Screening for Atrial Fibrillation

Our aim is to determine if screening for AF is effective and cost effective in reducing stroke and other key outcomes (both harm and benefit) compared to current practice. Our objectives are to: 1. Determine if screening for AF (including paroxysmal AF) is feasible in general practice in people aged 65 and over using a hand-held ECG recorder (Zenicor) 2. Perform an internal pilot cluster randomised controlled trial (RCT) to: -decide whether to proceed to an end-point powered cluster RCT -quantify psychological harms from screening -refine our screening programme 3. Perform a cluster RCT to determine if offering screening for AF (including paroxysmal AF) leads to reduced stroke and to quantify impact on mortality, cardiovascular disease and serious haemorrhage 4. Perform parallel qualitative studies to inform how to optimise screening both in the trial and if implemented nationally 5. Determine whether screening is cost effective Background 10% of ischaemic stroke is associated with newly diagnosed AF and a further 20% of ischaemic strokes are associated with known but untreated or undertreated AF. AF meets many of the classic criteria for screening, but there are weaknesses in the evidence base as identified by the UK Screening Committee. Uptake of anticoagulation is sub-optimal; screen-detected AF may have better prognosis than AF detected through routine clinical practice; and cost-effectiveness analyses (that consistently find that screening is cost-effective) have to make critical assumptions without RCT data on the effect of AF screening on stroke or mortality. An RCT powered to determine the impact of screening on clinical end-points such as stroke and mortality will resolve this controversy. Research Plan Workpackage (WP) 1: Feasibility and pilot work We will test feasibility of screening for AF (including paroxysmal AF) in six general practices in people aged 65 and over not on long term anticoagulation using a Zenicor hand-held ECG recorder. After a one-off screen, participants will be invited to check their heart rhythm using the ECG recorder over a two week period. We will test different recruitment methods (opportunistic; invitation sent by practice; patient invited to contact practice) with the aim of screening up to 400 patients per practice over a six month period. We will record key process criteria: proportion who consent to take part; proportion who complete both one-off and paroxysmal AF screening; proportion with AF identified; and use the results to refine the target population by reporting these in age strata. In an internal pilot, we will randomise larger practices (list size >7000) - 10 to screening and 20 to usual care. Prior to randomisation, we will send an information sheet and consent form to 800 patients in each practice that meet the inclusion criteria. Baseline data will be obtained from the GP electronic medical record in people who give consent. We will flag records with NHS digital for mortality and Hospital Episode Statistics (HES), and link with national stroke and myocardial infarction audits (SSNAP and MINAP). After twelve months, we will obtain data on new diagnoses of AF and use of anticoagulation from the GP electronic medical record. We will send postal questionnaires to a sample of 1,700 participants to assess psychological effects of screening. Key outcomes will be: AF detection rate; anticoagulation uptake rate, stratified by stroke risk (CHA2DS2-VASc score) and by AF type (paroxysmal or permanent). Parallel qualitative work (see WP3) will be carried out in both the feasibility and internal pilot to inform the design of the screening intervention. WP2: End-point powered cluster RCT We will recruit further practices to aim for 120,000 participants from 100 intervention and 200 control practices. We will follow up for an average of five years. Primary end-point: stroke. Secondary end points: all-cause mortality; cardiovascular events; bleeding requiring