My research interests include risk prediction using longitudinal and survival data, meta-analysis, cost-effectiveness analyses for public health interventions and novel designs for early phase trials.

I have worked on the joint modelling of growth and rupture rates in patients with abdominal aortic aneurysm (AAA) and have used the models to predict future risk of rupture for various AAA size categories. Through the synthesis of estimates from a large individual patient data meta-analysis we have been able to recommend cost-effective surveillance intervals for the NHS AAA Screening Programme.

I currently work on statistical methods in the development and assessment of cardiovascular risk scores and their predictive utility and am chief statistician in reviewing proposals submitted to the EPIC-CVD project. I am particularly interested in the comparison of traditional risk scores with dynamic (time-updated) predictions using repeatedly measured risk factors.