Analysis of large datasets

The quality and scope of large clinical and administrative datasets continues to improve, particularly in the UK which is home to the world’s largest joint register and the Patient Reported Outcome Measures (PROMs) programme which collects preoperative and postoperative data on four common elective surgical procedures. Both databases now include over one million patients. These resources offer the opportunity to observe the impact of interventions in real world settings and at a scale infeasible for clinical trials. However, exploitation of large observational datasets brings its own challenges, not least the mitigation of selection effects. Our research has focussed on the following aspects:

Extrapolation of event rates

Estimation of treatment effects

Estimating resource use

Key people: Mark Pennington, Zia Sadique

Comments are closed.