Developing and evaluating methods for the analysis of clinical trials

Clinical trials are a powerful tool for evaluating and understanding the effects of healthcare interventions on patients.
A trial’s analysis should address a relevant question and provide results that are informative, robust and make best use of the available data. The Analysis programme develops and evaluates such methods.

Our main areas of work

Covariate adjustment

Covariate adjustment uses background information on trial participants to better estimate the treatment effect. We focus on when, why and how to do covariate adjustment.


Estimands precisely describe the causal effect that we wish to estimate from a clinical trial. Our work focuses on how to choose, describe, and estimate different estimands.

Missing data

Most clinical trials have some missing data. We research methods for finding suitable assumptions about missing data and correctly using them in the analysis.

Sensitivity analysis

Sensitivity analysis checks the robustness of trial conclusions to different assumptions. We work on definitions, methods and application of sensitivity analysis.

Simulation studies

Simulation studies help us evaluate statistical analysis methods. Our ADEMP framework has changed practice in their planning and reporting.

Estimation and causal inference

Different estimands require different methods of estimation, which require different assumptions. We develop and evaluate estimators for different estimands.