Abstract
INTRODUCTION: Despite strong evidence supporting guideline-directed medical therapy (GDMT) for heart failure with reduced ejection fraction (HFrEF), a significant gap persists in the consistent application of these therapies. This shortfall has prompted organizations like the American College of Cardiology to recommend leveraging electronic health records (EHR) to optimize GDMT. This paper discusses the development of SmartHF, a clinical decision support system designed to enhance therapy adherence by effectively linking Fast Healthcare Interoperability Resources (FHIR)-based medication data with clinical algorithms tailored for the management of HFrEF. METHODS: The SmartHF system integrates FHIR-based medication data with clinical algorithms through a multi-step approach. Central to this process is data from CodeRx, a platform that utilizes streamlined data pipelines to map medication products to their ingredients using RxNorm. The methodology addresses the challenge of interpreting both structured and unstructured medication instructions, ensuring a precise linkage of product identifiers to algorithm-relevant ingredients and their corresponding strengths. Specific attention is given to the data granularity needed for distinguishing precise ingredients within complex formulations, such as sacubitril/valsartan and metoprolol salt form variants. RESULTS: The deployment of SmartHF involved rigorous testing using actual and synthetic patient datasets to validate its functionality. Results demonstrated the system's ability to process FHIR MedicationRequest data resources accurately, convert free-text dosing instructions into usable formats, and handle edge cases, including non-standard products and missing dose information. CONCLUSIONS: This article describes the potential of FHIR-based medication data integration for enhancing clinical decision support tools and improving care quality. It highlights the challenges and solutions for this integration.