Statistical Analysis

Statistical Analysis Working Group 

 

Statistical Analysis Working Group Co-Leads: Richard Emsley (King’s College London) and Beatriz Goulao (University of Glasgow)

To join: Complete the SAWG Expression of Interest form and send to methodologyhubs@liverpool.ac.uk

 

Objectives

The group’s aim is to improve methods for the design and analysis of randomised controlled trials (RCTs)  through collaborative methodological research and dissemination. The SAWG will support work on statistical design and analysis methods that could be written into a trial’s statistical analysis plan (rather general methods that might be applied to secondary trial data).

Its main objective is to facilitate collaboration among individuals working on similar topics relating to statistical analysis of randomised trials, with the aim of adding value to each other’s work, strengthening links between groups and identifying new areas for research. The group is focusing its efforts on preparing tutorial papers for applied journals and mainstream medical journals; presentations and lectures to increase uptake of methods amongst stakeholders; issuing guidance on methods (when to use them, how to use them, when they break); developing software to implement new statistical methods where required; teaching short courses.

We will develop a principled approach to the process of analysing trial data from the inception of study that can be used as a training resource for future clinical trial statisticians.

Research areas to target

  • AI and Automation in Statistical Analysis
  • Leveraging External and Routine Data
  • Advanced Outcome Measures
  • Environmental and sustainability considerations in design and analysis of clinical trials
  • Treatment Effect Heterogeneity and Subgroup Analysis
  • Estimands and Causal Inference
  • Handling Covariates and Missing Data
  • Statistical Analysis Plans (SAPs)
  • Efficient and Complex Trial Designs
  • Stakeholder Involvement in Statistical Analysis

 

Membership 

Members will include individuals and research groups working on developing and applying methods for statistical and design analysis and those with an interest in using these methods; for example, trial statisticians interested in discussing and finding solutions. However, members do not have to be statisticians and can include trial methodologists, clinicians or public partners with an interest in trial statistical design and analysis research.

We plan to have a ‘core’ group of up to 10 individuals (actively working on statistical methods) who will meet every 3 months. There will be a wider group of members and stakeholders for annual meetings.

 

Collaborations

  • UKCRC Registered Clinical Trials Unit (CTU) Network Statistics Working Group.
  • Funders (UKRI and NIHR) and charities with funding panels.
  • NIHR Statistics Group.
  • Other TMRP working groups (especially outcomes, stratified medicine, and adaptive designs.
  • Novartis UK.