This chapter will review the indications for using device-based remote monitoring for detection and management of atrial fibrillation and will critically examine recent data regarding the specific clinical utility of this modality.

Atrial fibrillation is the most common sustained arrhythmia and, although it is often symptomatic, episodes may be silent in up to 70% of patients. Several large epidemiologic studies with long follow-up durations have shown strong associations
between atrial fibrillation and both the development of heart failure and increased mortality.¹,²,³,⁴ It has also been repeatedly demonstrated in patients with cardiac implanted electronic devices (CIEDs) that onset of atrial fibrillation is associated with heart failure exacerbation, increased frequency of hospitalizations, inappropriate implantable cardioverter defibrillator (ICD) shocks, loss of cardiac resynchronization therapy (CRT), increased sympathetic tone, hemodynamic compromise, and thromboembolism.⁵,⁶,⁷,⁸

The advent of CIEDs with accurate atrial arrhythmia diagnostic capabilities has revealed a large burden of atrial fibrillation that is present in the cardiac population. When atrial arrhythmias detected by implanted devices are asymptomatic, they have been referred to as silent atrial fibrillation, subclinical atrial fibrillation (SCAF), atrial high rate episodes (AHREs), or CIED-detected AHREs. Although there are subtle differences in the definitions of these terms, they have all come to be used interchangeably. Many studies have been completed, and many more are ongoing, to determine the correct treatment course when atrial arrhythmias are detected; this chapter will summarize this important literature.

Alongside the development of accurate atrial diagnostics within implanted devices has been the growth of multiple complex and proprietary networks of wireless home monitoring specific to each manufacturer. Utilizing Internet connectivity, it is now possible to review individual patients’ atrial arrhythmia burden on a daily basis. This capability has tremendous promise for patient care, with the possibility of reducing strokes, decreasing heart failure, preventing cardiomyopathies, and likely substantially reducing health care costs. This is critical, because lack of symptoms from atrial fibrillation does not translate into freedom from adverse sequelae. Furthermore, remote monitoring may allow assessment of the efficacy of patient treatment regimens and the opportunity to modify therapy early in the course of disease before atrial fibrillation becomes a chronic condition. As this innovative diagnostic capability is generating large amounts of data, protocols for triage and automated management of this glut of new information are being developed.

The concept of the thrombotic triad wherein intravascular clot formation is seen to be dependent on key conditions in the vasculature has been attributed to Rudolph Virchow, although not without question and controversy.⁹ This construct postulates that inflammatory or similar changes in vascular endothelium, alterations in the constitution of the blood itself, and reduced flow (“stasis”) in the affected chamber are all required for the clotting of blood to occur and propagate. This heuristically successful concept has been the basis of much research on the relationship between atrial fibrillation and stroke. However, the dominant focus of this research activity until very recently has been stasis, and there is a large literature describing and characterizing alterations in flow velocity that occur within the left atrium (and particularly the left atrial appendage) during atrial fibrillation.¹⁰,¹¹,¹²,¹³,¹⁴,¹⁵,¹⁶,¹⁷,¹⁸,¹⁹,²⁰,²¹

When cardioversion of atrial fibrillation was first introduced, the underlying pathology in most patients who underwent the procedure was rheumatic, often with significant mitral stenosis.²² In these patients, the risk of clinically apparent thromboembolic complications in the days and weeks following the procedure was worrisome²³ and was a major driver of studies showing that oral anticoagulation is beneficial, not just in the peri-procedural period, but for long-term management of atrial fibrillation in this context. It is now well established that oral anticoagulation is a highly effective intervention for stroke prevention in atrial fibrillation and that the benefits are greatest in those patients who have the highest baseline risk (such as those with mitral stenosis or high CHA₂DS₂-VASc scores).²⁴,²⁵,²⁶,²⁷,²⁸ The theoretical framework within which this work has been done views atrial fibrillation as a cause rather than as a risk factor for stroke; this traditional and somewhat simplistic paradigm that atrial fibrillation is a direct cause of stroke has much supporting evidence.
For example, retrospective data have driven the commonly employed “48-hour rule,” whereby episodes of atrial fibrillation of this or longer (or unknown) duration are subject to rigorous anticoagulation (often with acute administration of heparin) for stroke prevention whether or not cardioversion is contemplated.¹⁵

Currently, there are excellent data suggesting that atrial fibrillation also has indirect effects that are potentially thrombogenic; robust animal models of induced atrial fibrillation show that there are both early and late structural changes in the left atrium that contribute to clot formation.²⁹ Atrial fibrillation itself has the effect of depleting calcium in atrial myocytes and promoting the development of apoptosis and interstitial fibrosis in the atrial wall; the genetic mechanisms and contractile effects of atrial fibrillation on atrial myocytes are now well characterized.²⁹,³⁰,³¹,³²,³³,³⁴,³⁵,³⁶,³⁷,³⁸,³⁹,⁴⁰,⁴¹ Finally, there is mounting evidence of specific changes in both platelet function⁴² and procoagulant protein behaviors⁴³ within left atrial blood (compared with blood sampled from right atrium or peripheral veins) that promote coagulation when atrial fibrillation or rapid pacing (mimicking atrial fibrillation) is present compared with sinus rhythm. Therefore, there is strong and growing evidence that atrial fibrillation is a powerful amplifier of both structural and functional changes in the left atrium that predispose to thrombosis far beyond the simple effect of the arrhythmia on attenuating flow velocities within the left atrial appendage. Furthermore, it is now clear that atrial fibrillation is intimately associated with all three limbs of Virchow triad and that there is a complex interplay of mechanisms that serve to promote left atrial thrombosis and stroke in the context of this arrhythmia.

It is becoming increasingly understood that atrial fibrillation is only one of many powerful causes of the structural and functional changes occurring in the left atrium that predispose to clot formation. Patients with longstanding hypertension, diabetes, sleep-disordered breathing, and other conditions that promote inflammatory processes leading to fibrosis are all at increased risk for left atrial thrombosis whether atrial fibrillation has ever occurred or not.⁴⁴,⁴⁵ A recent provocative study of patients with chronic atherosclerotic cardiovascular disease (Cardiovascular Outcomes for People Using Anticoagulation Strategies [COMPASS] trial)⁴⁶ tested the hypothesis of empirically prescribing combination therapy with aspirin and rivaroxaban for the prevention of major cardiovascular outcomes. There was a very dramatic beneficial effect in the treatment arm on reducing the combined end point of myocardial infarction, stroke, and cardiovascular death; all of the beneficial impact on this composite end point was in the reduction of stroke events and was independent of rhythm status.⁴⁶

Therefore, it has been recently recognized that the presence of atrial fibrillation is neither a necessary nor a sufficient cause of arterial thromboembolism, and this evolving understanding has dramatically shifted the theoretical paradigm within which new technologies such as remote monitoring for atrial fibrillation should be evaluated and implemented in clinical practice.

Although the relationship between atrial fibrillation and stroke has been the dominant driver of the development of remote monitoring for atrial fibrillation, it must be acknowledged that there are additional well-established indications for such monitoring. The use of long-term monitoring with implanted devices can assess the effects of pharmacologic or procedural therapies for atrial fibrillation, and data from such follow-up monitoring have been an informal requirement for the acceptability of research reports; recent guidelines emphasize this point.⁴⁷ The Lumos-T Safely Reduces Routine Office Device Follow-Up (TRUST) study demonstrated very effectively that (any) arrhythmia detection is significantly accelerated by remote monitoring compared with clinic-based follow-up of patients with implanted CIEDs.⁴⁸,⁴⁹ Most implanted CIEDs now permit acquisition of summary data showing evolution of arrhythmia events over time (Figure 3.1). This can be very helpful in facilitating management decisions for patients.

Subcutaneously implanted long-term monitoring devices (implanted cardiac monitors, ICMs) with battery longevity in the 2- to 3-year range are now available.
These devices are very easy to place in the anterior chest wall using a simple injection technique and utilize reliable remote monitoring systems that allow for detailed, routine, and on-demand assessment of rhythm status.⁵⁰ However, ICMs are both expensive and subject to diagnostic errors in atrial fibrillation detection because the device is dependent on detection algorithms based on R-R interval identification. Reduced specificity of atrial fibrillation diagnosis can result from oversensing of skeletal myopotentials and both atrial and ventricular premature beats. Therefore, it is essential that physicians scrutinize data from these devices.⁵¹ Implanted pacemakers and defibrillators with a functioning right atrial lead provide direct electrogram recordings and are technically superior in diagnostic specificity compared with ICMs.⁴⁷

Patients with implanted CIEDs have been found to have an incidence of previously unrecognized atrial fibrillation ranging from 30% to 60%.⁵²,⁵³ The substantial variation in atrial fibrillation prevalence reflected in these reports is related to the patient populations being studied, as well as to the definitions of atrial fibrillation determined by the specific device technology being utilized. Data from a selection of key trials that determined the incidence of device-detected atrial fibrillation in patients with implanted devices are summarized in Table 3.1. Older studies found an incidence of device-detected atrial fibrillation in about half of the population; many of these studies used devices without electrogram recording capabilities, and data were derived from summary reports of AHREs that could not be independently scrutinized and adjudicated. In addition, the older studies included all device patients, even those with a known prior history of atrial fibrillation. Nevertheless, this information about AHRE burden was strikingly predictive of outcome events.⁵⁴,⁵⁵,⁵⁶,⁵⁷ Studies specifically designed
to exclude patients with a history of atrial fibrillation, oral anticoagulation, or antiarrhythmic drug use found an incidence of newly detected or “silent” atrial fibrillation in about 30% of this filtered population of device patients (see Table 3.1).⁵⁸,⁵⁹,⁶⁰

More recently, several studies have been completed in patients with stroke risk factors alone, and no indication for an implanted pacing device to determine the incidence of atrial fibrillation measured by ICMs. This approach excludes any possibility that atrial fibrillation episodes are caused by the atrial lead or that such episodes occur as a sequel of the pathologic process that was the indication for CIED implant. Consistent with data from CIED studies, these reports show an incidence of previously unrecognized atrial fibrillation again of approximately 30% in patients with stroke risk factors.⁶¹,⁶²,⁶³ These data are summarized in Table 3.2.