Atrial Fibrillation (AF) is the most common heart rhythm disorder. People with AF are five times more likely to have a stroke compared to people without AF. AF-related strokes are more severe, more disabling and more often cause death. Many people with AF are prescribed medicines known as anticoagulants, which work to reduce the risk of an AF-related stroke. Unfortunately, anticoagulants also increase the risk of bleeding.
Doctors use two risk scores to help decide whether or not to prescribe anticoagulants to patients with AF: one gives the risk of AF-related stroke and the other gives the risk of major bleeding. However decision-making is not always easy, and many patients who need anticoagulants do not get them, and many patients who do not need anticoagulants are given them. A new tool, developed from data from a global AF study GARFIELD-AF, predicts the risk of AF-related stroke, bleeding and death in patients with AF. The aim of our study is to determine how well the GARFIELD-AF tool predicts AF-related stroke, bleeding and death in UK patients using anonymised electronic patient data from CPRD.
If successful, the tool will be implemented to work with electronic patient records in GP surgeries to help doctors decide who to give anticoagulants, so that all patients who need anticoagulants receive them, and patients who do not need anticoagulants do not receive them. This will help patients and the NHS by reducing AF-related strokes, deaths and bleeding in patients with AF.
Atrial fibrillation (AF), the most common clinically significant arrhythmia in the adult population, increases the risk of stroke five-fold. AF related strokes are more severe with increased mortality and disability than strokes in people without AF. Anticoagulation therapy can reduce up to two-thirds of this stroke risk and reduce mortality by up to 25%. However, anticoagulation increases the risk of bleeding. Risk scores are therefore used to assess stroke and bleeding risk to guide decisions on anticoagulation, and current NICE guidelines recommend the use of CHA2DS2VASc tool to assess stroke risk and HAS-BLED to assess bleeding risk.
The GARFIELD-AF tool is a novel risk prediction tool that simultaneously assesses a patients risk of mortality, AF-related stroke and major bleeding. It was developed using data from 39,898 patients in the global GARFIELD-AF registry and predicts mortality, stroke and bleeding better than the existing tools. The GARFIELD-AF tool offers a more accurate and integrated method to facilitate decisions on anticoagulation of patients with AF. Our study aims to validate the GARFIELD-AF tool using UK electronic primary care records.
We will use anonymised patient data from the CPRD, linked with Hospital Episode Statistics (HES) data, and ONS mortality data. We will include all patients aged ?18 years with a diagnosis of non-valvar AF with a minimum validated data for one year from date of diagnosis and up to 5 years from diagnosis. The primary outcome will be death from any cause, first recorded diagnosis of stroke or systemic embolism, or major bleeding within one year of diagnosis. We will also look at the 2-year and 5-year predictive ability of the tool.
Overall performance, discrimination and calibration will be assessed using the ROC statistic, the D statistic, the R2 statistic, the c-statistic and graphically.
All-cause mortality, Stroke and Systemic embolism, major Bleeding
David Fitzmaurice - Chief Investigator - University of Warwick
Patricia Apenteng - Corresponding Applicant - University of Warwick
David Prieto-Merino - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Diane Eaton - Collaborator - Anticoagulation UK
Karen Pieper - Collaborator - Thrombosis Research Institute
Keith Fox - Collaborator - University of Edinburgh
Rishi Caleyachetty - Collaborator - University of Warwick
Siew Wan Hee - Collaborator - University of Warwick
Trudie Lobban - Collaborator - Atrial Fibrillation Association
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation