Quality control survey for insulin, thyroxine, TSH, cortisol, gastrin, digoxin, beta-2 microglobulin and IgE including 249 institutes in Japan: a big variation due to kit to kit difference

The assay values using different commercially available kits were compared. To assess the quality of the assay, standard curves for the assay were analysed by a data-processing program. Quality of the assay in terms of precision profile was significantly different from kit to kit, as representatively shown is insulin radioimmunoassay. In contrast, the quality of the assay was not significantly different between users and kit-manufacturers. This results indicates that the assay technique of the users was in the same level as in kit-manufacturers. Furthermore, the fact that the ''between assays'' variation was almost identical with the ''within kits'' variation warrants similar technical levels among the different institutes. Thus, to improve the assay results, quality of kits must be improved. The magnitude of variation of the assay value due to difference in kits (''between kits'') significantly exceeded that due to other factors in insulin, TSH, T4 cortisol and gastrin assays. Among many reasons for this big variation, difference in standards was obvious in insulin and gastrin assay. This result calls for some international efforts to set up a common reference standard for each radioimmunoassay to minimize kit to kit variation