Minimizing interlaboratory variation in routine assays of serum cholesterol through the use of serum calibrators

A cholesterol standardization program was developed in The Netherlands for clinical laboratories that use their analytical results as indicators of cardiovascular risk. Participants with sufficient precision but inaccurate results are encouraged to use serum-based calibrators as a means to decrease bias. A regional pilot survey and thereafter a national one were carried out in 31 and 138 laboratories, respectively, to investigate whether use of the serum-calibration procedure could improve routinely performed cholesterol measurements. To participants in our national quality-control surveys were mailed three calibrators (pooled human serum) and five human-serum controls, for analysis of cholesterol on four different days. Reference method values had been assigned to the calibrators by the Netherlands Reference Laboratory, applying the U.S. Centers for Disease Control modification of the Abell-Kendall procedure. Using the calibration curves based on results for the serum calibrators decreased the between-laboratory variation (SD) by at least 50%, with improved accuracy for non-enzymatic methods. The initial within-laboratory precision was generally good, and improvements in it were small. To our knowledge, standardization results involving so many routine hospital laboratory procedures have not been previously reported.