Abstract
The measurement of plasma cholesterol is subject to particularly stringent requirements of bias and imprecision because clinically important action limits lie within the bulk of the population distribution. A pilot scheme was conducted with the aim of improving the accuracy of the assay of cholesterol in 26 laboratories in Scotland and Northern Ireland. This involved the distribution of three sets of authentic human sera with varying cholesterol concentrations assigned by the reference Abell-Kendall method which was established and validated in a central laboratory.
The precision (coefficient of variation) of cholesterol testing was 1·71% across all laboratories at the first distribution and this did not alter subsequently. Mean bias was initially 4·07% (median 4·14%) relative to the Abell-Kendall assigned values and this figure had fallen to 1·12% (median 0–92%) by the third distribution.
We conclude that through a standardization scheme it is possible in a relatively short time to improve bias significantly and meet the criterion of < 2–3% bias required for optimum cholesterol measurement.