Clinical prediction models are abundant in the medical literature, but their implementation is still sparse in clinical practice. Collaboration between data scientists, clinical researchers and clinicians at an early stage is crucial to be able to exploit the latest developments in data science and to tackle clinically relevant problems. Therefore, this session aims to disseminate the latest recommendations for development, evaluation and reporting of clinical prediction models to the data science community (e.g. TRIPOD-AI).

Program

Clinical prediction models: from development to external validation TRIPOD+AI: guidance for reporting clinical prediction models (seminar talks with Q&A, 60 mins)

Speakers

• Gary Collins, Professor, Director of Centre for Statistics in Medicine, University of Oxford, UK
• Paula Dhiman, Senior Researcher in Medical Statistics, University of Oxford, UK

Panel discussion P. Gary Collins, Paula Dhiman & Adam Hulman on How to build bridges between data science and clinical research? (30 mins)

Organizers

• Livie Yumeng Li, PhD Student, Department of Public Health, Aarhus University
• Anders Aasted Isaksen, Clinical Researcher, Steno Diabetes Center Aarhus
• Martin Rune Hassan Hansen, Clinical Researcher, Steno Diabetes Center Aarhus
• Adam Hulman, Associate Professor, Steno Diabetes Center Aarhus & Chairman of the DDSA Cross-Academy Collaboration Committee

Level

Introductory. Participation in the session neither requires deep prior knowledge in data science nor in clinical research, making the session easily accessible to a broad audience including pre-graduate students.