Activity Number:
|
325
- Semiparametric Regression in Matched Case-Crossover Studies
|
Type:
|
Topic Contributed
|
Date/Time:
|
Tuesday, August 1, 2017 : 10:30 AM to 12:20 PM
|
Sponsor:
|
ENAR
|
Abstract #323078
|
View Presentation
|
Title:
|
Semiparametric Bayesian Models for Unbalanced Matched Case-Crossover Approach
|
Author(s):
|
Inyoung Kim* and Feng Guo
|
Companies:
|
Virginia Tech and Virginia Tech
|
Keywords:
|
Case-crossover ;
Bayesian semiparametric ;
Time-Variant Risk Factor
|
Abstract:
|
We proposed a two-level mixed effect approach for unbalanced matched case-crossover studies with binary outcome. The proposed model accounts for information about variation within and between strata. We showed that the parameter estimation under properly specific distribution for the random effect is consistent. We proposed two alternative semiparametric models, a semiparametric Bayesian model with Dirichelet prior for the within stratum random effect and a model just specific the first two moments of the random effect. We developed two alternative appraoch to calculate the marginal likelihood of the conditional probability and showed that they are assymtotically equivalent. Our approach overcomes three major limitations comparing with two standard approaches for matched case-crossover studies.
|
Authors who are presenting talks have a * after their name.