Contemporary Causal Methods Special Interest Group Guest Speaker

Establishing causality using electronic health record (EHR) data is challenging and requires careful analysis and appropriate study design. Analytic methods are always evolving, so researchers need to keep up-to-date with best practice. At the Bennett Institute, we recently set up a Special Interest Group to discuss modern ways of analysing observational data, especially approaches that help understand causal associations.

In October we welcomed Associate Professor Clémence Leyrat who spoke about immortal time bias. Next month, we will welcome Associate Professor Nan van Geloven from Leiden University Medical Centre.

Nan van Geloven is Associate Professor in Biostatistics at the Leiden University Medical Center (the Netherlands), whose research focuses on the intersection of causal inference and clinical prediction models. Her methodological work addresses the complexities of integrating (time-varying) treatment effects into predictive frameworks. Dr van Geloven leads the ‘Causal Inference for AI’ network, is a member of the international STRATOS initiative, and serves as a work package leader within the Safe Causal Inference consortium. As the recipient of a 2025 ZonMw Vidi grant, she currently investigates the clinical impact of implementing algorithmic decision support in healthcare. Her work is applied across diverse clinical domains, including transplantation, intensive care, and transfusion medicine.

If you are in Oxford, please join us for her talk entitled: Accounting for treatment in clinical prediction models - the good, the bad and the ugly.

When? Thursday, 23 April, 12:00-13:30 (presentation and Q&A, followed by light lunch) Where? Room 4, Gibson Building, Radcliffe Observatory Quarter

Register for the event here.

For more information, or if you would like to be added to our mailing list to be contacted about future Contemporary Causal Methods SIG events, please email andrea.schaffer@phc.ox.ac.uk or william.hulme@phc.ox.ac.uk.