A new approach for clinical biological assay comparison and standardization: application of principal component analysis to a multicenter study of twenty-one carcinoembryonic antigen immunoassay kits.

BACKGROUND Principal component analysis (PCA) is a powerful mathematical method able to analyze data sets containing a large number of variables. To our knowledge, this method is applied here for the first time in the field of medical laboratory analysis. METHODS PCA was used to evaluate the results of a blind comparative study of 21 carcinoembryonic antigen (CEA) reagent kits used to determine CEA concentration in a panel of sera from 80 patients. RESULTS The mathematical technique first eliminated the variations attributable to the use of different calibrators. The PCA representation then gave a global view of the dispersion of the kits and allowed the identification of a main homogeneous group and of some discrepant kits. CONCLUSIONS PCA applied to the in vitro diagnostic reagent field could contribute to the standardization process and improve the quality of medical laboratory analyses. A standardization method using a panel of patient sera is proposed.