Remote Patient Monitoring

Part of the worlds-biggest-software-project initiative.

An open-source, AI-native platform that aggregates wearable and medical-device data into actionable clinical alerts and care team workflows -- replacing fragmented proprietary portals with a self-hostable, standards-based alternative.

Remote Patient Monitoring (RPM) is a platform for securely ingesting data from connected medical devices (blood pressure cuffs, pulse oximeters, glucose monitors, CGMs, ECG patches, smartwatches), aggregating it into a longitudinal patient record, and driving configurable clinical alerting, care team task management, and billing automation. It is designed for health systems, physician practices, and home health agencies that need to monitor chronic and post-acute patients at scale without locking into a single vendor's ecosystem.

Why Remote Patient Monitoring?

No credible open-source RPM platform exists. Every major player -- HealthArc, Prevounce, ThoroughCare, Biofourmis, Health Recovery Solutions, Validic, HealthSnap -- is proprietary SaaS with per-patient subscription pricing and no self-hosting option. Health systems have no way to own their monitoring infrastructure.
Enterprise-only AI analytics shut out smaller providers. Biofourmis offers the most advanced predictive analytics (FDA-cleared Biovitals engine), but its enterprise-only model and high cost of entry make it inaccessible to small practices and rural health providers where RPM could have the greatest impact.
Alert fatigue undermines monitoring programmes. Most platforms rely on static, population-level thresholds that generate excessive notifications. Nurses learn to ignore them, defeating the purpose of continuous monitoring. AI-assisted, per-patient dynamic thresholds are needed but only available in the most expensive tiers.
EHR integration remains painful. Health systems run diverse EHR environments (Epic, Cerner, Meditech) and FHIR R4 implementation quality varies widely. Current platforms offer inconsistent integration depth, with some still relying on HL7 v2 batch feeds rather than real-time FHIR or SMART on FHIR embedding.
Billing complexity erodes programme viability. Medicare RPM CPT codes (99453-99458, plus 2026 additions 99445, 99470) carry specific time-tracking and documentation requirements. Automating this without vendor lock-in or patented workflows (e.g. HealthSnap's patented eligibility tool) is an unmet need.

Key Features

Device Ingestion and Data Normalisation

Support for top device types: blood pressure cuffs, pulse oximeters, weight scales, glucose monitors, and activity trackers
CGM integration (Dexcom, FreeStyle Libre, Eversense) for diabetes programmes
Bluetooth and cellular device pairing via patient mobile app
Data normalisation into a single standard across device manufacturers
Motion artefact filtering and data quality confidence scoring

Clinical Alerting and AI-Assisted Triage

Configurable threshold alerts per patient or programme
AI-driven dynamic thresholds that learn per-patient baselines over time
Patient risk stratification and proactive escalation prioritisation
Anomaly detection distinguishing genuine clinical events from sensor noise
Deterioration forecasting using multi-signal ML models (24-48 hour lookahead)

Care Team Workflows

Care team dashboard with patient list, trend charts, and alert queue
Task assignment, escalation paths, and alert acknowledgement tracking
Care plan templates for chronic conditions
Asynchronous messaging and telehealth visit scheduling integration
Population-level analytics for programme performance

EHR Integration

FHIR R4 patient and observation resource export
SMART on FHIR launch capability for EHR-embedded experience
CDS Hooks integration for point-of-care clinical decision support
HL7 v2 support for legacy EHR environments
Bidirectional data sync with Epic, Cerner, and other major EHRs

Billing and Reimbursement

CPT time-tracking and billing report generation (99453, 99454, 99457, 99458)
Support for 2026 CMS codes (99445, 99470)
Billing eligibility pre-screening per patient
Multi-programme support: RPM, CCM, RTM, PCM, TCM
Generative AI narrative summaries for EHR progress note creation

Patient Engagement

Patient-facing mobile app (iOS and Android) with biometric readings display
Condition-specific education content
Automated reminders and engagement tools
Conversational AI check-ins and symptom triage (backlog)

AI-Native Advantage

Unlike incumbents that bolt on analytics as an enterprise upsell, this platform treats AI as a core layer available to all deployments. Per-patient dynamic alert thresholds replace static population-level rules, directly addressing the alarm fatigue problem that plagues RPM programmes. Generative AI produces structured EHR progress notes from raw monitoring data, cutting clinician documentation time. Multi-signal deterioration forecasting models can flag adverse events 24-48 hours before they become emergencies -- a capability currently locked behind Biofourmis' proprietary FDA-cleared engine and enterprise contracts.

Tech Stack & Deployment

The platform targets self-hosted, cloud, and hybrid deployment modes to serve both large health systems and small practices. All EHR integration is built on open standards: FHIR R4 for data exchange, SMART on FHIR for embedded launch, and CDS Hooks for point-of-care decision support (all maintained by HL7 International). Device connectivity uses open Bluetooth SIG Health Device Profiles where available, with adapter layers for vendor-specific APIs. The architecture assumes HIPAA-compliant infrastructure with AES-256 encryption at rest and TLS in transit, and is designed to support SOC 2 Type II certification requirements.

Market Context

The RPM market is experiencing accelerated growth in 2026, driven by the ageing US population, rising chronic disease prevalence, and expanded Medicare reimbursement for remote monitoring services. Incumbent pricing follows a per-patient per-month subscription model, creating significant cost at scale for health systems managing thousands of patients. Primary buyers are health system CIOs, physician practice administrators, and home health agency operators -- organisations that increasingly seek to own their data infrastructure rather than rent it from proprietary vendors.

Project Status

This project is in the research and specification phase. Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. See discussion for context.