Examining the change in risk of stroke and bleeding over time in patients with atrial fibrillation: A population-based cohort study using the Clinical Practice Research Datalink

Date of Approval
Application Number
Lay Summary
Atrial fibrillation is a common irregular heart rhythm (arrhythmia) and increases a person’s risk of stroke five-fold. The Atrial Fibrillation Better Care (ABC) pathway has been developed to help make a clear plan of care for patients and clinicians. It follows three steps: A - Avoid strokes; B - Better symptom management; C - Cardiovascular risk management. The ‘A’ part of the pathway is anticoagulation medication, therefore, the risk of stroke must be balanced against the risk of bleeding, but a person’s risk can change over time.

We will use a large anonymised database of patients in England which links electronic health records (e.g. GP visits, hospital visits) to answer the following five questions: (1) How does treatment adherent to the ABC pathway affect outcomes in atrial fibrillation patients? (2) How much does the risk of stroke and bleeding change over time in atrial fibrillation patients? (3) What is the best way to predict a person’s change in risk of stroke or bleeding? (4) How can the change in risk of stroke or bleeding over time best be used to help reduce someone’s overall risk? and (5) How frequently does risk need to be reassessed?

Clinicians and patients require clear information to assess the risks and benefits of medications to reduce stroke risk. We hope that the results of this study will indicate how changing risk can be best incorporated into decisions about anticoagulation therapy. The results will help to improve future care and decision making for atrial fibrillation patients
Technical Summary
Atrial fibrillation (AF) is the most common arrhythmia and increases a person’s risk of stroke five-fold. Oral anticoagulants (OACs) are recommended for the majority of individuals with AF to reduce their risk of stroke; however, anticoagulation also increases the risk of major bleeding. The CHA2DS2-VASc and HAS-BLED risk scores are advocated to assess risk for stroke and bleeding, respectively, in clinical practice. Both stroke and bleeding risk have the potential to change over time as patients age and develop new comorbidities. The change in CHA2DS2-VASc and HAS-BLED scores have been shown to perform better than baseline or follow-up scores in predicting stroke and major bleeding.
The Atrial Fibrillation Better Care (ABC) pathway is an integrated management pathway for patients with AF with three components: A - Avoid stroke with anticoagulation; B - Better symptom management and C - Cardiovascular risk management. Previous studies have shown that adherence to the ABC pathway at baseline is associated with lower risk of mortality, stroke, myocardial infarction, hospitalization and composite outcomes. We will treat adherence status as a time-dependent variable.
This project will use the Clinical Practice Research Datalink (CPRD), which is a representative sample of the primary care population in England. We will address the following four aims: 1. To compare the performance and flexibility of risk prediction models incorporating time-dependent risk factors. 2. To examine how changes in ABC pathway adherence over time impact the risk of adverse outcomes. 3. To model the trajectory of the CHA2DS2-VASc and HAS-BLED risk scores and other risk factors, and their associations with outcomes including incident stroke and major bleeding. 4. To evaluate the trajectory of the net clinical benefit of anticoagulation (the difference between the decrease in risk of a stroke on OAC and the increase in risk of a major bleed on OAC).
Health Outcomes to be Measured
Mortality; Ischaemic stroke or systemic embolism; Heart failure - incident or prevalent; Major bleeding; Cardiovascular disease; Myocardial infarction; Dementia; Hospitalisation – all-cause or cardiovascular

Deirdre Lane - Chief Investigator - University of Liverpool

David Stevens - Corresponding Applicant - University of Liverpool

Gregory Lip - Collaborator - University of Liverpool

Ruwanthi Kolamunnage-Dona - Collaborator - University of Liverpool

Stephanie Harrison - Collaborator - University of Liverpool

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