If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Insertable cardiac monitors (ICMs) are increasingly used for cardiac rhythm diagnosis with expanding indications. Little has been reported about their use and efficacy.
To evaluate the clinical utility of a novel ICM (Biotronik Biomonitor III) including the time to diagnosis in unselected patients with different ICM indications.
Patients from two prospective clinical studies were included to determine the diagnostic yield of the ICM. The primary endpoint was time to clinical diagnosis per implant indication or to the first change in atrial fibrillation (AF) therapy.
A total of 632 patients were included with a mean follow-up of 233±168 days. Of 384 patients with (pre-)syncope, 34.2% had a diagnosis at 1 year. Most frequent therapy was permanent pacemaker implantation. Of 133 patients with cryptogenic stroke, 16.6% had an AF diagnosis at 1 year, resulting in oral anticoagulation. Of 49 patients with an indication for AF monitoring, 41.0% had a relevant change in AF therapy based on ICM data at 1 year. Of 66 patients with other indications, 35.4% received a rhythm diagnosis at 1 year. Moreover, 6.5% of the cohort had additional diagnoses: 26 of 384 patients with syncope, 8 of 133 patients with cryptogenic stroke and 7 of 49 patients with AF monitoring.
In a large patient population with heterogeneous ICM indications, primary endpoint of rhythm diagnosis was achieved in ≈ 1 in 4 and additional clinically relevant findings in 6.5% of patients at short-term follow-up.
The use of insertable cardiac monitors (ICMs) is increasing due to expanding clinical indications from unexplained syncope to detection of subclinical atrial fibrillation (AF) in cryptogenic stroke, management of AF and clarification of suspected symptomatic arrhythmias. The available literature on ICMs is focused on specific clinical settings or according to implant indications but the diagnostic yield of ICMs in populations with heterogenous implant indications is not well reported.
Here, we evaluate the clinical efficacy of this ICM regarding the time to diagnosis in 632 unselected patients with different ICM indications and compare the diagnostic yield for atypical indications to that for the most common indication of unexplained syncope.
The present analysis includes patients with a BIOMONITOR III device inserted as part of the completed BIO|MASTER.BIOMONITOR III study (NCT04025710) or the ongoing BIO|STREAM.ICM registry (NCT04075084). Both studies were initiated to satisfy post-market clinical follow-up requirements for regulatory purposes and their enrollment criteria were minimal, including any clinical indication for an ICM implant, age ≥18 years old, and patient acceptance of remote monitoring. The subject must not be pregnant or be participant of another clinical study. To be included in the present analysis, patients had to have signed informed consent, and to have the insertion of the device plus any follow-up data reported by May 6, 2022. The research is conducted according to the principles of the Declaration of Helsinki with Institutional Ethics Committee approval at each site.
The main characteristics of BIOMONITOR III ICM are a large sensing vector (≈70 mm, composed of 47.5-mm long rigid component and a 30.5-mm long ﬂexible “antenna” adaptable to the curvature of the body), a miniaturized ICM profile (8.3 mm x 4.3 mm, weight 5 g), and a dedicated tool for a simplified injection-like insertion procedure with pocket formation and ICM placement in one step.
A 1-minute ECG strip is recorded upon arrhythmia detection, and 7.5-minute ECG recordings can be triggered by the patient. The ICM uses the established Biotronik Home Monitoring system to automatically transmit up to six ECG strips, arrhythmia detection statistics, and sensing performance parameters once a day via wireless links without active patient participation.
In both BIO|MASTER.BIOMONITOR III (“Master”) and BIO|STREAM.ICM (“Stream”) studies, the activation of the Home Monitoring feature was mandatory. The “Master” study enrolled patients between October 2019 and February 2021, with follow-up visits planned for the first month and at 3 and 12 months. The last patient completed the study on April 6, 2022 and the current publication was initiated when the data of the “Master” study were complete. The “Stream” study has been enrolling patients since October 2019 and will continue for a longer period. It collects long-term real-life data without mandatory study procedures and follow-ups; a patient completes the study when the investigator decides to explant the ICM or not to use ICM information any further.
For the present analysis, ICM indications were classified as (i) syncope or presyncope (hereinafter referred to as “syncope”), (ii) screening for AF after cryptogenic stroke, (iii) monitoring and management of AF, and (iv) others. Investigators reported when the respective diagnosis had been made based on a detection of the ICM; for example, pause, bradycardia, or tachyarrhythmia detection in syncopal patients, or AF detection followed by oral anticoagulation therapy prescription in cryptogenic stroke. Specifically, the classification or diagnostic criteria for all arrhythmias and pauses were left to the individual investigator’s discretion. In patients with other indications, a diagnosis was individually compared with the indication, for example, AF detection in patients with palpitations. The primary endpoint of the present analyses was the time to the clinical diagnosis based on ICM information. Only in patients with an ICM for AF management, the primary endpoint was the first relevant change in AF therapy including change of anti-arrhythmic medication, direct current cardioversion or catheter ablation procedures.
Data were used as entered until May 6, 2022. For the time to event analyses, patients were censored at the last study contact (e.g., in-office or remote follow-up, diagnosis, adverse event, study termination). Other results are presented with standard summary statistics (mean value ± standard deviation, median value and interquartile range according to data distribution) and as absolute and relative frequencies. In the time-to-event analyses, 6-month and 1-year probability estimates according to Kaplan-Meier are provided with 95% confidence intervals. The analyses were performed using the R (R Development Core Team, https://www.R-project.org/) statistical software.
This study included 632 patients (166 “Master”, 466 “Stream”) from 55 investigational sites in 11 countries, predominantly Germany (25.8% of patients), France (17.0%), Spain (8.3%), and Australia (7.3%). The patients were 64 ± 16 years old and 42.4% were females (Table 1). The ICM indications were syncope (60.8% of patients), cryptogenic stroke (21.0%), management of AF (7.8%), and others (10.4%). The indications summarized as “others” (n=66) included palpitations (n=20), clarification of ventricular (VT, n=12) or supraventricular (SVT, n=17) tachyarrhythmia, search for AF after peripheral embolism (n=5), possible arrhythmias (n=3), clarification of bradycardia or conduction system disturbances (n=2), and miscellaneous (n=7). Patient characteristics differed considerably across the four indication groups with those with ICM implanted for AF monitoring having more cardiovascular risk factors such as hypertension, heart failure, diabetes, valvular heart disease and coronary artery disease.
Table 1Baseline characteristics of included patients according to ICM indication.
All patients (N = 632)
(Pre)syncope (N = 384)
Crypt. stroke (N = 133)
AF monitor. (N = 49)
Other (N = 66)
Age – years
63.9 ± 16.3
65.1 ± 17.3
62.3 ± 14.3
64.6 ± 12.1
59.3 ± 16.4
BMI – kg/m2
27.6 ± 5.4
27.2 ± 5.2
27.5 ± 4.9
29.7 ± 6.2
28.6 ± 6.4
Stroke or TIA
History of arrhythmia
Sick sinus disease
Any cond. disturb.
Data are mean ± standard deviation, N (% of available data), or N (for NYHA classes).
AF, atrial fibrillation; arrhythm., arrhythmia; BMI, body mass index; CAD, coronary artery disease; cond., conduction; COPD, chronic obstructive pulmonary disease; Crypt., cryptogenic; disturb.; disturbance; ICM, insertable cardiac monitoring; monitor., monitoring and management of known AF; NYHA, New York Heart Association; pulm., pulmonary; SVT, supraventricular tachyarrhythmia; TIA, transient ischemic attack; VT, ventricular tachyarrhythmia.
At the time of data freeze, the 632 patients had a mean follow-up of 233 ± 168 days (median 220, interquartile range 76-365). Of all patients, 583 had an indication other than AF monitoring and 135 (23.2%) of them received an ICM-based diagnosis for the clinical question for which the device was implanted. Table 2 shows the number of patients for each indication and the diagnostic yield.
Table 2Diagnostic yield in 632 patients with follow-up
Pts with diagnosis† or change in therapy‡
Cryptogenic stroke (n=133)
AF monitoring (n=49)
†Diagnosis in (pre)syncope, cryptogenic stroke, and “other” indications.
‡Change in therapy in the AF monitoring indication.
§Kaplan-Meier estimates of the proportion of patients with a diagnosis with 95% CI.
More specifically, of 384 patients with syncope indication, 98 (25.5%) received a diagnosis based on the ICM. The ICM-detected arrhythmias were pause (n=69, 70.4% of cases), bradycardia (n=18, 18.4%), SVT other than AF (n=5, 5.2%), VT (n=3, 3.1%), AF (n=1, 1.0%), and other or not specified (n=2, 2.0%). As a result, reported medical interventions included permanent pacemaker implantation (n=71), medication change (n=12), catheter ablation (n=3), and cardioverter-defibrillator implantation (n=3). The Kaplan-Meier estimates of the proportion of patients with a diagnosis were 22.1% at 6 months and 34.2% at 1 year (Table 2, Figure 1). In seven patients, a syncope was reported without documentation of an arrhythmia by the ICM. Notably, 26 of these patients (6.8% of 383) had AF detected and received treatment accordingly.
Of 133 patients with cryptogenic stroke, 18 (13.5%) had an AF diagnosis based on the ICM. In all cases, oral anticoagulation was subsequently prescribed, and two patients also underwent catheter ablation for AF. The Kaplan-Meier estimates of the diagnostic yield were 12.6% at 6 months and 16.6% at 1 year (Table 2). In addition, other non-AF diagnoses from the ICM in eight out of 133 patients (6.0%) with cryptogenic stroke resulted in the following interventions: (i) closure of patent foramen ovale following exclusion of AF after 1 year of ICM monitoring in two patients; (ii) permanent pacemaker implantation following detection of a significant pause in four patients; (iii) coronary artery bypass grafting in one patient following detection of VT and subsequent coronary angiography; and (iv) detection of a regular SVT with ventricular rate of 200 beats/min in one patient.
Of 49 patients with an indication for AF monitoring, 19 (38.8%) had a relevant change in AF therapy based on ICM data. Ten patients underwent 11 catheter ablation procedures for AF (n=7), atrial flutter (n=2), atrioventricular nodal re-entry tachycardia (n=1), and frequent premature ventricular complexes (n=1). Six patients underwent AF cardioversions, and one patient was started on a new antiarrhythmic medication. The Kaplan-Meier estimates of the diagnostic yield were 29.2% at 6-month and 41.0% at and 1-year (Table 2), respectively. Further ICM-based diagnoses in seven (14.2%) out of 49 patients with AF monitoring indication resulted in implantation of a permanent pacemaker (n=6) or a cardioverter-defibrillator (n=1).
Of 66 patients with other indications, 19 (28.8%) received a diagnosis based on the ICM. The diagnoses were AF (n=6), SVT (n=4), pause (n=3), VT (n=2), bradycardia (n=1), sinus tachycardia (n=1), ventricular extrasystoles (n=1) and absence of VT recurrence following ablation (n=1). The Kaplan-Meier estimates of the diagnostic yield were 27.4% at 6 months and 35.4% at 1 year. Concerning different types of indication, six of 26 patients with palpitations had heterogeneous arrhythmias diagnosed (2 AF, 1 SVT, 1 VT, 1 pause, 1 VES) and three of seven patients after peripheral embolism had an arrhythmia diagnosis (2 AF, 1 SVT).
In our study of 632 patients with heterogeneous ICM implant indications over an average of 233 days of follow-up, we obtained the following principal findings. First, the primary endpoint achieved per implant indication was ≈ 1 in 4 with 25.6% for syncope, 13.5% for AF in cryptogenic stroke, 38.8% for relevant change in AF management based on AF monitoring and 28.8% for the remaining indications. Second, additional ’incidental’ ICM diagnoses were achieved in 6.5% of the cohort: 26 of 384 patients (6.8%) with syncope, eight of 133 patients with cryptogenic stroke (6.0%) and seven of 49 patients (14.3%) with AF monitoring. Third, the overall diagnostic yield of ICM per implant indication ranged from 16.6 to 41.0% at 12 months. Taken together, cardiac rhythm monitoring with ICM appears to contribute to successful diagnosis and resultant management changes that are likely to improve patient outcomes.
The most established indication for ICM use is syncope whereby prolonged monitoring is more likely to yield an arrhythmic diagnosis than intermittent monitoring. In our population, most patients (≈61%) received the device for the clarification of the cause of a syncope or presyncope. In a large meta-analysis of 41 studies of adults who underwent ICM implantation for unexplained syncope, the proportion of subjects ﬁnally diagnosed with arrhythmic syncope was 26.5%, ranging from 5.4% to 55.6%.
Our results confirm good diagnostic yield for this ICM indication, with 22.0% of patients having an arrhythmic diagnosis at 6 months, and 34.2% (95% CI, 27.8%-40.0%) at 1 year.
The second largest group were patients with cryptogenic stroke (≈21%). AF detection in our cohort was 12.6% at 6 months and 16.6% (95% CI, 10.6%–25.4%) at 1 year. These numbers are slightly higher than those of the landmark CRYSTAL AF study (8.9% at 6 months, 12.4% at 1 year),
In our study, all patients had oral anticoagulation prescribed after AF diagnosis, to prevent further thromboembolic event.
A relatively small group of patients (≈8%) received the device for the monitoring of known AF as this indication is not reimbursed in some of the participating countries. This is a relatively new application of ICM, typically in patients after AF ablation, in whom clarification of AF burden can influence disease managment.
Nearly one in five patients underwent a repeat ablation during the relatively short follow-up period in our study, suggesting that the ICM can have a relevant influence on the management of these patients. Interestingly, there was a relatively high yield of cardiac conduction abnormalities or ventricular arrhythmia in this cohort resulting in implantation of cardiac implantable electronic devices. This is not surprising as these conditions often share common risk factors that lead to atrial remodeling.
However, the most interesting results in our study may be those pertaining to the remaining patients. The reason why these patients received their device were symptoms other than syncope that could be a consequence of unknown arrhythmias, or a physician’s interest to describe more precisely known or assumed arrhythmias (to improve diagnosis). In a few cases, the proof of AF was attempted in patients that did not have a cryptogenic stroke but peripheral embolism. What patients in this heterogeneous indication group had in common was that their symptomatic burden and clinical risks were judged by physicians to be high enough to justify the insertion of an ICM. The ICM contributed to diagnosis in 27.4% of these patients at 6 months and 35.4% at 1 year, which is even numerically slightly higher than the diagnostic yield in patients with syncope.
Importantly, a considerable number of patients in our study received an unexpected diagnosis. The ICM provides added benefits due to the embedded algorithms to detect various rhythm related abnormalities. This result affirms a role for not disabling the in-built algorithms beyond the main indication for the ICM implantation, although this decision may result in an increased ECG classification workload with likely additional device alerts.
There are concerns that an invasive procedure should not be easily justified for a purely diagnostic device. The timescale of the time to diagnosis in our cohort makes clear that extracorporeal devices cannot be worn long enough to yield the same diagnoses. The low complication rate of approximately one serious adverse device effect in a hundred patients reported from a previous analysis of our population (which were mostly device extrusions due to insufficient wound closure) is however reassuring
and should be weighed against a diagnosis rate of approximately one in four patients, as demonstrated in this analysis.
Some limitations must be noted. First, we combined data from two studies with a different design. Further, some patients contribute only a relatively short follow-up duration as the “Stream” study is ongoing. Nevertheless, the large number of patients is a strength, and the knowledge of the diagnostic yield and resultant outcome of the patients are clinically relevant.
This study confirms the clinical utility of ICMs in a large population of patients with heterogeneous indications with ≈ 1 in 4 resulting in relevant diagnosis per implant indication and additional 6.5% achieving additional clinically relevant findings at short-term follow-up.
The authors would like to thank Anja Viehrig and Sabrina Hoche for study management, Katharina Ingel and Irina Köhler for statistical evaluations, Beate Wenzel for scientific support and Dejan Danilović for editorial assistance.
The role of implantable cardiac monitors in atrial fibrillation management.
On behalf of the BIO|MASTER.BIOMONITOR III study and BIO|STREAM.ICM registry investigators
Conflict of interest statement
DL declares support by Mid-Career Fellowships from The Hospital Research Foundation and reports that the University of Adelaide has received on his behalf lecture and/or consulting fees from Abbott Medical, Boehringer Ingelheim, Biotronik, Medtronic, MicroPort CRM, and Pfizer. BP declares being consultant for Abbott, Biotronik, Boston Scientific, Microport. PC declares having no conflicts of interest. EM reports speaker fee/consulting fee from Deutsche Gesellschaft für Kardiologie, Medtronic, Astra Zeneca, Bristol-Myers Squibb, Philips/Spectranetics, and Biotronik. GB declares having no conflicts of interest. DH reports educational grants, consultant or speaker fees, and fellowship support from European Heart Rhythm Organization (EHRA), Abbott, Medtronic, Biotronik, Boston Scientific, Biosense Webster, Novartis, Bayer, Pfizer, and Spectranetics/Philipps. AB, RE, JM, TG, VMS and AH declare having no conflicts of interest. JS is an employee of Biotronik. TD reports speaker fee from Biotronik (educational grant). All authors received patient fees for the reporting of study-related data, which ere paid to their institutions.
Clinical trial registration numbers: NCT04025710 and NCT04075084.
The study was supported by Biotronik SE & Co. KG, Berlin, Germany. The sponsor assisted in study design, data collection/analysis/interpretation, and preparation of this report. The authors had full access to the data and affirm the accuracy of this work.
⁃The diagnostic yield of an insertable cardiac monitor in a large population with heterogenous indications was ∼ 1 in 4 within 8 months.
⁃Additional incidental diagnoses were achieved from the insertable cardiac monitor in 6.5% of the patients.
⁃The overall diagnostic yield of an insertable cardiac monitor ranged from 16.5 to 41% at 12 months.