Skip to content

Upcoming

Lloc: Sala C5016 (Campus Nord)

A multiple testing procedure for binary and time-to-event endpoints. Marta Bofill

Many promising drugs identified with effectiveness in phase II fail in phase III of cancer trials. Common efficacy endpoints for phase II trials include binary endpoints such as objective response or overall response rate; whereas in phase II trial, the usual primary endpoints are time-to-event such as the overall survival or progression free survival. In this context, the integrated planning for phase II/III trials could play an important role for reducing cost and achieving trials with shorter follow-ups. Motivated by the improvement in the efficiency of phase II trial designs which would improve the success rate in phase III trials, we propose a bivariate statistic test for equality of survival functions and equality of proportions for two-arm controlled trials with binary and time-to-event endpoints.

We consider the comparison of two groups by means of the weighted difference in survival times and the difference in proportions, for the survival and binary endpoint, respectively. The association between the two endpoints to account for the covariance structure of the bivariate test will be discussed.

Lloc: Biblioteca de la secció d'Estadística de la Facultat de Biologia (Edifici Aulari)

Causal Inference. Klaus Langohr and Jordi Cortés

This session aims to support the practicing statistician in making it work: causal inference form observational data in a potential outcomes framework. How to chose among the many versions of exposure, the target population and indeed the estimand in a given setting? Different estimands serve different purposes while lending themselves (in)directly to natual constraints imposed on given datasets. We spend ample time exploring this in various contexts before explaining key features of estimation methods relying on either the `no unmeasured confounders’ assumption or the availability of `instrumental variables’. We talk about some of the following methods: outcome regression and stratification or inverse probability weighting. We illustrate key results and estimation properties with published case studies, using either the original data or simulations. Hands on sessions will guide participants in using R with the data. We also present the daggity platform.

Lloc: Sala C5016 (Campus Nord)

Titol: TBA. Oleguer Plana