A Browser-Native, Offline-Capable Clinical Decision Support Architecture with Native FHIR R4 Interoperability: System Design and Technology Validation

Description This work presents the design, implementation, and technical verification of a browser-native clinical decision support system (CDSS) architecture capable of operating entirely within a standard web browser while generating standards-compliant HL7 Fast Healthcare Interoperability Resources (FHIR) Release 4. Unlike conventional CDSS platforms that depend on enterprise electronic health records, application servers, or persistent network connectivity, the proposed architecture provides self-contained clinical decision support with native interoperability, enabling deployment in digitally limited healthcare environments. The architecture is implemented as a single HTML5 application using vanilla JavaScript without external frameworks, installation procedures, or build dependencies. Clinical knowledge is represented as configurable JSON-based rule objects, allowing modification or extension of clinical logic independently of the application core. Offline functionality is achieved through dynamic Service Worker generation, while interoperability is supported through native generation of HL7 FHIR Release 4 Bundles comprising eight resource types: Patient, Encounter, Observation, Condition, MedicationRequest, ServiceRequest, CarePlan, and QuestionnaireResponse. A twelve-rule inpatient diabetes module was developed solely to demonstrate the architecture and illustrate its disease-independent design. Technical verification evaluated four fundamental aspects of the platform: conformance of exported HL7 FHIR Release 4 resources to the published specification; offline operation across multiple browser engines; integrity and isolation of the rule-execution mechanism; and end-to-end functional verification of all application modules and diabetes decision rules using standardized synthetic patient scenarios. This publication reports an engineering proof of concept intended to demonstrate architectural feasibility rather than clinical effectiveness. The platform was evaluated without human participants, patient data, or clinical interventions, and no claims regarding diagnostic accuracy or patient outcomes are made. Clinical deployment requires independent validation of disease-specific knowledge and prospective evaluation within real-world healthcare settings. Keywords: Clinical Decision Support System; Software Architecture; HL7 FHIR Release 4; Browser-Native Computing; Rule-Based Reasoning; Offline Health Technology; Interoperability; Technical Validation.