A0348
Title: Conditional copula models for correlated survival endpoints in individual patient data meta-analysis
Authors: Takeshi Emura - The Institute of Statistical Mathematics (Japan) [presenting]
Virginie Rondeau - Inserm (France)
Sofeu Casimir - University of Bordeaux (France)
Abstract: Existing copula models for the joint distribution of two failure times impose a simplifying assumption: the measure of correlation does not depend on covariates. This assumption is problematic when one tries to measure Kendall's tau between surrogate endpoint and true endpoint in a meta-analytical setup. We suggest extending the existing copula models so that Kendall's tau depends on covariates. Our newly proposed model, a joint frailty-conditional copula model, can effectively implement meta-analyses. In order to facilitate our approach, we develop an original R function "condCox.reg" and make it available in the R package joint.Cox. We apply the proposed method to a gastric cancer dataset with 3288 patients in 14 randomized trials from the GASTRIC group. This data analysis concludes that Kendall's tau has different values between the surgical treatment arm and the adjuvant chemotherapy arm.
