Risk of stroke following discontinuation of oral anticoagulant treatment for atrial fibrillation: a cohort study

Date of Approval
Application Number
Lay Summary

Atrial fibrillation is an irregular and often rapid heart rate. Although atrial fibrillation itself usually isn't life-threatening, it is a serious medical condition that sometimes requires emergency treatment. Atrial fibrillation can lead to blood clots. If a blood clot forms, it could dislodge from the heart and travel to the brain, block blood flow to the brain and cause a stroke. The risk of developing a stroke is about 5 times higher in patients with atrial fibrillation than in patients without atrial fibrillation.
Life-long use of blood thinners, like warfarin, is recommended to reduce the formation of blood clots and to prevent strokes. However, it is known that approximately 50% of patients with atrial fibrillation stop its use within one year.
People who discontinue their blood thinners carry an increased risk of stroke. However, the actual level of risk after discontinuation of blood thinners is unclear.
To better understand the consequences of stopping use of blood thinners we will undertake a study in patients with newly diagnosed atrial fibrillation who started treatment with blood thinners. We will compare the risk of stroke among those who then discontinued use of blood thinners with those who continued use of blood thinners.

Technical Summary

The objective is to estimate the risk of stroke in patients with atrial fibrillation (AF) who are initially treated with vitamin K antagonists (VKAs), comparing patients discontinuing VKAs with those currently using VKAs.
This will be a population-based cohort study in patients treated with VKAs for AF, analysed using a nested case-control approach. The cohort of patients with an incident AF treated with VKAs will be identified from the Clinical Practice Research Datalink (CPRD) and Hospital Episode Statistics (HES). For each patient with a stroke during the observational period (cases), controls without a stroke will be selected from the VKA-treated AF cohort. Strokes will be identified from CPRD, HES and Office of National Statistics (ONS) mortality data.
Data analysis
The relative risk of VKA discontinuation compared to VKA continuation with regard to stroke will be estimated from crude and adjusted odds ratios (ORs) derived from conditional logistic regression for matched case–control data using current VKA use as the reference category.

Health Outcomes to be Measured

Incident strokes occurring during the study period after AF diagnosis


Carlos Martinez - Chief Investigator - Institute for Epidemiology, Statistics and Informatics GmbH (Pharma Epi)
Ben Freedman - Collaborator - ANZAC Research Institute
Christopher Wallenhorst - Collaborator - Institute for Epidemiology, Statistics and Informatics GmbH (Pharma Epi)


HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation