|
Activity Number:
|
188
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301535 |
|
Title:
|
Noncompliance-Corrected Effect of Randomized Highly Active Antiretroviral Therapy on Incident Aids or Death Using Inverse Probability-of-Censoring Weights
|
|
Author(s):
|
Lauren E. Cain*+ and Stephen R. Cole
|
|
Companies:
|
Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health
|
|
Address:
|
615 N. Wolfe St. E7133, Baltimore, MD, 21205,
|
|
Keywords:
|
Compliance ; HIV/AIDS ; Inverse probability weight ; Trial, randomized
|
|
Abstract:
|
In 1996--7, AIDS Clinical Trial Group study 320 randomized 1156 HIV+ US patients to combination antiretroviral therapy (ART) or highly active ART with equal probability. In one year of follow up, 96 patients incurred AIDS or died, 51 dropped out, and 290 dropped out or stopped their assigned therapy. Noncompliance likely results in a null-biased intent-to-treat hazard ratio (HR) of AIDS or death comparing highly active ART to combination ART: which was 0.75 (95% confidence limits [CL]: 0.43, 1.31) for follow up within 15 weeks. Noncompliance correction using inverse probability-of-censoring weights yielded a 63% stronger HR of 0.46 (95% CL: 0.25, 0.85). Weights were estimated conditional on randomization arm and measured baseline and time-varying covariates. These methods may help resolve discrepancies between randomized studies with differing amounts of compliance.
|
