|
Activity Number:
|
433
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
ENAR
|
| Abstract - #305579 |
|
Title:
|
Statistical Model and Sample Size Determination for the QuitNexus Study: Accounting for Within-Cluster Variability in the Denominator of the Counseling Rate
|
|
Author(s):
|
Tina D. Cunningham*+ and Robert E. Johnson
|
|
Companies:
|
Virginia Commonwealth University and Virginia Commonwealth University
|
|
Address:
|
Department of Biostatistics, Richmond, VA, 23298-0032,
|
|
Keywords:
|
cluster randomized trial ; sample size
|
|
Abstract:
|
QuitNexus is a referral system that supports tobacco cessation. A cluster-randomized controlled trial has been proposed to study the effectiveness of the model. The primary outcome is practices' rate of delivery to their population of adult smokers of one or more counseling sessions by the state quitline during a one year period. This rate is comprised of three parts: number of patients enrolled in quitline, adult population size, and proportion of the adult population who are current smokers. This proportion cannot be identified via the practices' data and must be estimated from a sample. The number of current smokers constitutes the denominator of the rate and thus usual generalized linear models do not apply. We present a model for the rate of enrollment, allowing for across-practice and within-practice variation and give advice concerning sample sizes at practice and patient levels.
|
