Arrhythmias & electrophysiologyCohort study

Accuracy of Implantable Loop Recorders

JACC Clin ElectrophysiolPublished

  • multicenter, multivendor database
  • random sample selected based on a prospectively determined power analysis
  • alerts independently adjudicated by trained physicians
  • 25 centers

Authors' conclusion

ILRs demonstrate variable accuracy in arrhythmia detection, with a substantial burden of false-positive alerts across all vendors despite artificial intelligence-based and other algorithmic enhancements. Further improvement of alert algorithms to reduce clinic burdens is needed.

Study context

Population
Patients with implantable loop recorders (medtronic (MDT) LINQ II, BSX LUX-Dx, Biotronik (BIO) Biomonitor III, Abbott (ABT) Confirm Rx), January 2022 to March 2023random sample (∼283 per device) from a database of 437,351 electrocardiogram-verified episodes from 6,756 patients
Participants
1,140 patients
Exposure
arrhythmia detection accuracy among currently used implantable loop recorders (ILRs)

Main finding

As reported in the abstract

Total alerts

total alerts

25,826

alerts per patient

median of 5 alerts (IQR: 2-14)

Observational data: associations, not effects.

Further findings

Atrial tachycardia/AF alerts (share of alerts)

PPV

OutcomeMDTBSXBIOABT
PPV after adjudication0.560.730.230.51
Bradycardia detection PPV · MDT primarily due to undersensing0.360.96Not reported in the abstractNot reported in the abstract
Pause detection PPVNot reported in the abstract0.70Not reported in the abstract0.01
Tachycardia alert PPV0.290.610.620.27
  • Pause detection PPV: ABT 79.8% due to undersensing

Causes of false-positive atrial tachycardia/atrial fibrillation (AF) alerts with BIO devices

ectopy
75.8%
combined ectopy with oversensing/noise
5.3%
isolated oversensing
5.2%
sinus arrhythmias
3.9%
indeterminate reason
3.7%
noise
3.6%
bradycardia
2.5%

Original publication

Accuracy of Implantable Loop Recorders. JACC Clin Electrophysiol. Published 13 February 2026. doi:10.1016/j.jacep.2025.12.039

Summarised from the published abstract; every number is checked against it