The ability to utilise external information when designing and analysing clinical trials brings many benefits, such as maximising the evidence provided by the trial, reducing the sample size required (particularly important for rare disease trials) and improving the generalisability of trial results. This course provides participants with a range of the latest statistical methods that can be used to incorporate external information and thus improve the efficiency and robustness of clinical trials. We also cover potential biases and pitfalls that may arise, and how to address them.

The following topics will be included:

• Bayesian methods that form prior distributions from elicited and (multiple) external data sources.
• Bayesian and hybrid approaches (e.g. assurance) that account for uncertainty in sample size calculations.
• Methods that facilitate borrowing of historical information or data from within the same trial (e.g. master protocols).
• Frequentist methods (e.g. propensity score weighting) that use external data, such as cohort studies and routinely collected healthcare records, to form synthetic control groups and generalise results from less representative trials to wider patient populations.
• Application to real clinical trials, including trials for rare diseases.

As well as the necessary theory, we will cover computational approaches to implement the methods and practical issues, such as funder and regulator views.

Course Tutors:
Professor James Wason, Biostatistics Research Group, Newcastle University
Dr Faye Williamson, Biostatistics Research Group, Newcastle University
Dr Kevin Wilson, School of Mathematics, Statistics & Physics, Newcastle University