Decades in the Making: The Evolution of Digital Health Research Infrastructure Through Synthetic Data, Common Data...

Area of Science:

• Digital Health Research
• Health Informatics
• Clinical Epidemiology

Background:

• Traditional medical research relies heavily on randomized controlled trials (RCTs), which are costly and limited in scope.
• There's a growing need for health research to evolve, utilizing real-world data (RWD) for large-scale, longitudinal insights.
• RWD, collected during routine care via digital health infrastructure, offers real-life, long-term outcome data.

Purpose of the Study:

• To describe the evolution of digital health research infrastructure.
• To address challenges in harnessing RWD for efficient, ethical, and compliant digital health research.
• To present novel methods enhancing RWD utility in health research.

Main Methods:

• Discusses the evolution of synthetic data generation, common data models, and federated learning.
• Emphasizes the importance of cross-sector collaboration in digital health research.
• Highlights the integration of RWD with traditional evidence from RCTs.

Main Results:

• Identifies key challenges in RWD utilization: data quality, integration, governance, compliance, analytics, and translation.
• Showcases novel methods like common data models, federated learning, and synthetic data to overcome RWD silos.
• Demonstrates a significant shift in clinical evidence generation over 25 years, incorporating RWD alongside RCTs.

Conclusions:

• Digital health research requires new data infrastructure, akin to drug trial support systems.
• Federated learning, common data models, and synthetic data are key to unlocking RWD potential.
• Lessons learned offer a model for other jurisdictions needing RWD infrastructure for digital health research.