10 Apr 2017 09:30am to 10:30am

Meta-analyses pooling results from different clinical trial designs

Event Location
Francois Curtin
University of Geneva

In the development of drugs or therapeutic interventions, the use of trials with different designs is frequent. In particular parallel group and cross-over trials are often used. When there is a need to pool the results of such studies into a meta-analysis, combining results from these different trials is not straightforward. We present first simple methods based on weighted average approaches to combine metrics from parallel and cross over trials. Regression methods based on generalized estimating equations as well as multilevel and Bayesian methods introducing random effects allow to combine the trial designs with the possibility to adjust for trial design specific parameters such as carry-over effect. Comparisons of the methods based on examples are presented.  Although statistical methods are available to pool results, the main limitation of the pooling exercise is the availability of the data, in particular those of cross-over trials, in order to implement the models successfully.