From Real-World Data to Clinical Visualization: Developing a Data Model for Wound Care

Abstract

Introduction: The rapid expansion of digital health is changing healthcare, integrating clinical research and everyday practice. Real World Data from Electronic Health Records is the cornerstone for generating Real World Evidence that informs decision-making. Chronic wounds require continuous monitoring and intervention, and traditional wound assessment remains inefficient and subjective. The openEHR standard offers a solution for creating persistent, longitudinal health records. However, the true value of such data is unlocked only when it is effectively reused. Bridging the gap between robust clinical data capture and its effective secondary use remains a challenge, and one effective approach is the development of analytical packages and visualization dashboards that can provide actionable insights. This work aims to design and validate a dual-purpose data model for wound care that supports both primary data collection in clinical practice and secondary use for analytics and visualization. Methodology: This work followed a methodology of sequential, interdependent tasks, including clinical modelling, semantic mapping, and data visualization. The initial phase established the clinical requirements for a comprehensive wound dataset. Subsequently, the core information model was designed using the openEHR standard, with archetypes and templates authored in Archetype Designer and validated against the Clinical Knowledge Manager (CKM). The resulting model's technical conformance was validated using a data repository based on the Open-Source EHRbase platform. To ensure semantic interoperability, existing terminology bindings from validated archetypes were leveraged, complemented by a manual mapping of additional concepts to standard clinical vocabularies, including Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT), using the SNOMED CT Browser and the Observational Health Data Science and Informatics (OHDSI) Athena platform. Finally, the model's dual-purpose utility was validated by developing an interactive Apache Superset dashboard to translate the structured data into actionable clinical and management insights. Results: An archetype-based openEHR model for wound care was developed, comprising 53 data points across key clinical domains such as patient demographics, risk assessment scales (e.g., Braden), wound characteristics (e.g., dimensions, tissue type), and treatment protocols. To enable a holistic patient view for secondary use, the model's design anticipates interoperability with ancillary systems. This approach resulted in a consolidated dataset of 107 data points for the visualization dashboard, integrating contextual clinical variables from sources such as laboratory information systems (e.g., inflammatory markers) and pharmacy records. The hybrid mapping approach, combining pre-validated archetype bindings with targeted manual curation, ensured semantic fidelity. An interactive, two-panel dashboard ("Clinical" and "Management") was successfully implemented in Apache Superset, providing dynamic visualizations of key indicators designed to address clinical challenges such as tracking wound progression, identifying infection risks, and evaluating treatment effectiveness, thereby demonstrating the value of the collected data for quality monitoring. Conclusion: This work presents a conceptual and methodological framework for integrating primary and secondary data use in wound care through a dual-purpose information model. It demonstrates how leveraging openEHR for robust data persistence, combined with standardized terminologies and modern business intelligence tools, can enhance interoperability and unlock the value of real-world data. The proposed solution provides a scalable foundation to support data-driven decision-making, ultimately contributing to improved clinical outcomes and more efficient healthcare management.