The use of outpatient cardiac telemetry (also known as mobile cardiac outpatient telemetry or MCOT) as a diagnostic alternative in patients who experience infrequent symptoms (less frequently than every 48 hours) suggestive of cardiac arrhythmias (i.e., palpitations, dizziness, presyncope, or syncope) does not meet Primary Coverage Criteria for cost-effectiveness.  There is a lack of data indicating this testing is more effective than other cardiac event recorders (e.g. 30 day recording) and the procedure is more costly.  

 

Other uses of outpatient cardiac telemetry, including but not limited to monitoring effectiveness of antiarrhythmic therapy and detection of myocardial ischemia by detecting ST segment changes, does not meet Primary Coverage Criteria.

 

Published guidelines of national and/or international workgroups of medical experts, widely used medical compendia, or technology assessments published by independent technology assessment organizations have classified non-covered uses as investigational or as questionable or of unknown benefit.

 

For any contract without Primary Coverage Criteria, the use of outpatient cardiac telemetry (also known as mobile cardiac outpatient telemetry or MCOT)  as a diagnostic alternative in patients who experience infrequent symptoms (less frequently than every 48 hours) suggestive of cardiac arrhythmias (i.e., palpitations, dizziness, presyncope, or syncope) or for any other reason,  including but not limited to monitoring effectiveness of antiarrhythmic therapy and detection of myocardial ischemia by detecting ST segment changes, is considered investigational and is not covered.  Investigational services are an exclusion in the member benefit certificate.

AEMs are a well-established technology that are most typically used to evaluate episodes of cardiac symptoms (palpitations, dizziness, syncope), which, due to their infrequency, would escape detection on a standard 24- to 48-hour Holter monitor. Other proposed uses include monitoring the efficacy of antiarrhythmic therapy and evaluating ST segment changes as an indication of myocardial ischemia. However evidence is inadequate to validate these uses of AEMs. Although serial EKG monitoring has often been used to guide antiarrhythmic therapy in patients with symptomatic sustained ventricular arrhythmias or survivors of near sudden cardiac death, it is not known what level of reduction of arrhythmic events constitute successful drug therapy. Furthermore, the patient’s cardiac activity must be evaluated before and during treatment, such that the patient can serve as his or her own control. The routine monitoring of asymptomatic patients after myocardial infarction is additionally controversial, especially after the Cardiac Arrhythmia Suppression Trial (CAST) showed that patients treated with encainide or flecainide actually had a higher mortality. While Holter monitoring has been used to detect ST segment changes, it is unclear whether ST segment changes can be reliably detected by an AEM. The interpretation of ST segment change is limited by instability of the isoelectric line, which is in turn dependent on meticulous attention to skin preparation, electrode attachment, and measures to reduce cable movement.

 

The published literature regarding outpatient cardiac telemetry was reviewed, with a specific focus on whether or not outpatient cardiac telemetry was associated with incremental benefit compared to the use of ambulatory event monitors. Of specific interest was the benefit of real-time monitoring in an ambulatory population, presumably considered to be at a lower level of risk from significant arrhythmia such that an electrophysiologic study or inpatient telemetry was not required. One published study was identified that described the outcomes of a consecutive case series of 100 patients.  Patients with a variety of symptoms were included, including most commonly palpitations (47%), efficacy of drug treatment (25%), dizziness (24%), or syncope (19%). Clinically significant arrhythmias were detected in 51% of patients, but half of these patients were asymptomatic. The authors comment that the automatic detection results in an increased diagnostic yield, but there was no discussion of its unique feature, i.e., the real-time analysis, transmission, and notification of arrhythmia.

 

One recent study reported on the potential role of mobile telemetry in the follow-up of patients who have ablation procedures for atrial fibrillation.  Studies presented as meeting abstracts suggest that mobile cardiac outpatient telemetry (MCOT) can detect more significant arrhythmias than standard loop recordings, but the impact of these findings on clinical outcomes is uncertain. In addition, the role of this device in the diagnosis and treatment strategy of patients with possible cardiac arrhythmias is unknown.

 

Rothman and colleagues recently reported results comparing MCOT to standard loop recording.  This study involved 305 patients who were randomized to the LOOP recorder or MCOT and who were monitored for up to 30 days. The unblinded study enrolled patients at 17 centers for whom the investigators had a strong suspicion of an arrhythmic cause of symptoms including those with symptoms of syncope, presyncope, or severe palpitations occurring less frequently than once per 24 hours and a nondiagnostic 24-hour Holter or telemetry monitor within the prior 45 days. Test results were read in a blinded fashion by an electrophysiologist. Study exclusions were Class IV heart failure, myocardial infarction within the prior 3 months, unstable angina, history of sustained ventricular tachycardia or ventricular fibrillation, and complex ectopy with an ejection fraction less than 35%.

 

While 305 patients were randomized, results from 266 were analyzed using patients who completed at least 25 days of monitoring, 132 in the LOOP group and 134 in the MCOT group. Of the 39 patients who did not complete the protocol, 20 (13 MCOT and 7 LOOP) did not complete the study due to non-compliance (non-wearing) with the device. Patients were predominantly female with a mean age of 56 years. Approximately 20% had a history of heart disease, 50% hypertension, and 5% heart failure.

 

A diagnostic endpoint (confirmation/exclusion of arrhythmic cause of symptoms) was found in 88% of MCOT patients and 75% of LOOP patients (p = 0.008). The difference in rates was due primarily to detection of asymptomatic (not associated with simultaneous symptoms) arrhythmias in the MCOT group consisting of rapid atrial fibrillation and/or flutter (15 patients vs. 1 patient) and ventricular tachycardia defined as more than 3 beats and rate greater than 100 (14 patients vs. 2 patients). These were thought to be clinically significant rhythm disturbances and the likely causes of the patients’ symptoms. The paper does not comment on the clinical impact (changes in management) of these findings in patients for whom the rhythm disturbance did not occur simultaneously with symptoms. In this study, the median time to diagnosis in the total study population was 7 days in the MCOT group and 9 days in the LOOP group. Of note, prior studies of the autotrigger loop recorder have also shown similar findings that are viewed as improvements over traditional LOOP recordings that require patient activation.

 

A subset of only 50 patients (related to device availability) received the newer autotrigger loop recorders. This study protocol does not allow for adequate comparisons between the autotrigger and the MCOT device.

 

The autotrigger loop recorders have become a part of the standard diagnostic approach to patients who have infrequent symptoms that are thought likely to be due to arrhythmias. Therefore, this is the test to which newer technologies must be compared. Currently, the literature does not provide any adequate comparative data for the autotrigger device compared to mobile cardiac telemetry. Further study of MCOT is needed to replicate the Rothman study and to compare MCOT with the autotrigger loop recorder.

 

There are limited data to show any potential value for the continuous monitoring aspects of MCOT.

 

 

 

 

 

 

 

The available published evidence is considered insufficient to determine that MCOT improves the net health outcome for patients with suspected arrhythmias or for the detection of AF in patients following catheter ablation or following cryptogenic stroke.    

 

 

In 2016, Solomon and colleagues evaluated the diagnostic yield for potentially high-risk arrhythmias with 14 days of continuous recording with the Zio Patch among 122,454 patients (122,815 recordings) included in a manufacturer registry (Solomon, 2016). Patients included in the series all underwent monitoring with the device from November 2011 to December 2013. Mean wear time was 9.6 days. Overall, there were 22,443 (18%) patients with sustained ventricular tachycardia, 1766 (1.4%) patients with sinus pauses of 3 seconds or more, 521 (0.4%) patients with AF pauses of 3 seconds or more, 249 (0.2%) patients with symptomatic pauses, and 1468 (0.4%) with high-grade heart block, which were considered potentially high-risk arrhythmias. After 24 and 48 hours of monitoring, 52.5% and 65.5%, respectively, of potentially high-risk arrhythmias were detected. Seven days of monitoring identified 92.9% of potentially high-risk arrhythmias.

 

 

 

A prospective study that evaluated the incidence of asymptomatic AF episodes following AF ablation (using an implantable cardiac monitoring) followed 50 patients with cardiac monitoring over 18 months postablation (Verma, 2013). Based on symptoms alone, 29 (58%) of 50 patients were arrhythmia-free after ablation; based on occurrence of symptoms or the detection of AF on intermittent (every 3 month) ECG or Holter monitor, 28 (56%) patients were arrhythmia-free postablation. Six (12%) patients had arrhythmias detected on implantable monitoring alone.

 

Brachmann and colleagues reported on 3-year follow-up from the CRYSTAL-AF trial in 2016 (Brachmann, 2016). At the closure of the trial, 48 subjects had completed 3 years of follow-up (n=24 in each treatment group). By 3 years, the HR for detecting AF for ICM-monitored versus control patients was 8.8 (95% CI, 3.5 to 22.2; p<0.001).

 

Giada and colleagues conducted an RCT of 2 diagnostic strategies in 50 patients with infrequent (≤1 episode per month) unexplained palpitations: an ILR strategy (n=26) vs a conventional strategy (n=24) including 24-hour Holter, 4 weeks of ambulatory ECG monitoring with an external recorder, and an electrophysiologic study if the 2 prior evaluations were negative) (Giada, 2007). Prior cardiac evaluation in eligible patients included standard ECG and echocardiography. Rhythm monitoring was considered diagnostic when a symptom-rhythm correlation was demonstrated during spontaneous palpitations that resembled pre-enrollment symptoms. In the conventional strategy group, a diagnosis was made in 5 (21%) subjects, after a mean time to diagnosis of 36 (±25) days, based on external ECG monitoring in 2 subjects and electrophysiologic studies in 3 subjects. In the ILR group, a diagnosis was made in 19 subjects (73%; vs conventional group, p<0.001) after a mean time to diagnosis of 279 (±228) days.

 

In 2004, Farwell and colleagues reported results of an RCT comparing the diagnostic yield of an ILR (Reveal Plus, Medtronic) with a conventional diagnostic strategy in 201 patients with unexplained syncope (Farwell, 2004). Eligible patients were evaluated at a single institution for recurrent syncope and had no definitive diagnosis after a basic initial workup (including 12-lead ECG, Holter monitoring in patients with suspected cardiac syncope, upright cardiac sinus massage, and tilt-table testing). At last follow-up, more loop recorder patients (33%) had an ECG diagnosis than control patients (4%; HR for ECG diagnosis; 8.93; 95% CI, 3.17 to 25.19; p<0.001).Seven of the loop recorder patients had a diagnosis made with the device’s auto-trigger feature. In the loop recorder group, 34 patients had an ECG-directed therapy initiated (vs 4 in the control group; HR=7.9; 95% CI, 2.8 to 22.3). No device-related adverse events were reported.

 

 

 

 

 

Hindricks and colleagues evaluated the accuracy of auto-triggered ILRs in 247 patients at high risk for paroxysmal AF (Hindricks, 2010). All patients underwent simultaneous 46-hour continuous Holter monitoring, and the authors calculated the performance characteristics of the loop recorder using physician-interpreted Holter monitoring as the criterion standard. The sensitivity of the loop recorder for detecting AF episodes of 2 minutes or more in duration was 88.2%, rising to 92.1% for episodes of 6 minutes or more. AF was falsely identified by the loop recorder in 19 of 130 patients who did not have AF on Holter monitoring, for a false-positive rate of 15%. AF burden was accurately measured by the loop recorder, with the mean absolute difference between the loop recorder and Holter monitor of 1.4% (±6.4%).

 

Hanke and colleagues compared an auto-trigger ILR with 24-hour Holter monitoring done at 3-month intervals in 45 patients who had undergone surgical ablation for AF (Hanke, 2009). After a mean follow-up of 8.3 months, the ILR identified AF in 19 (42%) patients in whom Holter monitoring recorded sinus rhythm.

 

In 2015, Afzal and colleagues reported on a systematic review and meta-analysis of studies comparing ILRs with wearable AEMs for prolonged outpatient rhythm monitoring after cryptogenic stroke (Afzal, 2015). The review included 16 studies (total N=1770 patients): 3 RCTs and 13 observational studies. For ILR-monitored patients, the median monitoring duration was 365 days (range, 50-569 days), while for wearable device-monitored patients, the median monitoring duration was 14 days (range, 4-30 days). Compared with wearable device AEMs, ILRs were associated with significantly higher rates of AF detection (23.3% vs 13.6%; odds ratio, 4.54; 95% CI, 2.92 to 7.06; p<0.05).

 

In 2015, Ziegler and colleagues reported on a large (N=1247) set of patients undergoing ILR monitoring for AF detection after a cryptogenic stroke who were identified from the manufacturer’s registry (Ziegler, 2015). Over a median follow-up of 182 days, a total of 1521 episodes of AF were detected in 147 patients. Overall, 42 (29%) patients had a single episode of AF and 105 (71%) patients had multiple episodes. The overall detection rate 12.2% (at 182 days) was somewhat higher than that reported in CRYSTAL AF.

 

Sanders and colleagues reported on the diagnostic yield for AF with the Reveal Linq device, a miniaturized ILR with a detection algorithm designed to detect AF in a nonrandomized, prospective trial (Sanders, 2016). The study included 151 patients, most of whom (81.5%) were undergoing monitoring for AF ablation or AF management. Compared with Holter-detected AF, the ILR had a diagnostic sensitivity and specificity for AF of 97.4% and 97.0%.

 

In 2015, Mittal and colleagues reported on safety outcomes related to the use of an ILR, the Reveal LINQ device, based on data from 2 studies, the Reveal LINQ Usability study and the Reveal LINQ Registry (Mittal, 2015). The Usability study enrolled 151 patients at 16 European and Australian centers; adverse events were reported for the first month of follow-up. The Registry is a multicenter post-marketing surveillance registry, with a planned enrollment of at least 1200. At the time of analysis, 161 patients had been enrolled. For Registry patients, all adverse events were recorded when they occurred. The device version used in these studies measures 7 × 45 × 4 mm3, and is inserted with a preloaded insertion tool via a small skin incision. In the Usability study, 1 serious adverse event was recorded (insertion site pain); in the Registry study, 2 serious adverse events were recorded (1 case each of insertion site pain and insertion site infection). The rates of infection and procedure-related serious adverse events in the Usability study were 1.3% and 0.7%, respectively, and were 1.6% and 1.6%, respectively, in the Registry study.

 

In an uncontrolled case series, Tayal and colleagues retrospectively analyzed patients with cryptogenic stroke who had not been diagnosed with AF by standard monitoring (Tayal, 2008). In this study, 13 (23%) of 56 patients with cryptogenic stroke had AF with MCOT. Twenty-seven asymptomatic AF episodes were detected in the 13 patients; 23 of these were less than 30 seconds in duration. In contrast, Kalani and colleagues reported a diagnostic yield for AF of 4.7% (95% CI, 1.5% to 11.9%) in a series of 85 patients with cryptogenic stroke (Kalani, 2015). In this series, 82.4% of patients had completed transesophageal echocardiography, cardiac magnetic resonance imaging (cMRI), or both, with negative results; the authors proposed that the use of advanced cardiac imaging may alter the underlying prevalence of AF in patients labeled as having cryptogenic stroke.

 

 

 

 

Assessing the prevalence of asymptomatic AF is difficult because of the lack of symptoms. Those who are asymptomatic have been estimated to constitute approximately a third of all patients with AF (Savelieva, 2000). Studies have suggested that most paroxysmal episodes of AF are asymptomatic (Israel, 2004; Page, 1994). It is uncertain whether patients with paroxysmal AF have a stroke risk comparable to those with persistent or permanent AF; some studies have suggested the risk of stroke is similar (Hart, 2000; Hohnloser, 2007) while a 2016 systematic review of 12 studies (total N=99,996 patients) suggested the risks of thromboembolism and all-cause mortality were higher with nonparoxysmal compared to paroxysmal AF (Ganesan, 2016). The clinical management of symptomatic and asymptomatic AF is the same. Anticoagulation should be initiated if reduction in risk of embolization exceeds complications due to increase bleeding risk regardless of AF symptoms.

 

Screening for AF in asymptomatic patients could be either systematic or targeted to high-risk populations. European guidelines for screening for AF are based on a 2007 large‐cluster RCT of opportunistic pulse taking versus systematic screening with 12‐lead ECG or standard care in general practice. This RCT showed that systematic and opportunistic screening detected similar rates of AF and both were superior to standard care (Fitzmaurice, 2007). The mechanisms of how and when to screen for AF in unselected populations have not been well-studied.

 

In 2015, Turakhia et al reported results for a single-center noncomparative study evaluating the feasibility and diagnostic yield of a continuous recording device with longer recording period (Zio Patch) for patients with risk factors for AF (Turakhia, 2015). The study included 75 patients older than age 55 with at least 2 of risk factors for AF (coronary disease, heart failure, hypertension, diabetes, or sleep apnea), without a history of prior AF, stroke, TIA, implantable pacemaker or defibrillator, or palpitations or syncope in the prior year. Of the 75 subjects, 32% had a history of significant valvular disease and 9.3% had prior valve replacement. Most subjects were considered at moderate to high risk of stroke (CHA2DS2-VASc scores ≥2 in 97% of subjects). AF was detected in 4 (5.3%) subjects, all of whom had CHA2DS2-VASc scores of 2 or greater. All patients with AF detected had an initial episode within the first 48 hours of monitoring. Five patients had detected episodes of atrial tachyarrhythmias lasting at least 60 seconds.