Table 2

Baseline probabilities and diagnostic accuracy of conventional testing versus ICM for unexplained syncope

Parameter	Mean	Source
Mortality
Monthly probability for general population	Age and sex-specific	US Centers for Disease Control16
HR for cardiac vs no syncope	2.01	Soteriades et al 20022
HR for vasovagal vs no syncope	1.08	Soteriades et al 20022
Cardiac cause of syncope
Proportion of unexplained syncope patients with arrhythmia	60.4%	Solbiati et al, 201718; meta-analysis of 4381 patients
Proportion of arrhythmia patients with bradycardia*	68.7%
Proportion of arrhythmia patients with ventricular tachycardia and ventricular fibrillation*	10.2%
Proportion of arrhythmia patients with supraventricular tachycardia or atrial fibrillation*	21.1%
Syncope recurrence
Monthly probability of syncope recurrence	5.1%	EaSyAS15; monthly risk corresponding to event rate of 0.6 events/year
Injury due to syncope
Probability of major injury per syncope event	4.8%	Bartoletti et al, 200819
Probability of minor injury per syncope event	24.7%	Bartoletti et al, 200819
Probability of ECG capture
Conventional testing	18.9%	Farwell et al, 200643
Reveal LINQ ICM	Month 0=70% Month 1=82% Month 2=78% Month 3=87% Month 4=90% Month 5 onward=97%	Musat et al, 201744
Battery life
Reveal LINQ ICM	3.0 years
Adverse events
Risk of AE requiring ICM explant	First cycle=0.00734 Subsequent cycles=0	Pooled data from LINQ ICM usability and registry studies
Probability of ICM removal (explant)
Probability of removal on diagnosis of arrhythmic syncope	80%	Assumption
Probability of removal on diagnosis of non-arrhythmic syncope	100%	Assumption

*See online supplemental files for a breakdown of how these probabilities were calculated.
AE, adverse event; EaSyAS, Eastbourne Syncope Assessment Study; ECG, electrocardiogram; HR, hazard ratio; ICM, insertable cardiac monitor; US, United States.