Results of a province-wide quality assurance program assessing the accuracy of cholesterol, triglycerides, and high-density lipoprotein cholesterol measurements and calculated low-density lipoprotein cholesterol in Ontario, using fresh human serum.

To evaluate laboratory performance, eight to 13 samples of fresh human serum from volunteers were sent to 250 laboratories in the Canadian province of Ontario licensed to perform lipid analysis. Fresh human specimens were used because of potential matrix effects with processed materials. We show that on all survey samples, 71% (range, 63% to 82%) of participating laboratories are within +/- 5% of the target cholesterol value and that 93% are within +/- 10%. The goal of the National Cholesterol Education Program for 1992 is total error of no more than +/- 9% for 95% of results. The unblanked triglycerides results show that on all samples 40% (14% to 59%) of participants are within +/- 5% and 68% (range, 31% to 86%) are within +/- 10% of the target value. For triglycerides results from 0.9 to 2.0 mmol/L, 80% or more are within +/- 0.2 mmol/L. Between 2.0 and 3.0 mmol/L, 90% are within +/- 0.3 mmol/L of the target values. For high-density lipoprotein cholesterol, for all samples 35% (range, 24% to 50%) of laboratories are within +/- 5% and 68% (range, 55% to 88%) are within +/- 10%. A range of 80% to 95% of participants are within +/- 0.2 mmol/L of the target values. For calculated low-density lipoprotein cholesterol, 51% and 62% of the laboratories surveyed are within +/- 5%, with 83% and 89% within +/- 10% of the target values. We conclude that the laboratory measurement of lipids is approaching the degree of accuracy and precision required for clinical purposes, and that the use of fresh human serum samples is a viable approach to their proficiency testing.