Arrhythmias & electrophysiologyResearch article
Safety and Outcomes of Intracardiac Versus Transesophageal Echocardiography for Left Atrial Appendage Closure in the Very Elderly: Propensity Score Matched Real-World Outcomes From a Large US Network
- Propensity score matched real-world outcomes
- Propensity score matching (1:1)
- TriNetX US Collaborative Network
- Cox proportional hazards models
Authors' conclusion
In this large propensity-matched cohort of very elderly patients undergoing left atrial appendage occlusion, ICE and TEE demonstrated comparable rates of mortality, stroke, and device thrombosis at 90 days and 1 year. ICE was associated with a higher rate of device leak at 1 year, warranting careful procedural technique and follow-up surveillance. Prospective studies are needed to define optimal intraprocedural imaging strategies in this high-risk population.
Study context
- Population
- Patients aged ≥80 years with atrial fibrillation who underwent percutaneous left atrial appendage occlusion from 2015 to 2025patients per group after propensity score matching
- Participants
- 2,913 patients
- Comparison
- Intracardiac echocardiography (ICE)- versus transesophageal echocardiography (TEE)-guided left atrial appendage occlusion
- ICEn = 2,913
- TEEn = 2,913reference
Main finding
Mortality · 90 days
Hazard ratio
1.18
95% CI 0.81–1.73
As reported in the abstract
Further findings
| Outcome | hazard ratio (95% CI) | |
|---|---|---|
| Mortality · 1 year · similar between groups | 0.93 (0.76–1.13) | |
| 0.512Hazard ratio (log scale) |
Stroke, device thrombosis, pericardial effusion, pericardiocentesis, tamponade, device leak · 90 days
- No significant differences between ICE and TEE
Stroke and device thrombosis · 1 year
- Remained similar between groups
Baseline characteristics after matching
- mean age
- 83 years
- women
- 44%
- Baseline characteristics were well balanced after matching
Safety
| Outcome | hazard ratio (95% CI) | |
|---|---|---|
| Device leak · 1 year · higher incidence with ICE | 1.81 (1.11–2.97) | |
| 0.20.5125Hazard ratio (log scale) |
Original publication
Summarised from the published abstract; every number is checked against it