Should patients with diabetes be routinely screened for atrial fibrillation?

KEYWORDS:

• Atrial fibrillation
• diabetes mellitus
• stroke prevention
• screening

Atrial fibrillation (AF), the most common cardiac arrhythmia, is more likely to occur in patients with type 2 diabetes mellitus (DM2), with the risk increasing by 30–40% [1]. Higher levels of HbA1c levels are associated with a higher risk of developing AF among patients with DM2. Additionally, patients with DM2 tend to have diminished cardiac awareness, which may result in asymptomatic episodes of AF. Given the potential consequences of AF diagnosis in DM2 patients, this population may need systematic AF screening.

With the emergence of wearable heart monitors, population-based arrhythmia screening is closer than ever to becoming feasible. With the technology now matured, it is crucial to identify the appropriate target population for screening. To determine if patients with diabetes should undergo routine screening for atrial fibrillation, we need to systematically evaluate the screening process in a systematic and organized way.

First, we start with a general rationale for screening.

• What arrhythmias should we be screening for? Atrial fibrillation is the arrhythmia that makes the most sense, as AF is highly prevalent and finding AF carries therapeutic implications even in the absence of symptoms.

• Is it then plausible that screening for AF may help patients? To this, the answer is ‘yes.’ The studies published so far, even though with some limitations due to low sample size and other several bias, seem to suggest that implementation of screening strategies to identify atrial fibrillation with a subsequent prescription of OAC, appears to reduce the risk of outcomes over follow-up [2].

Second, we need to identify the population that should be screened with three key questions in mind:

• What is the expected yield in that population (i.e. how much AF will be found)?

• What are the ramifications of clinically asymptomatic AF in that population (i.e. what is their AF – related thromboembolism risk)?

• What is the cost of screening (or to flip the question, ‘What is the health care cost benefit of finding one case of AF and getting the patient on oral anticoagulation (OAC)?’)

We have learned from well-designed randomized clinical trials (RCTs), such as the SCREEN-AF trial, that screening older patients can result in higher detection rates for AF, leading to guideline-directed institutions of OAC [3].

Third, we need to determine the best way to screen patients. Currently, patch monitors and wearables (such as SmartWatches) hold the most promise. The commercially available Apple Watch Series 4 SmartWatch is among the market leaders and offers two mechanisms for AF detection (rhythm detection with R-R variability and single lead ECG). With rhythm detection, the positive predictive value of an irregular pulse was 0.84 for AF [4].

From a general accessibility standpoint, the SmartWatch paradigm makes the most sense. This provides longer- term monitoring with less actual cost to the patient. The issue is that data produced by the wearable needs to find its way to a physician who has a caring relationship with the patient. Further, the provider must know how to interpret and factor the data into a treatment plan. Mass Smartwatch screening programs have significant implications for the physicians asked to review the tracings; this raises some questions:

• How should information flow from a wearable device to an electronic health record?

• Are there enough such providers available to interpret and triage the data?

• What compensation do they receive for their time and expertise?

• What are the medical-legal implications for physicians when data sent to them is misinterpreted or not interpreted?

Fourth, we need to determine the duration threshold of AF that matters. The recently published NOAH-AFNET6 data show that in patients with a pacemaker or defibrillator, an atrial high-rate episode cutoff of ≥6 minutes is too aggressive for initiating OAC (edoxaban), which would be the primary intervention from the screening [5]. This study was terminated early for futility (no significant decrease in stroke risk with OAC) and safety (more bleeding with OAC). In contrast, results from the very recently published ARTESIA trial show that in patients with a moderate risk CHADS-VASc score, treatment of pacemaker/defibrillator diagnosed atrial high-rate episodes (with a cutoff of ≥6 minutes) with OAC (apixaban) reduces stroke/systemic embolism; of note, risk of bleeding was also increased in the OAC arm [6]. Further research (which may include sub-group analyses of NOAH-AFNET6 and ARTESIA) is needed to provide a better sense of the best cutoff. This is critical because having clear ‘positive’ and ‘negative’ criteria is essential to a good screening test.

Back then, to our original question of whether we should screen diabetic patients for AF. In our opinion, this makes sense for patients aged ≥75. Why? Given the high prevalence of AF in this patient population, the likelihood of detecting AF will be greater. For such patients, the CHADS-VASc score will be at least 3 due to age and DM2; hence, anticoagulation therapy will be indicated. In the future, with AI-generated algorithms, we may expand the target population by identifying a high-risk subset of younger patients with DM2 who warrant AF screening. How should we screen? Based on the results of the SCREEN-AF trial, sequential 2-week patch monitors in elderly patients yielded higher AF diagnosis than usual care; on the other hand, in the VITAL-AF trial [7], single-lead ECG screening at primary care visits in patients aged 65 and older did not increase AF diagnosis. When examining health care costs, it seems that a wearable (Smartwatch) would be more cost-effective than serial patch monitors, which are quite expensive. The main issue is establishing a workflow and reimbursement mechanism for providers over-reading rhythm tracing downloads from the wearable. When will the screening alter patient care? We do not have a good understanding of the time duration cutoff for AF that should trigger OAC. Sub-studies of the NOAH AFNET6 and ARTESIA trials could be helpful.

As our screening tools improve and risk stratification algorithms refine, we move closer to a healthcare environment where large-scale systematic AF screening can be feasible and cost-effective and improve patient outcomes.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.