Title: Choosing covariates in the analysis of cluster randomised trials

Co-supervisors: Dr S Bremner&  Professor S Eldridge

This studentship focuses on cluster randomised trials, which are
increasingly popular in health services research. A gap in knowledge in
relation to such trials is determining which covariates to include in the
trial analysis. Appropriate procedures are better understood in individually
randomised trials, but in cluster randomised trials covariates can occur at
two levels, the cluster and the individual level, and there may be
complicated relationships between covariates at different levels. There has
been a small amount of work in this area but a comprehensive review is
needed particularly for non-continuous and non-normally distributed
outcomes.

This studentship offers the chance to explore existing literature in this
area both for individually and cluster randomised trials, to use data from
some of the many cluster randomised trials within the PCTU to understand
empirically the relationships between different covariates and outcomes, and
use simulation to explore these relationships further. The student will gain
knowledge and experience in these three areas, and develop highly relevant
guidelines for clinical trial investigators, a priority area for NIHR
research.

For an informal discussion about this project, please contact Dr Stephen
Bremner
Tel: 020 7882 2547  Email:s.a.bremner@qmul.ac.uk

The closing date for applications is 14 February 2011. To apply, please
download the School of Medicine&  Dentistry application form from our
website:http://www.qmul.ac.uk/postgraduate/apply/index.html
and return to the Graduate School Office - email:pgsmd@qmul.ac.uk
Please note that we accept postal and electronic applications.
Referees may email references directly topgsmd@qmul.ac.uk