Association Between Cardioembolic Sources On Transesophageal Echocardiogram In Acute Ischemic Stroke And Atrial Fibrillation In Follow-Up Holter: A Retrospective Cohort Study In A Third Level Hospital In Buenos Aires, Argentina.
Rocio Marquez1, Alejandra Heriz1, Facundo Escandon1, Juan Pablo Rodriguez1, Rodrigo Sanjinez1, Otto Vega1, Guido Stupenengo1, Natalia Balian1, Jeremias Ayerbe1, Marcelo Rugiero Rodriguez1
1Hospital Italiano de Buenos Aires
Objective:
To describe the proportion of adults with acute ischemic stroke (AIS) in which cardioembolic sources were found on transesophageal transesophageal (TEE) and the association with atrial fibrillation (AF) in follow-up Holter.
Background:
The identification of AF is highly relevant as it increases five-fold the risk of stroke. Holter monitoring is often used although it is controversial in terms of cost-effectiveness, especially in developing countries. This could be due to low detection rate (3-8%) because of the inclusion of unselected patients. 
Design/Methods:
A retrospective cohort study conducted between October 1st, 2015 and September 30th, 2022 which included adults with AIS who underwent TEE and follow-up holter.  Major and minor cardioembolic sources were defined according to the Demographical, clinical and paraclinical (magnetic resonance imaging and TEE) data were collected from electronic records.
Results:

369 patients were included. 9.8% (36) had AF. Median age was 76 years (IQR 17) and 46.6% (172) were male. A cardioembolic source on TEE was found in 31.4% (116), [5.4% major, 27.9% minor].

AF was not associated neither with minor cardioembolic sources (p=0.12) nor with total cardioembolic sources (p=0.16) in a bivariable analysis. However, AF was significantly associated with major cardioembolic sources on TEE (p=0.03). This association was not found (p=0.07) when adjusted by age, gender and minor cardioembolic sources in a multivariable logistic regression analysis. Age and male gender were associated with AF, OR 1,04 (CI95%, 1,01-1,08, p=0,048) and OR 0,43 (CI95%, 0-18-0.94, p=0,04) respectively.


Conclusions:

In the population included 9.8% had AF. Major cardioembolic sources were statistically associated with AF in a bivariable analysis, but not when adjusted by age, gender and minor sources. No association with minor sources was found. 


10.1212/WNL.0000000000204125