FHIR for Remote Patient Monitoring: Architecture and Data Structure

Details

Implementing Remote Patient Monitoring (RPM) requires reliably moving measurements from wearables and home medical devices into professional care workflows. The typical architecture involves a smartwatch or blood pressure monitor sending readings via Bluetooth to a smartphone or gateway, which then validates the payload, adds timestamps, and securely transmits it to a healthcare platform. FHIR's Observation resource is central for representing these measurements. Designed for clinical assertions and measurements, Observation can reference multiple entities like Patient, Device, Encounter, etc. The structure allows standardized coding of results (e.g., oxygen saturation) using LOINC, ensuring context regarding the specimen or measurement method. Beyond simple data logging, FHIR integrates with resources such as Condition, QuestionnaireResponse, and Provenance to provide clinical context. This enables an alert engine to evaluate multiple data sources—including trends, symptoms, and device quality—to support a clinician's dashboard. Follow-up actions are represented by resources like Task, CarePlan, and ServiceRequest, forming a comprehensive care process supported by FHIR workflows. Crucially, the text emphasizes that while FHIR provides structured, traceable, interoperable data to support timely human decisions, it does not diagnose the patient or replace clinical judgment.