Activity Number:
|
320
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 14, 2002 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Biopharmaceutical Section*
|
Abstract - #301074 |
Title:
|
Comparison of Methods to Analyze Coarse Immunogenicity Data
|
Author(s):
|
William Wang*+ and Eric Zhi and Ivan Chan
|
Affiliation(s):
|
Merck Research Laboratories and University of Minnesota and Merck Research Laboratories
|
Address:
|
785 Jolly Road, Building C, Blue Bell, Pennsylvania, 19422, US
|
Keywords:
|
coarse data ; immunogenicity ; maximum likelihood ; multiple imputation
|
Abstract:
|
Coarse data arise in vaccine clinical trials when an immunologic response is measured by a serial dilution assay, which may give a range of response instead of the exact value in the absence of standard calibration curve. A conventional method treats the lower limit of the range as the exact value in estimating the population mean immunologic response or in treatment comparisons. Ignoring the data coarseness, this method may cause bias and underestimation of the variance of parameter estimates. In this talk, we explore some alternative methods for analyzing coarse data, including the maximum likelihood (ML) method and the multiple imputation method. We carried out simulation studies to compare the performance of these methods under the log-normality assumption. The results suggest that the ML method for coarse data performs best under a wide variety of parameter settings.
|