Technology

“Enabling ubiquitous data pipelines and exploiting them for intelligent health requires:

• Adopting standards — Fast Healthcare Interoperability Resources (FHIR) are emerging as the de facto language for health data exchange globally. There has been great progress made in the development and adoption of FHIR, and it is often specifically articulated in national data strategies that call for a participation in common data exchange platforms. FHIR skills and resources need to be continuously mastered, exploited and governed (see Healthcare Provider Interoperability Benchmarks and Trends, 3Q23).
• Advancing interoperability — A new class of interoperability hubs and health data management platforms is emerging. These advances signal a new generation of capabilities that will address historic constraints associated with health data integration and exchange and will force a reengineering of data pipelines (see Quick Answer: Emergence of Health Data Management Platforms Signal a New Generation of Capabilities).
• Building data engineering competency — Healthcare and life science organizations must become proficient in data engineering skills in order to support next generation best practices in data management and to orchestrate data in a manner that will allow nontechnical users to easily locate and interrogate the data they need (see 5 Ways to Enhance Your Data Engineering Practices).
• Harnessing the IoT for healthcare — Ingesting streaming data from ambient listening devices, computer vision, beacon technology, geolocation, medical devices and consumer wearables is needed to fully materialize goals of health at home, patient-enabled health and remote monitoring (see Healthcare Delivery Organization IoT Scale Demands a Platform Approach).
• Enabling privacy enhancing computation — Enabling organizations to share the insights contained within data for research, public health and machine learning, without sharing the regulated data itself. These technologies include synthetic data and federated learning (see Three Critical Use Cases for Privacy-Enhancing Computation Techniques).”