[Problems related to determination of reference interval and clinical decision limit in enzyme immunoassay].

It is generally difficult to determine reference interval in enzyme immunoassay (EIA) at an individual laboratory because of many sources of variations for analytes measured by EIA, low analytical precision and high reagent cost of EIA. Large within-kit variations can be reduced by strict control of assay procedures so that the reference interval can be determined jointly and transferred among users of the same kit. To set clinical decision limits (CDL) it is necessary to consider low frequency of abnormality in tumor markers and hormones when they are measured for case finding. The CDL which minimize an expected cost can be determined when costs for false positive/negative results are given together with pretest probability, sensitivity and specificity.