ASSAY VALIDATION USING THE CONCORDANCE CORRELATION-COEFFICIENT

被引:328
作者
LIN, LIK
机构
[1] Baxter Healthcare Corporation, Round Lake, IL 60073
关键词
ACCURACY; CONCORDANCE; PRECISION; REPRODUCIBILITY;
D O I
10.2307/2532314
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
A new reproducibility index, called the concordance correlation coefficient, has been proposed (Lin, 1989, Biometrics 45, 255-268) to evaluate the reproducibility of an assay method, or instrument. This index includes measurements of precision and accuracy. Based on certain criteria about the allowable losses in precision and accuracy, one can compute the sample size requirement. One can also meaningfully validate the reproducibility using the confidence interval approach. For illustration, an application using actual data is given.
引用
收藏
页码:599 / 604
页数:6
相关论文
共 4 条
[1]   WHY PROOF OF SAFETY IS MUCH MORE DIFFICULT THAN PROOF OF HAZARD [J].
BROSS, ID .
BIOMETRICS, 1985, 41 (03) :785-793
[2]   SIGNIFICANCE TESTING TO ESTABLISH EQUIVALENCE BETWEEN TREATMENTS, WITH SPECIAL REFERENCE TO DATA IN FORM OF 2 X 2 TABLES [J].
DUNNETT, CW ;
GENT, M .
BIOMETRICS, 1977, 33 (04) :593-602
[3]   A CONCORDANCE CORRELATION-COEFFICIENT TO EVALUATE REPRODUCIBILITY [J].
LIN, LI .
BIOMETRICS, 1989, 45 (01) :255-268
[4]   HOW TO ESTABLISH EQUIVALENCE BETWEEN TREATMENTS - A ONE-SIDED CLINICAL-TRIAL IN PEDIATRIC ONCOLOGY [J].
RODARY, C ;
COMNOUGUE, C ;
TOURNADE, MF .
STATISTICS IN MEDICINE, 1989, 8 (05) :593-598