Aims To evaluate the impact of Home Monitoring™ (HM) technology on detection and treatment of atrial fibrillation (AF).
Methods and results One hundred and sixty-six patients (mean age 73 ± 10 years, 112 males) received HM devices [121 pacemakers, 22 implantable cardioverter defibrillators (ICDs), and 23 ICDs combined with cardiac resynchronization therapy]. A nurse consulted the cardioreports on the website and submitted the alerts for AF (new onset AF, persistent AF, and AF burden of >10% for >5 consecutive days) to the physician. During a mean follow-up of 488 ± 203 days, 42 patients (26%) had alerts for AF. Twenty-two patients had no history of AF before implant. In nine patients, no further action was taken. For the remaining 33 patients, an unscheduled follow-up was performed: in 16, antiarrhythmic drug therapy was introduced/modified; in 15, anticoagulation was started; in 2, antiplatelet drugs were introduced; in 7, an external cardioversion was performed; in 2 patients, the device was reprogrammed to avoid intermittent atrial undersensing during AF; and in 4 patients, no further action was taken. In four cases the arrhythmia was not confirmed (false positive). The median time to the first intervention for AF was 50 days (148 days before the scheduled follow-up).
Conclusion The HM technology allowed early detection of AF in paced patients and early reaction to optimize medical treatment.
Atrial fibrillation is the most common arrhythmia in clinical practice and is associated with increased morbidity and mortality.1–3 It has been demonstrated that many patients with pacemakers and implantable cardioverter defibrillators (ICDs) have arrhythmia recurrences during the follow-up, also when there is no history of atrial arrhythmias at implant, and that many episodes are asymptomatic.4,5 Furthermore, it has also been shown that either a high burden of atrial fibrillation, regardless of symptoms, or arrhythmia episodes lasting more than 24 h are independent predictors for stroke and mortality.6–8
Diagnostics of last-generation devices allow us to do detailed monitoring of arrhythmia, including asymptomatic episodes.9,10 However, benefits of continuous monitoring of atrial fibrillation by implanted devices are limited because data can be reviewed only during scheduled follow-ups (usually every 6 months) or in case of patient symptoms.
Home Monitoring™ (HM) technology (Biotronik GmbH & Co. KG, Berlin, Germany) is a long-distance telemetry system that provides automatic transmission of data stored in the pacemaker memory on daily basis to a Service Centre. Here, information is decrypted and uploaded in a secure website, which can be accessed by the physicians. Automatic alerts are sent to the clinical staff in case of trigger events.11
A potential benefit of the HM technology in patients with atrial fibrillation is the prevention of arrhythmia-related adverse events by early detection and early reaction to arrhythmia recurrences mainly for asymptomatic episodes.12–14 Furthermore, unnecessary out-patient visits may be reduced and healthcare resource allocation may be optimized.15–17
Aim of our study was to evaluate the impact of the HM technology on detection and treatment of atrial fibrillation in pacemaker and ICD patients.
From April 2006 to March 2008, 166 patients (mean age 73 ± 10 years, 112 males) received HM devices [121 dual-chamber pacemakers, 22 dual-chamber ICDs, and 23 ICDs combined with cardiac resynchronization therapy (CRT-D)]. In the pacemaker group, implant indications were sinus node disease in 55% patients, atrio-ventricular or intra-ventricular blocks in 24%, and neuromediated syncope in 21%. In the ICD group, implant indications were primary prevention of sudden cardiac death in 73% and secondary prevention in 27%. Forty-five patients in the pacemaker group (37%) and 4 in the ICD group (9%) had a history of atrial fibrillation before implant. Clinical characteristics of enrolled patients are summarized in the Table 1. Antithrombotic and antiarrhythmic drug therapy at implant is reported in the Table 2.
Antithrombotic and antiarrhythmic drug therapy at implant
Class 1A antiarrhythmics
Class 1C antiarrhythmics
Main features of Home MonitoringTM
The main features of HM have been described previously.17 In summary, data are retrieved from the implanted device through a wireless receiver for long-distance telemetry and forwarded to a unique Service Centre, by connecting to the GSM (Global System for Mobile Communication) network in the form of encrypted ‘short message system’ (sms) messages. The Service Centre anonymously decodes, analyses, and organizes the data and posts them on a secure website. Transmission is automatic and patient independent. Reports and alerts are available for the clinical staff on daily basis. The HM features transmits data every night and notifies Event Report whenever pre-selected critical events have occurred.
Device and alerts programming
Pacemaker and ICD programming was tailored according to the patient's individual clinical profile. Mode switch function was switched on with an atrial fibrillation detection rate of 160 bpm. Spontaneous atrio-ventricular conduction was preserved as far as possible either by programming a long atrio-ventricular delay or by switching on the IRS plus algorithm, which periodically checks for spontaneous atrio-ventricular conduction and, if present, automatically lengthens the atrio-ventricular delay. The atrial sensitivity value was set by applying a safety margin of 3 after measuring the atrial electrogram amplitude in sinus rhythm. Anyway, the values of atrial sensitivity higher than 1.0 mV were not allowed. The total atrial refractory period was programmed at the nominal value of 425 ms. The post-ventricular atrial blanking was programmed at the nominal value of 56 ms. The pacemaker used in the study is provided by an internal electrogram recording of atrial tachyarrhythmias, high ventricular rate episodes, and pacemaker-mediated tachycardias. At each follow-up, all internal electrogram recordings were reviewed and classified as true arrhythmias or R-wave far-field oversensing.
All the patients received and were trained to use the transmitter Cardiomessenger, which is designed to retrieve data from the implanted device and to forward them to the Service Centre. All enrolled patients provided written informed consent on the use of HM system and data management.
Alerts for atrial fibrillation were defined as follows: new onset atrial fibrillation, atrial fibrillation burden of >10% for more than 5 consecutive days or persistent atrial fibrillation (defined as two consecutive days showing 100% of atrial fibrillation burden). An expert nurse received the alerts from the HM service as email in a dedicated box and as short message service (sms) through a commercial mobile phone. The nurse consulted the cardioreports on the website and submitted the critical cases to the responsible physician for clinical decision. Settings for trigger alerts could be modified after the first intervention according to clinical judgment.
Standard in-hospital follow-up
Standard in-hospital visit intervals were extended because of HM remote control. Implantable cardioverter defibrillator patients had scheduled follow-up every 6 months (instead of 3 months) whereas pacemaker patients were checked once a year (instead of every 6 months).
Normally distributed continuous variables were expressed as mean ± standard deviation, while median values with the inter-quartile ranges were reported for not-normally distributed variables. Ninety-five per cent confidence interval was calculated for rates and percentages. Survival rates were estimated by means of the Kaplan–Meier method. Paired comparisons of not normal variables were performed with the Wilcoxon sign test. Differences were considered significant when P is <0.05.
The follow-up lasted on average for 488 ± 203 days. Three patients were excluded from the analysis because of permanent atrial fibrillation at implant (one patient implanted with ICD) or because the implanted device (Lexos A+ single lead, Biotronik GmbH & Co. KG, Berlin, Germany) did not allow detailed monitoring of atrial fibrillation (two patients). One hundred and sixty-three patients had data suitable for the analysis.
Atrial fibrillation recurrences and burden
At HM data analysis, 109 patients (67%) had no episodes of atrial fibrillation during the follow-up, including 23 with a prior history of atrial fibrillation. Fifty-four patients (33%) had at least one episode of atrial fibrillation lasting more than 15 min. Among them, 22 had a prior history of atrial fibrillation. Arrhythmia duration was >2.5 h (daily burden >10%) in 42 (26%), >6 h (daily burden >25%) in 34 (21%), >12 h (daily burden >50%) in 25 (15%), and >24 h (daily burden 100%) in 18 (11%), as represented in Figure 1.
Kaplan–Meyer curves for survival from atrial fibrillation occurrence. Survival from episodes lasting more than 2.5, 6, 12, and 24 h are represented separately.
Home MonitoringTM web data analysis and clinical interventions
Forty-two patients (26%, 37 with pacemaker and 5 with ICD) had alerts for atrial fibrillation. Patient with atrial fibrillation alerts had on average 2.0 ± 1.2 atrial fibrillation detections (range from 1 to 6). Twenty-two patients (52%, 19 with pacemaker and 3 with ICD) had no history of atrial fibrillation before implant was done. Among the pacemaker patients, atrial fibrillation alerts were triggered in 22 of 67 with sinus node disease (33%), in 9 of 29 with atrio-ventricular block (31%), and in 6 of 25 with neuromediated syncope (24%). In nine patients, no further action was taken since the patient was judged to be already on optimal treatment. In detail, in six patients with alert triggered by an atrial fibrillation burden slightly above 10% for >5 consecutive days an extra follow-up was not scheduled since they were judged to be at low risk according to their clinical profile. In three patients, of whom one with recent isthmus ablation and two with a prior myocardial infarction, all in anticoagulation, a ‘wait and see’ decision was taken.
For the remaining 33 patients, an unscheduled follow-up was performed, in-hospital in 26 and by phone contact in 7. Five patients had both in-hospital and by phone unscheduled follow-ups. Among the 33 patients who had an unscheduled follow-up, 24 (73%) were asymptomatic, six (18%) were mildly symptomatic, and only three (9%) had severe symptoms, which require urgent clinical evaluation, and two cases require hospitalization because of progression of heart failure after atrial fibrillation recurrence. After the additional follow-up, the following reactions were established: in 16 patients (48%), antiarrhythmic drug therapy was introduced or modified; in 15 patients (45%), anticoagulation was started; among them, there was no history of atrial fibrillation in 6 patients and anticoagulation therapy was introduced only because of HM data; in 2 patients (6%), antiplatelet drugs were introduced; in 7 patients (21%), an external cardioversion was performed. One ICD patient experienced undersensing of 2:1 atrial flutter. The arrhythmia could be detected by the HM analysis since the data showed a sudden increase of the mean ventricular rate combined with a sudden increase of the atrial sensing percentage. The patient received an external cardioversion, and sinus rhythm was restored. In two cases (6%), the device was reprogrammed to avoid intermittent atrial undersensing during atrial fibrillation: in one of them the atrial sensitivity was reprogrammed from 0.5 to 0.2 mV with significant improvement of the arrhythmia detection. In the other one, the atrial sensitivity was not actually modified, since higher values of atrial sensitivity led to R-wave far-field oversensing. In four cases (12%), no further action was taken and the ongoing therapy was confirmed. In detail, in two patients, antiarrhythmic therapy and anticoagulation were confirmed and, in one patient at low risk with no structural heart disease and relatively low atrial fibrillation burden, no therapy was added. One patient with persistent atrial fibrillation did not accept anticoagulation and cardioversion. Finally, in four cases (12%), the arrhythmia was not confirmed due to detection of far-field R-wave oversensing (false positive). In all of them, blanking periods were programmed at nominal values according to the study protocol. Reprogramming the atrial sensitivity from 0.5 to 1.0 mV significantly reduced far-field oversensing in two patients as shown in the case reported in Figure 3. Atrial sensitivity reprogramming was unsuccessful in the other two patients. Among the seven patients who underwent the electrical cardioversion, three had atrial fibrillation recurrence and went into permanent atrial fibrillation, while in four cases stable sinus rhythm was maintained on average after 232 days. Figure 4 illustrates the case of a patient with asymptomatic persistent atrial fibrillation recurrence early detected by the HM analysis. After electrical cardioversion, the HM analysis allowed us to confirm long-term persistence of stable sinus rhythm. In 10 patients with paroxysmal atrial fibrillation in whom the antiarrhythmic drug therapy was either introduced or modified, atrial fibrillation burden was analysed and compared 10 days prior to the drug therapy change and 10 days after one month of therapy effect stabilization. After changing of antiarrhythmic drug therapy the atrial fibrillation burden decreased from 7.5% ± 9.2% to 0.4% ± 1.0% (P < 0.02). Figure 5 presents the case of a patient with recurrent paroxysmal atrial fibrillation. After antiarrhythmic therapy introduction, the HM analysis demonstrated a reduction in the number and duration of tachyarrhythmia episodes.
Case report of a patient with inappropriate detection of atrial fibrillation because of R-wave far-field oversensing. (A) The Home Monitoring™ graph of mode switch number and duration. After follow-up and atrial sensitivity reprogramming, mode switch duration decreased to 0% indicating no further inappropriate arrhythmia detection. (B) The internal electrogram retrieved from pacemaker memory at follow-up that documented inappropriate mode switching due to R-wave far-field oversensing.
Case report of a patient with asymptomatic persistent atrial fibrillation recurrence detected by the Home Monitoring™ analysis. The patient was called back to the hospital. An electrical cardioversion was performed and sinus rhythm was restored. The HM analysis allowed us to confirm long-term persistence of stable sinus rhythm.
Case report of a patient with recurrent paroxysmal atrial fibrillation. After an additional follow-up, antiarrhythmic drug therapy was introduced. Home Monitoring™ reports demonstrated a reduction in the number and duration of tachyarrhythmia episodes over time.
The median time from the last in-hospital follow-up visit to the detection and the first intervention for atrial fibrillation was 50 days (inter-quartile interval: 10–244 days). The median value of reaction time to atrial fibrillation in advance to the scheduled follow-up was 148 days (inter-quartile interval: 75–170 days).
Main findings of the study
In the present study, atrial fibrillation episodes lasting >2.5 h were detected in 26% of the enrolled pacemaker/ICD patients. Atrial fibrillation was completely asymptomatic in 73% of patients with device-detected episodes. Through remote monitoring, atrial fibrillation detection and notification by the physician occurred on average 148 days earlier than with scheduled follow-up visits. An early warning of atrial fibrillation occurrence as soon as the arrhythmia has been detected and stored in the device memory may induce prompt clinical reaction by the physician and may prevent arrhythmia-related severe adverse events caused by long-lasting undetected atrial fibrillation. That is particularly meaningful in asymptomatic patients.
Atrial fibrillation occurrence in paced patients and prognostic value
Atrial fibrillation is very common in patients with implanted pacemakers and ICD, even when there is no history of the arrhythmia before implant. It has been reported that in patients implanted with dual-chamber pacemakers, device diagnostics show atrial fibrillation episodes in up to 50% of patients already after a short-term follow-up, regardless of indication for pacing and that the majority of the episodes are asymptomatic.4,5 In patients candidate for ICD implant, it has been demonstrated that 20% had a prior episode of atrial fibrillation before implant and that during the life-span of the device at least 50% may experience atrial fibrillation.18
Atrial fibrillation is not a benign arrhythmia; a high burden of atrial fibrillation, regardless of symptoms, or arrhythmia episodes lasting >24 h have been identified as independent predictors for stroke and mortality.6–8
Reliability of the Home MonitoringTM technology
Reliability and safety of remote control of implanted devices using the HM technology have been definitely demonstrated by several studies.11,15,17,19–22 It has been reported that the clinical staff may be alerted within 3 min for more than 97% of messages. Clinical studies showed that, in clinical practice, >90% of days could be actually monitored and that data and electrogram transmission were reliable and comparable with data retrieved during a standard in-hospital follow-up. Patient compliance and acceptance were good in all the studies.
Home MonitoringTM technology in atrial fibrillation management
The main potential advantage of remote control application in atrial fibrillation management is represented by early detection and early reaction to the arrhythmia occurrence.12–14 The current device diagnostic is very sophisticated and may give to the physician full information about arrhythmia episodes:9,10 number and duration, date and time of occurrence, onset mechanism, arrhythmia burden, effects of antitachycardia therapies, and heart rate during the arrhythmia. Unfortunately, this information is available only during in-hospital follow-up, usually scheduled every 6 or 12 months, significantly delaying the reaction to atrial fibrillation. This represents a great limitation, mainly for asymptomatic patients and for those with mild symptoms.
In the present study, 42 patients (26%) had alerts for atrial fibrillation and in 33 (20%) an unscheduled follow-up was performed on average 148 days in advance to the scheduled follow-up. The majority of patients with alert (73%) were asymptomatic. The HM-guided unscheduled follow-ups lead to clinically significant reactions to atrial fibrillation such as antiarrhythmic drug therapy introduction or modification (48%), anticoagulation starting (45%), or external cardioversion performing (21%).
An open issue is how to select appropriate alerts for atrial fibrillation and how to apply international guidelines for atrial fibrillation treatment23 in patient remote management. It is not known, for instance, if treatment of asymptomatic episodes detected by the device should be the same applied for symptomatic episodes. Anyway, it has been demonstrated that the risk for stroke is independent from symptoms.6,7
The present study was not designed to evaluate if the HM-induced atrial fibrillation reaction actually improved overall patient outcome. In our series we did not actually have stroke events. Among the patients submitted to electrical cardioversion, 57% showed long-term persistence of stable sinus rhythm. In the patients with paroxysmal atrial fibrillation, in whom antiarrhythmic drug therapy was introduced or modified, the burden of atrial fibrillation significantly decreased after the intervention. To this regard, it has to be kept in mind the spontaneous variability of atrial fibrillation burden as well as its uncertain clinical value.24,25
To evaluate the impact of remote monitoring on hard clinical end points, large randomized trials with longer follow-up are warranted.
The main limitation of the present observational study without control group is the lack of clinical outcome criteria. The study was aimed at evaluating changes in clinical management of patients with device-detected atrial fibrillation through remote monitoring, but it has not been proved if early atrial fibrillation detection by the HM combined with earlier clinical reactions had any positive impact on patient clinical status.
Alert definitions for atrial fibrillation were arbitrary. They were defined at the beginning of the study, according to the present knowledge. Criteria for drug therapy modification or device reprogramming as well as indication for electrical cardioversion were defined according to standard guidelines, but in each case they were established after individual medical judgement.
Remote control of patients with implanted devices allows early detection of atrial fibrillation, significantly in advance if compared with standard in-hospital follow-up, and early reaction to arrhythmia occurrence. Changes in pharmacological or electrical therapy may prevent severe complications, mainly in asymptomatic patients. Controlled studies are needed to evaluate if such early reaction may improve patient's clinical outcome.
The authors warmly thank Dr Maria Teresa Laudadio, PhD, for her help in data analysis and manuscript review. They are grateful to Dr Alessio Gargaro, PhD, for his help in the statistical analysis.
Conflict of interest: None of the authors received grants or consultancy fees for this work. R.P.R. and M.S. previously received minor consultancy fees from Medtronic and St. Jude Medical.
. Prevalence of supraventricular arrhythmias from the automated analysis of data stored in the DDD pacemakers of 617 patients: the AIDA Multicenter Study Group. Automatic Interpretation for Diagnosis Assistance. Pacing Clin Electrophysiol 1998;21:250-5.
; for the Atrial High Rate Trial Investigators. Asymptomatic atrial fibrillation in pacemaker recipients: incidence, progression, and determinants based on the atrial high rate trial. Pacing Clin Electrophysiol 2007;30:404-11.
; for the MOST Investigators. Atrial high rate episodes detected by pacemaker diagnostics predict death and strokes. Report of the atrial diagnostics ancillary study of the Mode selection Trial (MOST). Circulation 2003;107:1614-9.
; on behalf of the Italian AT500 Registry. Monitored atrial fibrillation duration predicts arterial embolic events in patients suffering from bradycardia atrial fibrillation implanted with antitachycardia pacemakers. J Am Coll Cardiol 2005;46:1913-20.
; On behalf of the Home CARE Phase 0 Study Investigators. Predicting mortality and rehospitalization in heart failure patients with Home Monitoring: The Home CARE pilot study. Clin Res Cardiol 2006;95:iii29-iii35.
. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation: full text: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the European Society of Cardiology Committee for Practice Guidelines. Europace 2006;8:651-745.