the stabilized liquid serum material. Results from the fresh-frozen materials indicate that the instrument systems evaluated allow laboratories to attain the National Cholesterol Education Program analytical performance goals.
