How to best prevent cardioembolic stroke?

Strokes affect 17 million people worldwide each year, and are a major cause of mortality and long-term disability. More than two-thirds of all strokes are due to cerebral ischaemia.1 Improvements in medical management of cardiovascular risk factors in the community have led to a significant decline in strokes caused by large artery atherosclerosis and small vessel disease. As a result, cardioembolic stroke has increased significantly and is now the leading cause of cerebral ischaemia in some populations.2 Atrial fibrillation is associated with a three- to five-fold increased risk of cardioembolic stroke. The prevalence of atrial fibrillation increases sharply with age, affecting almost 10% of individuals aged >80 years. Due to an aging population, the number of patients with atrial fibrillation may double and the number of atrial fibrillation-related strokes may triple in the next decades based on projections from high-income countries.3 Ischaemic stroke associated with atrial fibrillation is nearly twice as likely to be fatal as non-atrial fibrillation stroke. Recurrence is more frequent, and functional deficits are more likely to be severe among survivors.4 Since stroke is often the initial manifestation of embolism in atrial fibrillation, early detection is crucial for reducing disability and mortality. Current European Society of Cardiology guidelines recommend opportunistic screening for silent atrial fibrillation by pulse taking or electrocardiogram (ECG) rhythm strip in patients >65 years of age. Systematic screening may be considered in patients aged >75 years, or those at high stroke risk.5 Most of the evidence for opportunistic screening of patients aged >65 years comes from a single clustered randomized controlled trial, performed in 50 primary care centres in England (Screening for Atrial Fibrillation in the Elderly; SAFE trial).6 Practical nurses were trained in ECG recording and interpretation. A 12-lead ECG was only performed in the case of an irregular radial pulse. The detection rate of new cases of atrial fibrillation was 1.63% a year in the intervention practices and 1.04% in control practices. Systematic and opportunistic screening detected similar numbers of new cases.6 Interestingly, the study revealed that general practitioners could not accurately detect atrial fibrillation (sensitivity 80%, specificity 92%). The 12-lead ECG interpretation software showed slightly better results (sensitivity 83%, specificity 99%), but was not sufficiently accurate.7 Probably as a consequence, a recent US Preventive Services Task Force (USPSTF) Recommendation Statement concluded that the current evidence is insufficient to assess the balance of benefits and harms (e.g. ECG misinterpretation, misdiagnoses and unnecessary treatments) of screening for atrial fibrillation. Importantly, the USPSTF statement did not consider the costs.8 Recent technology advances have improved devices and algorithms for heart rhythm monitoring and atrial fibrillation detection, possibly reducing the burden of misdiagnoses. Blood pressure monitors and non-12-lead ECG devices have demonstrated a superior accuracy over pulse palpation.9 Single-lead ECGs connected to smartphones or smartwatches show accurate atrial fibrillation detection with sensitivity and specificity values >95% in large and diverse cohorts.10 Intermittent recordings from single-lead ECG devices produce an atrial fibrillation detection similar to 24-h Holter monitoring.11 Despite the limitations of the SAFE trial, a recent systematic review (of which the economic findings were based largely on the SAFE trial) concluded that a national screening programme for atrial fibrillation is likely to represent a cost-effective use of resources.12 This study suggested that opportunistic screening would be more likely to be cost-effective than systematic population screening. Nurse pulse palpation or modified blood pressure monitors would be appropriate screening tests, with confirmation by diagnostic 12-lead ECG interpreted by a trained General Practitioner, with referral to a specialist in the case of an unclear diagnosis.12 Against this background, the study from Proietti and colleagues adds an important piece of evidence in support of cost-effectiveness of systematic population screening for atrial fibrillation.13 They used a Markov-model based modelling analysis in 1000 hypothetical individuals that matched a national atrial fibrillation screening programme of the general population in Belgium (20 to >90 years). A nurse-led ECG screening with a validated handheld ECG (one single 30-s recording) allowed identification of 1.1% new atrial fibrillation patients.14 Their model accounted for a screening procedure undertaken yearly for 40 consecutive years and simulated the natural dynamics of the cohort. Given a willingness to pay for a quality-adjusted life year (QALY) of €4000, the screening programme was cost-effective not only for subjects ≥65 years and subjects ≥75 years, but also for the general population. Two points deserve special attention, which may have caused overestimation of the gains of the screening programme. The authors acknowledged both points in their limitations. First, the ‘Belgian Heart Rhythm Week’ was a national campaign on awareness about cardiac arrhythmias. It is reasonable to speculate that the campaign increased the chance of selection bias, since people with some unspecific symptoms may have been more interested in having a heart check. This is supported by the notion that in the age groups 30 to 34 years and 35 to 39 years the prevalence of atrial fibrillation was unusually high (3.6% and 2.3%, respectively).14 Thus, the studied population is unlikely to be representative of the general population. The unusually high prevalence of atrial fibrillation in the young may have favoured the conclusion that screening was cost-effective in the general population. A comparable atrial fibrillation screening study with handheld ECG devices was performed in The Netherlands. In this cohort, screening was offered to all individuals ≥65 years of age during seasonal influenza vaccination, which reduced the risk of selection bias. Also the Dutch screening programme was found to be cost-effective, but only at a willingness-to-pay of €20,000/QALY and at a higher incremental cost-effectiveness ratio per QALY.15 Secondly, Proietti and colleagues made the assumption that in all patients diagnosed with new onset atrial fibrillation, a generic non-vitamin K oral anticoagulant (NOAC) would be prescribed. Safety and efficacy data for the model were derived from randomized controlled trials. However, the proportion of NOAC users among ordinary atrial fibrillation patients is substantially lower than 100%16 and short-term clinical trial data do not always reflect long-term real-life efficacy and safety. In conclusion, the work of Proietti and colleagues adds to the evidence that systematic, and not only opportunistic, screening for atrial fibrillation is feasible and likely to be cost-effective in the population ≥65 years of age using modern devices. Even if implementation in guidelines may need additional evidence, rapid development of smartphone-based devices and algorithms, together with the increased awareness among customers, will enhance unsystematic population-based screening for atrial fibrillation.10 Using handheld ECG devices repetitively instead of once during a scheduled screening may further improve the atrial fibrillation detection rate.11 Using photoplethysmographic signals derived from built-in cameras of smartphones and smartwatches is an alternative approach to monitor the heart rhythm. While the accuracy of measuring the exact heart rate may be low, depending on the application,17 detection of an irregular heart rhythm compatible with atrial fibrillation seems possible with high sensitivity and specificity (both >98%) in selected populations.10 Novel signal processing approaches and machine-learning algorithms are needed to overcome problems like motion artefacts. However, continuous monitoring of the heart rhythm and early detection of arrhythmias with a smartwatch seems feasible in the future.10 Importantly, atrial fibrillation screening will only reduce the burden of cardioembolic stroke when combined with proper and complete atrial fibrillation treatment according to guidelines.5 A multifaceted and multilevel educational intervention aimed to improve use of oral anticoagulation in patients with atrial fibrillation and at risk of stroke resulted in a significant increase in the proportion of patients treated with oral anticoagulants (68% to 80% in one year). The intervention led to a significant reduction of stroke by 52%.16 This highlights the three important pillars of cardioembolic stroke prevention: early detection, guideline-based treatment, and educational interventions aimed to improve use of guideline-based treatment (Figure 1). The three pillars of cardioembolic stroke prevention. ECG: electrocardiogram; AF: atrial fibrillation. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author(s) received no financial support for the research, authorship, and/or publication of this article.