Using historical data to predict patients' future outcomes and exciting times for predictive, federated modelling
Description
In this episode, we explore developments in predictive modelling and its growing impact in research and for clinical practice, with a focus on methodology, quality and lessons learned. Our guest in this episode is Peter Rijnbeek, Assoc Prof, Health Data Sciences at Erasmus MC, Rotterdam, and he is the Coordinator for EHDEN, in the leadership committee of OHDSI, and leading OHDSI Europe, as well as the OHDSI Predictive Modelling Working Group (with Jenna Reps, Janssen R&D). We start with briefly reflecting on the end of year three for EHDEN and the focus on evidence generation within the network of Data Partners. Peter covers initial basics of what is predictive modelling, the opportunities and challenges today, and demonstrative use cases across therapeutic areas, from Type 2 Diabetes, to Dementia, through to COVID-19. This final episode of season 1 finishes with the exciting prospects for expansion in this field, inclusive of a novel prediction model library being developed, but with the need to ensure transparency and reproducibility, as well a critical need for education across the field through to end users, clinicians.
United States
