|
Activity Number:
|
243
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #307571 |
|
Title:
|
Using Objective Measures in Combination with Self-Report To Estimate Adherence
|
|
Author(s):
|
Richard Thompson*+ and Michael Griswold and Arlene Butz and Michele Donithan
|
|
Companies:
|
Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins University and Johns Hopkins Bloomberg School of Public Health
|
|
Address:
|
3133 Guilford Ave., Baltimore, MD, 21218,
|
|
Keywords:
|
sensitivity ; missing data ; self-report data ; study design ; electronic monitoring
|
|
Abstract:
|
Asthma is a major health risk for those under the age of 18. Adherence to prescribed asthma medication is often assessed purely through self-report, which is known to give biased estimates. We present methods to account for biases in self-report data through objective measures which are also surrogates for the outcome of adherence. We address this by: 1) relating the objective measure to the self-report measure, 2) adjusting the self-reported results, and 3) performing sensitivity analyses around assumptions on informative missingness mechanisms. Motivating data come from self-reported diary data of nebulizer use for rescue medication in conjunction with electronic monitoring. Our goal is to improve estimation and guide study design using self-report data. Specifically, given limit resources, what percent of the sample should be monitored in order to improve self-reported outcomes?
|
