Distributed data networks: a blueprint for Big Data sharing and healthcare analytics.

Area of Science:

• Health Informatics
• Data Science
• Network Infrastructure

Background:

• Distributed data networks are crucial for large-scale health research.
• The U.S. Food and Drug Administration's Sentinel Initiative serves as a model for healthcare-related distributed data networks.
• Developing robust analytic infrastructure is key to leveraging these networks effectively.

Purpose of the Study:

• To define the attributes of successful distributed data networks.
• To outline the necessary data and analytic infrastructure for building and maintaining such networks.
• To explore the application of machine learning in population health analytics using multisite administrative data.

Main Methods:

• Defining key attributes of distributed data networks.
• Examining the Sentinel Initiative as a case study for implementation.
• Discussing analytic infrastructure development based on six core pillars: consistency, reusability, flexibility, scalability, transparency, and reproducibility.
• Introducing a machine learning use case for population health analytics.

Main Results:

• Successful implementation of a large-scale, multisite, healthcare-related distributed data network (Sentinel Initiative).
• Identification of critical components for data and analytic infrastructure.
• Demonstration of how the six pillars enhance analytic infrastructure.
• Exploration of machine learning's potential in multisite administrative data analysis.

Conclusions:

• A well-defined infrastructure is essential for effective distributed data networks.
• The six pillars provide a framework for building sustainable and scalable analytic capabilities.
• Machine learning offers advanced opportunities for population health analytics using distributed data.