Real world data on left atrial appendage closure after severe gastrointestinal bleeding in anticoagulated patients with atrial fibrillation: frequency and associated factors.

Abstract Background There are no "real world" data on the actual impact of percutaneous left atrial appendage closure (LAAC) programs in the management of embolic event prevention after a severe gastrointestinal bleeding (GIB) in anticoagulated patients with atrial fibrillation (AF). Purpose Our main objective was to investigate the frequency of LAAC in anticoagulated AF patients surviving a severe GIB and to describe the differential features of these patients versus those medically managed in Andalusia (South of Spain). Methods All consecutive patients discharged alive from January 1st, 2021 to December 31st, 2022, after a severe GIB or intracranial haemorrhage, who were previously anticoagulated for AF, in all public Andalusian hospitals with LAAC program, and who could be eligible either for LAAC or direct anticoagulant treatment were included in a retrospective, multicentre study. This analysis focuses in patients with GIB. The frequency of a percutaneous LAAC strategy in the follow-up was registered. A comparative analysis of baseline features between those patients treated with LAAC and those medically managed was performed. Results Among a reference population of 7119044 inhabitants in 15 hospitals, 1403 patients were included in the study, and 1041 had a GIB and conform this study sample. Median age was 82 years p25-75, 76-87 years with 48.7% male patients. A LAAC procedure was performed in 70 patients (6.7%) at a median time of 5 2-10 months after discharge. After adjusting for confounding variables in a logistic regression model, LAAC closure was independently associated with younger age, previous bleeding events, angiodysplasia, lower diastolic blood pressure, direct anticoagulant or aspirin use at admission and discharge from a tertiary referral centre (Table). Variables with univariate association with LAAC (p0.10) but excluded in backward stepwise regression after showing no significant association with the dependent variable were diabetes mellitus (rate of LAAC 8.9% vs 5.3% in non-diabetics, p=0.02), hypercholesterolemia (8.0% vs 5.3%, p=0.08), ischaemic heart disease (10.4% vs 6.0%, p=0.03), at least moderate valvular disease (9.0% vs 5.9%, p=0.08), dementia (2.9% vs 7.5%, p=0.03), liver disease (10.3% vs 6.3%, p=0.099), anemia (7.2% vs 4.1%, p=0.14), and vitamin K antagonist use at admission (3.8% vs 7.6%, p=0.04). Conclusions In anticoagulated patients with AF who survived a severe GIB in all public hospitals with LAAC program in a large European region in 2021-2022, and who were also eligible for anticoagulation with DOAC, only 6.7% were managed with a LAAC strategy. Several baseline features were independently associated with higher or lower frequency of this approach.Table