Proyectos
Statistical Methods for Cluster Randomized Trials

The increasing popularity of cluster-randomized trials has created many opportunities and challenges for statisticians. In recent years a variation of the traditional cluster randomized trial has been proposed called stepped wedge design trial. However, several questions have been raised about the advantages and disadvantages of the stepped wedge design compared to the traditional cluster randomized design, in particular with regard to difference in power and sample size needed to properly measure the effect of an intervention. In this project, we will develop statistical methods to 1) study the advantages/disadvantages of the stepped wedge design vs. the cluster-randomized trial, 2) develop statistical models for bounded or even fractional response variables, 3) adjust them to the analysis of a stepped wedge study and 4) develop guidelines for data analysis and reporting of stepped wedge studies. In addition, we plan to implement the methods in freely available software, which will allow it to be used by other statisticians and public health researchers. Our proposal searches not only to give a statistical insight in the discussion above but also to propose new flexible statistical models in the analysis of the response variable when we deal with cross-sectional and longitudinal data.

Fecha de inicio: 01/02/2014
Fecha final: 31/07/2015
Estado DGI: En proceso
Instituciones Investigadoras:
PUCP
Instituciones Financiadoras:
DGI-PUCP
The University of Washington