Background Reliable data on clinical biomarkers are necessary in order to assess the health risks of populations and especially in assessing long-term trends related to disease incidence.

Methods Ten European laboratories participated in a two-phase quality control exercise of total cholesterol (TC) and high-density lipoprotein-cholesterol (HDL-C) analysis. The European Health Examination Survey Reference Laboratory prepared plasma batches for analysis, and provided target values for them. Two criteria were set for the precision and the systematic error (bias). Three plasma samples were analysed in duplicate on separate days (n=12).

Results In Round 1, all laboratories met the acceptable criterion (3%) for precision of TC. The mean bias of all laboratories was 0.99% (95% CI 0.03% to 1.95%). Six laboratories measured samples from Round 2. Five laboratories met the goal criterion of 3%; one failed to meet the acceptable criterion of 5%. The mean bias for HDL-C of the three batches of six laboratories was within goal limits (±5% from target) and that of all 10 within acceptable (±10%). The mean bias of all laboratories was 1.1% (95% CI −0.18 to 2.32). In Round 2 four laboratories met the goal criterion and one the acceptable criterion.

Conclusion The quality control exercise demonstrated that although the majority of the laboratories met the strict criteria for systematic error for TC and HDL-C, standardisation of methods is still needed to improve the accuracy of biomarker measurements of laboratories engaged in population health surveys. A protocol is recommended for obtaining reliable and comparable biomarker data between countries.