Other Lecture or Seminar

Statistical Methods in Assay Validation

Dr. Wendell C. Smith
Eli Lilly & Company

Start Date and Time: Mon, 17 Sep 2001, 4:30 PM

End Date and Time: Mon, 17 Sep 2001, 6:00 PM

Venue: LAEB B222

Abstract:

Development of a validated assay to support pharmaceutical drug development is a multi-faceted activity that provides many interesting challenges for laboratory scientists and statisticians. Many of these challenges are common to all analytic dilution assays. In this presentation, the "similarity" condition which requires a test sample to behave as though it were a dilution of the standard preparation will be reviewed as the prerequisite for fundamental validity of an analytic method. The impact of analytical and statistical sources of invalidity (error) on the accuracy and precision of test results will be illustrated for an ELISA assay. Statistical applications related to the estimation of assay performance characteristics and to the assessment of method suitability will be illustrated, with specific consideration given to the validation of in vitro bioanalytical quantitative assays.