Activity Number:
|
154
|
Type:
|
Contributed
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
Sponsor:
|
WNAR
|
Abstract - #310408 |
Title:
|
Identification of Treatment Effects in the Presence of Deviation from Randomized Assignment: Application to a Surgical Setting
|
Author(s):
|
Corwin Zigler*+
|
Companies:
|
University of California, Los Angeles
|
Address:
|
11610 Iowa Ave, Los Angeles, CA, 90025,
|
Keywords:
|
instrumental variables ; noncompliance ; latent class model
|
Abstract:
|
Often in medical settings, adherence to randomized assignment of competing treatments proves unethical or simply unrealistic. Without advantages of randomization, standard approaches may not allow useful comparison of treatment effects. An instrumental variables framework has emerged as a useful tool for identifying treatment effects in settings where treatment received differs from treatment assigned (Angrist, Imbens, and Rubin 1996, JASA and Imbens and Rubin 1997, Annals of Statistics). We apply this framework in an example involving surgical versus non-surgical treatment of mandibular fractures in which clinicians were allowed to override random treatment assignment. We use Bayesian methods that utilize data characterizing patient injury to estimate the effect surgical treatment on length of hospital stay.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |