Intelligent Monitoring to Predict Atrial Fibrillation

Overview

Atrial Fibrillation (AF) is the commonest arrhythmia worldwide, affects 5% of people over the age of 65 and increases the risk of stroke and heart failure. The investigators aim to detect clinical and subclinical episodes of atrial fibrillation lasting \>30 seconds to develop risk prediction models to identify patients at high risk for ischaemic stroke.

Atrial Fibrillation (AF) is the commonest arrhythmia worldwide, affects 5% of people over the age of 65 and increases the risk of stroke and heart failure. Among acutely unwell patients; arrhythmias and myocardial injury are common and associated with increased mortality, morbidity, and healthcare costs. Cardiovascular comorbidities in these high-risk patients include hypertension (47%), dyslipidaemia (29%) and ischaemic heart disease (11%). The investigators aim to detect clinical and subclinical episodes of atrial fibrillation lasting 30 seconds to develop risk prediction models to identify patients at high risk for ischaemic stroke. Data will serve to develop and validate bedside clinical decision support tools and digital twins. Patients who develop episodes of AF as part of acute illness, will suffer further episodes of AF within one year in over 20% of cases with 27% progressing to paroxysmal/permanent AF. The true incidence of AF is unknown in acutely unwell patients as a significant percentage of AF episodes remain undetected with conventional intermittent monitoring. Patients experiencing short self-terminating episodes of AF carry a 5-fold risk of developing continuous AF and double the risk of stroke and thromboembolic events. Patients suffering episodes of AF often remain asymptomatic but are at increased risk of heart failure and death at one year. Compared to routine intermittent manual measurement of vital signs, wireless continuous vital sign monitoring systems (wCVSM) detect deviations instantaneously with the option of alerting clinical staff in real time via mobile phone applications. Accurate categorization of alerts into false and true events is essential for developing intelligent software that can be embedded into monitoring systems. Continuous ECG and vital signs monitoring can detect AF episodes more reliable, trigger timely investigations and support longer term treatment plans. Changes in patient pathways and introduction of novel devices to alert healthcare staff on the potential of clinical events require buy-in from all stakeholders. It is therefore essential to evaluate user acceptance and to determine perceptions of users before rolling out a novel patient pathways or implementation of a new device within an organization. The investigators therefore wish to explore users\' views of the device, wearing the device and potential areas for improvement using questionnaires for patients and health care staff and by conducting semi-structured interviews with healthcare staff. Primary objective To determine the true cardiovascular event rate (defined as at least one of the following criteria: episodes of AF, New Regional Wall Motion Abnormalities, raised cardiac biomarkers hs-troponin T and NT-pro-BNP) versus false cardiovascular events detected by continuous wireless remote monitoring. Secondary objective To determine patient acceptability and usability for health care professionals of a novel remote monitoring device with automated alert function.

Conditions

Interventions

Eligibility

Locations (2)

Outcomes