PURPOSE:

The objectives of this session are to share learnings and insights from a data-driven hackathon.

DESCRIPTION:

Provision of real-world-evidence (RWE) is of great interest for all stakeholders within the health care sector. Where randomized controlled trials (RCTs) provide the unique ability to evaluate efficacy, the use of real-world-data (RWD), i.e. claims data, enable analyses of real-world clinical practice. However, neither data source is in itself without limitations with regard to internal and external validity, time horizon and collected parameters. We experimented with an innovative approach – a “Datathon” (data-driven Hackathon) - to bridge this gap. In our Datathon we had the ambition, by combining data and knowledge from both perspectives (RCT/real-world) with machine learning in an unconventional and intensive working event over 2,5 days to tackle different medical challenges within the healthcare sector. These challenges in the area of asthma and chronic lymphocytic leukemia were assigned to competing teams, resulting in innovative solutions and comprehensive insights for participants and stakeholders alike.

This breakout session starts with a background on the challenge of data driven healthcare projects and an outline of the Datathon approach. Insights from the conceptualizing phase, challenges and solutions from the perspective of data protection, data linkage, data science and general collaboration set-up are presented (20 min, Kümpel/Unger). Next, one of the winning teams of the Datathon will present its results for the prediction and identification of asthma patients with high risk of developing OCS related adverse events (20 min, Alig/Grotenrath). After a summary of learnings from the event (10 min), a discussion with the audience concludes the session (10 min).

This session may benefit any healthcare stakeholder with an interest in exploring an innovative, multi-disciplinary approach for RCT and RWD linkage and machine learning approaches in order to solve medical research questions and develop innovative care models.