Design & Analysis of Cluster Randomised and Stepped Wedge Trials
The course will have a hybrid format and students can choose to attend in person in London or online.
This course will provide attendees with a clear understanding of how to design and analyse cluster randomised trials and stepped wedge trials. Cluster randomised trials and stepped wedge trials are increasingly used to evaluate health and social interventions. This course will cover all aspects of designing and analysing these trials. This includes the rationale for using these designs, specific design issues, the randomisation process, sample size calculations, analytical methods, ethical considerations, trial reporting, and interpretation. We will include case studies from low, middle and high-income settings. Participants will gain practical experience of data analysis using the computer package Stata.
By the end of this course, attendees will be able to critique and apply a range of appropriate design approaches and analytical methods for cluster randomised and stepped wedge trials.
The course will cover:
- Key concepts of cluster randomised trials, including measures of between-cluster variation and the rationale for cluster randomisation
- Design of cluster-randomised trials, including cluster selection and randomisation procedures
- Calculation of sample size
- Analysis of cluster randomised trials using both cluster-level summaries and individual-level data (random effects models and generalised estimating equations)
- Design and analysis of stepped wedge trials
- Ethical considerations, data monitoring and reporting of CRTs
The course will have a hybrid format and students can choose to attend in person in London or online. Teaching will take place through a blend of lectures and hands-on, live, practical sessions. There will be a strong emphasis on the practical exercises where participants will have hands-on experience using Stata to analyse illustrative datasets from a variety of trials.
Students attending in person who do not have their own laptop will be given access to one for the practical sessions. For those attending online, sessions will take place using Zoom and therefore participants will need to have a reliable internet service and a PC or laptop with microphone and speakers. Students on the course without access to Stata will be given a temporary Stata licence for the duration of the course.
The course will take place from approximately 9:30am to 5.00pm British Summer Time (BST).
We require students to have a working knowledge of standard statistical methods used in epidemiological analyses (e.g. linear, Poisson and logistic regression) and of clinical trials. The course is taught in Stata. Detailed explanations of how to implement the methods in Stata will be provided, but applicants should have basic knowledge of Stata. Applicants should have a good command of English.