LibraryFamily 14 · Population Health & Epidemiology
Heart-failure decompensation early warning from wearables
Under reviewPilotDeveloping evidence
Undergoing in-depth clinical, technical and governance review.
Combines wearable vital-sign trends and patient-reported symptoms to flag likely decompensation 3–7 days before clinical presentation, prompting community-nurse contact.
Plain-language summary
AIChoose an audience and generate a tailored summary on demand. The AI uses only what is on this page.
Clinical context
- Clinical problem
- Monitor a patient's status
- Point of care
- Follow-up
- Nature of AI output
- An alert
- Clinical specialty
- Cardiology
- Care setting
- Home / remote
- Patient population
- Adults with chronic heart failure enrolled in a remote-monitoring programme.
- Intended use
- Combines wearable vital-sign trends and patient-reported symptoms to flag likely decompensation 3–7 days before clinical presentation, prompting community-nurse contact.
Technology
- AI technique
- Classical machine learning, Statistical model
- Input data
- Vital signs, Patient-reported data, Waveforms (ECG, EEG…)
- Output type
- Alert
- Autonomy level
- Informs a human (advisory)
- Model provenance
- Built in-house
- Model version
- hf-1.0
- Built on a general-purpose model
- No
Deployment
- Status
- Pilot
- Country
- Netherlands
- Deployment date
- 15 February 2025
- Sites
- 1
Regulatory & governance
- EU AI Act risk tier
- High-risk
- High-risk basis
- Annex III use case
- Medical device
- No
- EU MDR class
- Not a device
- CE marking
- Not required
- FDA status
- Not applicable
- ISO 14971 risk class
- Medium
- GDPR processing basis
- Public interest
- GDPR DPIA
- Completed
- Data identifiability
- Pseudonymised
- Explainability method
- Intrinsic
- Human oversight model
- A community nurse contacts the patient on every amber alert; the AI never escalates to hospital without a human.
NICE evidence standards
- ESF tier
- Tier B — inform / monitor
- Evidence category
- Category 2
Performance summary
- Headline metric
- Sensitivity
- Value
- 0.79
- Subgroup performance assessed
- Yes
- Known bias signals
- Higher false-positive rate for atrial-fibrillation patients; under recalibration.
Similar deployments
AILooking for close matches in the Library…
Safety signals
Post-deployment concerns flagged by clinicians, patients or monitoring (proposal §3.7). High or critical signals re-enter a validated use case into maturation automatically.
Patient reports too many amber alerts during exercise; threshold under review.
Raised 17 May 2026, 21:01
LowOpen
Evidence records
Studies and evaluations attached to this use case.
Prospective observational
Sensitivity: 0.74
Population: Prospective pilot, 250 patients
Subgroup: AF subgroup
Retrospective validation
Sensitivity: 0.79
Population: Remote-monitoring HF cohort, 700 patients
Contributors
- Deploying organisation
- [demo] Erasmus MC · Hospital / health system · Netherlands
- AI vendor
- —
- Product name
- —