Reference values versus consensus values: a reaction to the article of Baldan et al.

Abstract In a recent article Baldan et al. concluded that reference values can be used in a proficiency test as assigned values and that this does not increase the cost of the proficiency test. They feel that the values found during an homogeneity test could be used as reference values. In this article, it is argued that the conclusions of Baldan et al. have only a very limited validity. Due to the difference in goals, homogeneity data cannot be used for the determination of reference values without a drastic change in set-up. And in practice, in many fields of application, the situation is quite different than Baldan et al. suggest. Thus in many proficiency testing studies it is not technically possible to use reference values and in cases where it is possible, the costs will increase significantly if reference values are used instead of consensus values.