An Original Study on Performance-Optimized EMR-to-HL7 FHIR Conversion Using a Lightweight Library

Background: Heterogeneous electronic medical record (EMR) systems and institution-specific data structures continue to limit interoperability and large-scale utilization of healthcare data. Although Health Level Seven (HL7) Fast Healthcare Interoperability Resources (FHIR) has been adopted as an international standard, existing conversion approaches often require extensive preprocessing, high implementation costs, or deep system-specific expertise, restricting their applicability, particularly in small and medium-sized hospitals. To address these constraints, we propose a lightweight EMR-to-HL7 FHIR conversion library optimized for small- and medium-sized healthcare providers that operate with limited system resources. Methods: The library adopts a modular architecture comprising data preprocessing, reference management, structural transformation using transform maps, terminology translation, and validation modules. The proposed approach was implemented using the HL7 Application Programming Interface (HAPI) FHIR and evaluated with anonymized EMR data extracted from multiple hospitals in South Korea, with performance and validation results compared against a conventional HAPI FHIR client-based conversion method. Results: This study proposes a standardized FHIR-based medical data conversion library that enables the efficient transformation of diverse EMR data structures into interoperable FHIR. The proposed library achieved approximately 30% lower single-request conversion latency compared to a conventional HAPI FHIR client-based conversion pipeline under identical hardware and runtime conditions. Conclusions: The proposed conversion method provides a lightweight and adaptable solution for EMR-to-FHIR transformation, improving interoperability with reduced implementation effort and supporting scalable medical data exchange across diverse healthcare environments.