# Oral anticoagulation in patients with gastrointestinal bleeding and new‐onset atrial fibrillation: A population‐based registry‐linkage study

## TL;DR

This study finds that patients with new-onset atrial fibrillation and a history of gastrointestinal bleeding are less likely to start oral anticoagulation therapy, even as its use increases overall.

## Contribution

The study provides population-based evidence on how gastrointestinal bleeding affects oral anticoagulation initiation in atrial fibrillation patients.

## Key findings

- Patients with gastrointestinal bleeding had a lower probability of initiating oral anticoagulation therapy.
- Obscure gastrointestinal bleeding origin or concurrent bleeding further reduced oral anticoagulation initiation likelihood.
- Oral anticoagulation use increased from 2010 to 2018 but remained lower in patients with gastrointestinal bleeding.

## Abstract

Limited data exist on the prevalence of gastrointestinal bleeding (GIB) in patients with new‐onset atrial fibrillation (AF) and the impact of GIB on the initiation of oral anticoagulation (OAC) therapy.

A population‐based registry‐linkage study included all patients diagnosed with new‐onset AF in Finland during 2010–2018 who had available laboratory data and a definite indication for OAC therapy. The primary outcome was OAC initiation within 90 days following AF diagnosis. Factors associated with OAC initiation were assessed using modified Poisson regression.

Among 117 997 patients with new‐onset AF, 6628 (5.6%) had GIB, of which 5336 occurred more than 30 days prior to AF diagnosis, and 1292 were temporally (±30 days) associated with new‐onset AF (GIBTAF). Patients with GIB compared to those without GIB were older (mean age 78.3 vs. 75.3 years), more frequently men (48.5% vs. 41.9%), and had more comorbidities. The occurrence of GIB was associated with a lower probability of initiating OAC (RR 0.84, 95% CI 0.81–0.86). Among patients with GIB, an obscure origin of GIB (RR 0.93, 95% CI 0.88–0.99) or GIBTAF reduced the likelihood of OAC initiation (RR 0.72, 95% CI 0.66–0.79). The initiation of OAC did not depend on the known GIB bleeding site (lower vs. upper). Overall, the initiation of OAC therapy increased from 2010 to 2018 but remained consistently lower in patients with GIB.

Prior and concurrent GIB is common among patients with new‐onset AF, and despite the overall increasing use of OACs, they remain less utilized in patients with GIB.