|
Activity Number:
|
341
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305832 |
|
Title:
|
Tolerance Regions for Multivariate Random Effects Model with Application to Beat-to-Beat QT-TQ Interval Data
|
|
Author(s):
|
Kimberly Crimin*+ and Joseph McKean
|
|
Companies:
|
Pfizer Inc. and Western Michigan University
|
|
Address:
|
Eastern Point Road, Groton, CT, 06340,
|
|
Keywords:
|
QT prolongation ; GEE ; tolerance regions ; random effects
|
|
Abstract:
|
QT prolongation has been responsible for the withdrawal and re-labeling of numerous compounds, making it important to understand how the ECG signal varies over time. The development of automatic ECG signal acquisition systems has made it possible to collect continuous ECG data. The duration of any cardiac cycle is largely dependent on the preceding diastolic interval, measured by the TQ interval. When the TQ interval is extinguished, the probability of an arrhythmia increases. Tolerance regions are used in clinical applications to define the 'normal' range of clinical or laboratory measurements. We present a method to calculate tolerance regions for bivariate random effects models and apply these methods to continuous QT-TQ interval data collected on 'healthy' individuals. The variance was estimated using GEE. We investigate methods for computing tolerance factors in a MC study.
|
