Job description

Observational data is frequently used to draw causal inferences when a randomised trial is not feasible or ethical. The ‘target trial’ framework attempts to harness design principles from trials to improve the standard of observational epidemiology. Researchers are stimulated to first specify a hypothetical target trial, before trying to emulate it their data, in order to avoid common analytical pitfalls. The approach has helped to resolve conflicting results between observational and randomised studies in studying the effect of hormone therapy on coronary heart disease. It has been applied more recently to guide treatment strategies for antiviral therapy in HIV epidemiology and judge the effectiveness of Covid-19 vaccines.

This project will focus on several open methodological problems within the target trial framework, and will involve developing new models and estimation approaches for emulating target trials using observational data. The methods will be illustrated using case studies in collaboration with medical researchers at the Ghent University Hospital.

Responsible:
Oliver Dukes,
Stijn Vansteelandt,
Department of Applied Mathematics, Computer Science and Statistics, Faculty of Science, Ghent University, Belgium

Job profile

• Master in statistics, biostatistics, mathematics, computer science or equivalent.
• A motivation for conducting methodological scientific research with a view to defending a PhD.
• Demonstrable quantitative skills and interest in applications of statistics in biomedical research.
• Experience with programming in a statistical package, eg R, Stata, SAS.
• Language skills: Excellent written and spoken English.
• Generic competences: Proactive, communicative, accurate, able to work independently and in team.

How to apply

Please send a CV and cover letter to oliver.dukes@ugent.be. In your cover letter, please briefly describe your interest and familiarity with causal inference.