GENERATOR HEART FAILURE DataMart: An integrated framework for heart failure research

Area of Science:

• Cardiology
• Health Informatics
• Data Science

Background:

• Heart failure (HF) is a complex syndrome with diverse causes and clinical trajectories.
• Current HF research models are limited by biases from highly selected clinical trial populations.
• Real-world data offers significant potential to advance clinical care pathways for heart failure patients.

Purpose of the Study:

• To develop a framework, the GENERATOR HF DataMart, for managing clinical and research processes in heart failure.
• To create a robust system for leveraging real-world data to generate evidence and improve patient care.

Main Methods:

• Integration of retrospective patient information into a multimodal, longitudinal data repository.
• Implementation of disease trajectory toolsets and predictive models for heart failure evolution.
• Adherence to CODE-EHR framework standards for methodological scaffolding.

Main Results:

• The HF DataMart enables dynamic filtering of patient subsets based on demographics, biomarkers, and clinical events.
• Demonstrated utility in selecting retrospective cohorts, exploring data associations, and identifying study participants.
• Successful development of multi-parametric predictive models for re-hospitalization and clustering patients by ejection fraction variation.

Conclusions:

• The GENERATOR HF DataMart effectively utilizes institutional real-world data to generate evidence for heart failure management.
• The platform provides an infrastructure for integrated patient support, continuous monitoring, and remote care.
• This system assists patients, caregivers, and healthcare professionals in managing heart failure.