JdS2012


 English   -  Français  

Résumé de communication



Résumé 65 :

The role of statistical methodology in clinical research - shaping and influencing decision making
Bretz, Frank
Novartis

The clinical development of new drugs has become increasingly challenging, inefficient and costly. It is therefore imperative to consider and implement innovative study designs and statistical analysis methods to increase the information value and improve decision-making. In this presentation we illustrate the opportunities for applying advanced statistical methodology with two examples. First, we consider response-adaptive designs for dose-finding studies. To allocate new cohorts of patients in an ongoing study, we use optimal designs that are robust under model uncertainty. In addition, we use a Bayesian shrinkage approach to stabilize the parameter estimates over the successive interim analyses. This allows us to calculate updated parameter estimates and model probabilities which are then used to determine the optimal design for the new cohorts. The resulting designs are robust with respect to model misspecification and in addition adapt efficiently to the information accrued during an ongoing study. Second, we consider structured multiple hypotheses test problems arising in confirmatory studies, such as comparing several treatments with a common control, combined non-inferiority and superiority testing, assessing the benefit of a drug for multiple endpoints, testing a treatment at different dose levels in an overall and a subpopulation, or any combination thereof. We propose a simple iterative graphical approach to construct and perform advanced multiple test procedures that address the given study objectives. The resulting test procedures are represented by directed, weighted graphs, where each node corresponds to an elementary hypothesis, together with a simple algorithm to update such graphs while sequentially testing the individual hypotheses.