Arrhythmias & electrophysiology
Diagnostic accuracy for detecting atrial fibrillation using a novel machine learning algorithm in a blood pressure monitor
Authors' conclusion
These data support both home and professional use of these novel Omron BPMs for the detection of AF.
Study context
- Population
- Subjects with and without atrial fibrillation (AF) at 5 US sites267 AF, 292 non-AF
- Participants
- 559 subjects
- Comparison
- Omron blood pressure monitors (BPMs) with machine-learning AF screening vs Microlife BPM, against simultaneous 12-lead electrocardiogram (ECG)
- Outcome
- Sensitivity, specificity and accuracy for AF diagnosis vs 12-lead ECG
Methodological detailsmulticenter, prospective, 5 US sites, cardiologist blinded to BPM results
- Multicenter
- Prospective
- 5 US sites
- Cardiologist blinded to BPM results
Performance
As reported in the abstract
AF detection vs 12-lead ECG
| Test or model | Sensitivity | Specificity | Accuracy |
|---|---|---|---|
| HEM-7371T1-AZ | 95.1% (95% CI 91.8%-97.4%)Sensitivity on a scale from 0 to 100 percent | 98.6% (95% CI 96.6%-99.7%)Specificity on a scale from 0 to 100 percent | 97.0% (95% CI 95.2%-98.2%) |
| Microlife BPM | 78.5% (95% CI 73.1%-83.3%)Sensitivity on a scale from 0 to 100 percent | 97.6% (95% CI 95.1%-99.0%)Specificity on a scale from 0 to 100 percent | 88.4% (95% CI 85.5%-91.0%) |
Sensitivity
HEM-7371T1-AZ95.1% (95% CI 91.8%-97.4%)
Microlife BPM78.5% (95% CI 73.1%-83.3%)
Sensitivity on a scale from 0 to 100 percentSpecificity
HEM-7371T1-AZ98.6% (95% CI 96.6%-99.7%)
Microlife BPM97.6% (95% CI 95.1%-99.0%)
Specificity on a scale from 0 to 100 percentAccuracy
HEM-7371T1-AZ97.0% (95% CI 95.2%-98.2%)
Microlife BPM88.4% (95% CI 85.5%-91.0%)
Original publication
Diagnostic accuracy for detecting atrial fibrillation using a novel machine learning algorithm in a blood pressure monitor. Heart Rhythm. Published 30 April 2024. doi:10.
Summarised from the published abstract; every number is checked against it