|
Activity Number:
|
522
- Contributed Poster Presentations: Biometrics Section
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #308009
|
|
|
Title:
|
A Joint Model for the Analysis of Recurrent Events and a Dependent Terminal Event: Application to a Large Cardiovascular Outcomes Trial
|
|
Author(s):
|
Shahidul Islam* and Carl Rosenberg and John C. LaRosa and Michael Szarek
|
|
Companies:
|
SUNY Downstate Medical Center and SUNY Downstate Medical Center and SUNY Downstate Medical Center and SUNY Downstate Medical Center
|
|
Keywords:
|
|
|
Abstract:
|
While analyses of cardiovascular (CV) outcomes clinical trials typically focus on the first event in a composite of related nonfatal and fatal events, patients may experience multiple nonfatal events during follow-up, and terminal events interrupt the nonfatal events process. Methods that can account for multiple nonfatal events and the possible dependency between nonfatal and fatal events can therefore provide added insights into a disease process beyond an analysis of first events. The purpose of this project was to compare the shared frailty model (SFM) and joint frailty model (JFM) in analyzing nonfatal and fatal events to conclude when JFM is warranted over SFM. We applied these models to nonfatal and fatal events observed in a large CV trial. Simulation study demonstrated significant bias when using SFM compared to JFM when there is an association between risk of nonfatal and fatal events. JFM quantified the strength of the association between nonfatal and fatal events by analyzing them jointly. We demonstrated that JFM is warranted when there is an association between nonfatal and fatal events and ignoring this association can produce inaccurate results for nonfatal events.
|
Authors who are presenting talks have a * after their name.
