|
Activity Number:
|
119
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 9, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #300803 |
|
Title:
|
Bias in a Placebo-controlled Study Due to Mismeasurement of Disease Status and Regression Effect
|
|
Author(s):
|
Hung-Mo Lin*+ and Robert H. Lyles and John M. Williamson
|
|
Companies:
|
Pennsylvania State University and Emory University and Centers for Disease Control and Prevention
|
|
Address:
|
College of Medicine, A210, Health Eval Sci., Hershey, PA, 17033,
|
|
Keywords:
|
bias ; misclassification ; placebo-controlled trial ; regression to the mean ; intervention
|
|
Abstract:
|
We raise the concern of whether the use of a placebo group in a randomized clinical trial is sufficient to eliminate bias in the assessment of the effectiveness of a drug when enrollment into the trial prior to intervention requires diagnosis of a dichotomous disease, and the diagnostic test is subject to uncertainty. Due to misclassification and the regression effect, the observed difference in the proportions of diseased individuals between the treatment and placebo groups at follow-up will be equal to the true difference multiplied by the positive predictive value at screening and the difference between the sensitivity and the false-positive value at follow-up. Thus, measurement error of disease status before and after administering the intervention attenuates the intervention effect. We provide the statistical methods required for variance estimation and inference about the intervention effect after correcting for RTM. Validation data corresponding to both the screening and follow-up conditions are necessary to provide additional information on the validity of the diagnostic test.
|
