TY - JOUR
AU1 - Ollitrault,, Pierre
AU2 - Jacon,, Peggy
AU3 - Auquier,, Nathanaël
AU4 - Champ-Rigot,, Laure
AU5 - Ben Kilani,, Mouna
AU6 - Vandevelde,, Florence
AU7 - Pellissier,, Arnaud
AU8 - Ferchaud,, Virginie
AU9 - Legallois,, Damien
AU1 - Defaye,, Pascal
AU1 - Anselme,, Frédéric
AU1 - Milliez,, Paul
AB - Abstract Aims No data exist concerning the clinical performances of the subcutaneous implantable cardioverter-defibrillator (S-ICD) atrial fibrillation (AF) detection algorithm. We aimed to study the performances and implications of the latter in a ‘real-world’ setting. Methods and results Between July 2017 and August 2019, 155 consecutive S-ICD recipients were included. Endpoint of the study was the incidence of de novo or recurrent AF using a combined on-site and remote-monitoring follow-up approach. After a mean follow-up of 13 ± 8 months, 2531 AF alerts were generated for 55 patients. A blinded analysis of the 1950 subcutaneous electrocardiograms available was performed. Among them 47% were true AF, 23% were premature atrial contractions or non-sustained AF, 29% were premature ventricular contractions or non-sustained ventricular tachycardia, and 1% were misdetection. Fourteen percent (21/155) patients had at least one correct diagnosis of AF by the S-ICD algorithm. One patient presented symptomatic paroxysmal AF not diagnosed by the S-ICD algorithm (false negative patient). Patient-based sensitivity, specificity, positive, and negative predictive values were respectively 95%, 74%, 38%, and 99%. Among patients with at least one correct diagnosis of AF, 38% (8/21) had subsequent clinical implications (anticoagulation initiation or rhythm control therapies). Conclusion The S-ICD AF detection algorithm yields a high sensitivity for AF diagnosis. Low specificity and positive predictive value contribute to a high remote monitoring-notification workload and underline the necessity of a manual analysis. Atrial fibrillation diagnosis by the S-ICD AF detection algorithm might lead to significant therapeutic adjustments. Atrial fibrillation, Subcutaneous defibrillator, Algorithm, Sensitivity What’s new? The subcutaneous implantable cardioverter-defibrillator (S-ICD) detection algorithm as a 95% patient-based sensitivity for atrial fibrillation diagnosis. Both low specificity and positive predictive value contributes to a high remote monitoring notification workload, with the necessity of a manual analysis of each alert. Correct diagnosis of atrial fibrillation by the S-ICD might have significant clinical implications for the patient. Introduction Atrial fibrillation (AF) is the most frequent cardiac arrhythmia worldwide, with an increased risk of stroke, systemic embolic events, and mortality. Early diagnosis, long-term anticoagulation, and invasive strategies such as AF catheter ablation decrease adverse outcomes. Among other cardiac implantable electronic devices (CIEDs), implantable cardioverter-defibrillators (ICDs) are validated diagnostic tools for AF detection, with additional value of remote monitoring.1–4 This is of clinical importance because ICD recipients are at high risk of AF and associated adverse outcomes.5,6 Cardiac implantable electronic devices with a transvenous atrial lead remain the gold standard for AF detection3,7 but, as the majority of transvenous ICDs are single chamber,8,9 various algorithms have been developed to allow detection of AF by other mean that rapid ventricular response.10 The latter are based on RR interval variability analysis, similarly to those integrated in insertable cardiac monitors (ICMs), with a minimum reported sensitivity of 80% depending on the algorithm and the characteristics of AF.7,11–14 Recently, the entirely subcutaneous ICD (S-ICD Emblem™, Boston Scientific, USA) has emerged as a safe and efficient alternative to the single-chamber ICD for sudden cardiac death prevention, in selected patients.15 As a consequence, there is a continuous growth in implantations over Europe.16 The latest S-ICD integrates a dedicated detection algorithm (AF Monitor™, Boston Scientific, USA) designed to assist for the diagnosis of AF with irregular ventricular response. However, there is currently no data concerning clinical performances and implications of the latter.17 Hence, the goal of the present study was to evaluate clinical performances and implications of the S-ICD AF detection algorithm, in a ‘real-world’ setting combining on-site and remote follow-up. Methods Inclusion Between July 2017 and December 2019, consecutive patients ≥18-year-old implanted with a last-generation S-ICD (A219 Emblem™) in four academic medical centres were included. A minimal follow-up of 3 months was required. The patient was excluded in case of complete atrioventricular block. Atrial fibrillation detection algorithm AF Monitor™ is a dedicated AF detection algorithm nominally enabled in every A219 Emblem™ S-ICD. Rhythm analysis is performed on 192-interval windows over each 24-h period. The algorithm integrates two components: ventricular scatter analysis (VSA) and heart rate density index analysis (HRDIA). Ventricular scatter analysis is based on beat-to-beat RR interval variability (unstable if ≥5 b.p.m.; stable if <5 b.p.m.; or random). Heart rate density index analysis is based on RR interval distribution and rate (narrow distribution and low rate; or wide distribution and high rate). An AF alert is generated if: (i) the 192-interval window meets both VSA (unstable or random) and HRDIA (wide distribution and high rate) criteria; (ii) the next window meets at least the HRDIA criterion; and (iii) the total duration of the episode reaches 6 min. A maximum of one AF alert per 24 h is stored in the S-ICD, with a 44-s subcutaneous electrocardiogram (sECG). A maximum of 250 AF alerts with sECG are stored in the device memory, and any additional episode overwrite the oldest sECG episode (the AF alert notification still remain in the device memory). Burden of AF for the last 100 days is estimated as the daily duration of AF (in h/day) and the overall number of days with at least one AF alert (in days). The S-ICD AF detection algorithm is independent from shock zones programming. Follow-up All patients underwent both remote and on-site follow-up. On-site follow-up was scheduled every 6–12 months (depending on local practice), including clinical assessment, 12-lead standard electrocardiogram (ECG), and other paraclinical investigations at the discretion of the electrophysiologist. In patients with both an S-ICD and a pacemaker or a cardiac resynchronization therapy, interrogation was performed searching for AF episodes. Remote monitoring (Latitude™ NXT, Boston Scientific, USA) was scheduled every 3 months, with additional on-demand and automatic transmissions if necessary. Episode-based metrics A blinded analysis of all stored AF alerts was performed by two senior electrophysiologists. Each alert was labelled as true AF, premature atrial contractions (PACs) or non-sustained AF (NSAF), premature ventricular contractions (PVCs) or non-sustained ventricular tachycardia (NSVT), misdetection or non-diagnostic. In the absence of consensus between the two electrophysiologists, previous episodes/alerts from the same patient could be used for diagnostic purpose. All stored S-ICD shock and diverted charge episodes were also reviewed, searching for inappropriate shocks related to AF, atrial flutter, and/or oversensing. Patient-based metrics Each patient was classified as: (i) true positive, if having at least one AF alert related to true AF; (ii) false positive, if having every AF alert unrelated to true AF (e.g. PVCs, PACs, etc.); (iii) false negative, if having a diagnosis of AF made by any other modality than the S-ICD AF detection algorithm (e.g. 12-lead ECG during on-site follow-up, symptom-triggered Holter monitoring, etc.); and (iv) true negative, otherwise. Patient-based sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the S-ICD AF detection algorithm were calculated. Endpoints The endpoint of the study was the incidence of de novo or recurrent AF (or atrial flutter). As a single correct diagnosis of AF is sufficient for risk stratification, the primary objective of the study was to evaluate patient-based metrics analysis, rather than episode-based metrics. The secondary objective of the study was to evaluate the clinical implications of the S-ICD AF detection algorithm in terms of anticoagulation initiation, rhythm control strategies, and inappropriate shock occurrence. Statistical analysis Continuous variables are presented as mean ± standard deviation or median (quartile 1–quartile 3) as appropriate, and categorical variables are given as number of subjects with the attribute (percentage). For continuous variables, a Student’s t-test or a Mann–Whitney U test was performed as appropriate. The χ2 test was used for the analysis of categorical variables. The incidence of de novo or recurrent AF during follow-up in S-ICD recipients is currently unknown. As the estimated minimum incidence of AF ≥6 min is 5% at 6 months in patients with a single-chamber ICD,5 we calculated a minimum sample size of 149 patients for a confidence interval of 95% and an absolute precision of 3.5%. Survival freedom from AF alert and AF diagnosis were calculated using the Kaplan–Meier method. A two-tailed P-value <0.05 denoted statistical significance. Analyses were conducted using IBM SPSS Statistics for Macintosh (Version 23.0, IBM, Chicago, IL, USA). Ethics Approval for this study was obtained from the local Ethics Committee and was in accordance with the declaration of Helsinki. This report was prepared according to the ‘Strengthening the Reporting of Observational Studies in Epidemiology’ statement.18 All subjects provided written informed consent for clinical and remote monitoring data analysis. Results Baseline characteristics Between July 2017 and December 2019, 168 patients implanted with an S-ICD were screened. Eight patients were excluded because of follow-up duration <3 months, three patients because of the absence of remote monitoring, and two because of complete atrioventricular block and permanent ventricular pacing. A hundred and fifty-five patients were finally included; their baseline characteristics at the time of S-ICD implantation are detailed in Table 1. Eight patients were in persistent atrial fibrillation and one patient in persistent atrial flutter at S-ICD implantation. Table 1 Baseline characteristics of the study population (n = 155) . n = 155 . Demographics  Age (years) 50 ± 15  Male gender 132 (85)  Height (cm) 172 ± 16  Weight (kg) 79 ± 18  Body surface area (m2) 1.93 ± 0.28 Medical history  Hypertension 43 (28)  Diabetes 16 (10)  Stroke/TIA 12 (7.7)  Congestive heart failure 106 (68)  LVEF (%) 39 ± 14  CHA2DS2-VASc 2 ± 1   0 23 (15)   1 46 (30)   2 38 (25)   3 22 (14)   4 19 (12)   ≥5 7 (4.5)  AF/atrial flutter 33 (21)  Pacemaker/CRT 2 (1.3) ICD implantation  Sinus rhythm at implantation 146 (94)  Primary prevention 98 (63)  Cardiomyopathy   Ischaemic/coronary artery disease 62 (40)   Dilated 40 (26)   Hypertrophic 11 (7.1)   Arrhythmogenic (genetic, myocarditis) 11 (7.1)   Valvular 2 (1.3)   Congenital 7 (4.5)  Primary inherited arrhythmia syndrome 11 (7.1)  Idiopathic VF 11 (7.1) Medications  Anticoagulant   Direct oral anticoagulant 26 (17)   Vitamin-K antagonist 25 (16)  Anti-arrhythmic   Class I 2 (1.3)   Class III 16 (10) . n = 155 . Demographics  Age (years) 50 ± 15  Male gender 132 (85)  Height (cm) 172 ± 16  Weight (kg) 79 ± 18  Body surface area (m2) 1.93 ± 0.28 Medical history  Hypertension 43 (28)  Diabetes 16 (10)  Stroke/TIA 12 (7.7)  Congestive heart failure 106 (68)  LVEF (%) 39 ± 14  CHA2DS2-VASc 2 ± 1   0 23 (15)   1 46 (30)   2 38 (25)   3 22 (14)   4 19 (12)   ≥5 7 (4.5)  AF/atrial flutter 33 (21)  Pacemaker/CRT 2 (1.3) ICD implantation  Sinus rhythm at implantation 146 (94)  Primary prevention 98 (63)  Cardiomyopathy   Ischaemic/coronary artery disease 62 (40)   Dilated 40 (26)   Hypertrophic 11 (7.1)   Arrhythmogenic (genetic, myocarditis) 11 (7.1)   Valvular 2 (1.3)   Congenital 7 (4.5)  Primary inherited arrhythmia syndrome 11 (7.1)  Idiopathic VF 11 (7.1) Medications  Anticoagulant   Direct oral anticoagulant 26 (17)   Vitamin-K antagonist 25 (16)  Anti-arrhythmic   Class I 2 (1.3)   Class III 16 (10) Data are expressed as mean ± standard deviation or number (percentage), as appropriate. AF, atrial fibrillation; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; TIA, transient ischaemic attack; VF, ventricular fibrillation. Open in new tab Table 1 Baseline characteristics of the study population (n = 155) . n = 155 . Demographics  Age (years) 50 ± 15  Male gender 132 (85)  Height (cm) 172 ± 16  Weight (kg) 79 ± 18  Body surface area (m2) 1.93 ± 0.28 Medical history  Hypertension 43 (28)  Diabetes 16 (10)  Stroke/TIA 12 (7.7)  Congestive heart failure 106 (68)  LVEF (%) 39 ± 14  CHA2DS2-VASc 2 ± 1   0 23 (15)   1 46 (30)   2 38 (25)   3 22 (14)   4 19 (12)   ≥5 7 (4.5)  AF/atrial flutter 33 (21)  Pacemaker/CRT 2 (1.3) ICD implantation  Sinus rhythm at implantation 146 (94)  Primary prevention 98 (63)  Cardiomyopathy   Ischaemic/coronary artery disease 62 (40)   Dilated 40 (26)   Hypertrophic 11 (7.1)   Arrhythmogenic (genetic, myocarditis) 11 (7.1)   Valvular 2 (1.3)   Congenital 7 (4.5)  Primary inherited arrhythmia syndrome 11 (7.1)  Idiopathic VF 11 (7.1) Medications  Anticoagulant   Direct oral anticoagulant 26 (17)   Vitamin-K antagonist 25 (16)  Anti-arrhythmic   Class I 2 (1.3)   Class III 16 (10) . n = 155 . Demographics  Age (years) 50 ± 15  Male gender 132 (85)  Height (cm) 172 ± 16  Weight (kg) 79 ± 18  Body surface area (m2) 1.93 ± 0.28 Medical history  Hypertension 43 (28)  Diabetes 16 (10)  Stroke/TIA 12 (7.7)  Congestive heart failure 106 (68)  LVEF (%) 39 ± 14  CHA2DS2-VASc 2 ± 1   0 23 (15)   1 46 (30)   2 38 (25)   3 22 (14)   4 19 (12)   ≥5 7 (4.5)  AF/atrial flutter 33 (21)  Pacemaker/CRT 2 (1.3) ICD implantation  Sinus rhythm at implantation 146 (94)  Primary prevention 98 (63)  Cardiomyopathy   Ischaemic/coronary artery disease 62 (40)   Dilated 40 (26)   Hypertrophic 11 (7.1)   Arrhythmogenic (genetic, myocarditis) 11 (7.1)   Valvular 2 (1.3)   Congenital 7 (4.5)  Primary inherited arrhythmia syndrome 11 (7.1)  Idiopathic VF 11 (7.1) Medications  Anticoagulant   Direct oral anticoagulant 26 (17)   Vitamin-K antagonist 25 (16)  Anti-arrhythmic   Class I 2 (1.3)   Class III 16 (10) Data are expressed as mean ± standard deviation or number (percentage), as appropriate. AF, atrial fibrillation; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; TIA, transient ischaemic attack; VF, ventricular fibrillation. Open in new tab Episode-based metrics After a mean follow-up of 13 ± 8 months, a total of 2531 AF alerts were generated, with 16 ± 64 AF alerts/patient (more than 250 AF alerts in three patients with persistent AF). Among those, 1950 AF alerts (77%) had a stored sECG and underwent blinded analysis (Figure 1). Nine hundred and twenty-two episodes were true AF (47%), 457 (23%) were PAC or NSAF, 559 (29%) were PVC or NSVT, 11 (0.6%) were misdetection, and 1 (0.05%) was non-diagnostic (data extraction error). Figure 1 Open in new tabDownload slide Blinded analysis of AF alerts and representative examples (2.5 mm/mV, 25 mm/s). Data are expressed as percentage. AF, atrial fibrillation; ICD, implantable cardioverter-defibrillator; FU, follow-up; NSAF, non-sustained AF; NSVT, non-sustained ventricular tachycardia. Figure 1 Open in new tabDownload slide Blinded analysis of AF alerts and representative examples (2.5 mm/mV, 25 mm/s). Data are expressed as percentage. AF, atrial fibrillation; ICD, implantable cardioverter-defibrillator; FU, follow-up; NSAF, non-sustained AF; NSVT, non-sustained ventricular tachycardia. Patient-based metrics Thirty-five percent (55/155) patients had at least one AF alert at the end of follow-up. Survival freedom from first AF alert is illustrated in Figure 2. Among them, 21 patients (21/155, 14%) had at least one true AF episode detected by the S-ICD. In addition, one patient was diagnosed AF without any AF alert by the S-ICD (false negative; diagnostic by symptoms-triggered Holter monitoring). Survival freedom from AF (in the 146 patients in sinus rhythm at implantation) is illustrated in Figure 2 (92 ± 2.7% at 6 months). Figure 2 Open in new tabDownload slide Kaplan–Meier survival freedom from AF alert, and de novo or recurrent AF diagnosis in patient in sinus rhythm at implantation. AF, atrial fibrillation. Figure 2 Open in new tabDownload slide Kaplan–Meier survival freedom from AF alert, and de novo or recurrent AF diagnosis in patient in sinus rhythm at implantation. AF, atrial fibrillation. The patient-based true positives, false positives, false negatives, true negatives, sensitivity, specificity, PPV, and NPV of the S-ICD AF detection algorithm are detailed in Table 2. Table 2 Patient-based metrics of the S-ICD AF detection algorithm in the study population (n = 155) . AF . No AF . . AF Alert True positives False positives PPV n = 21 n = 34 38% No AF Alert False negatives True negatives NPV n = 1 n = 99 99% Sensitivity Specificity 95% 74% . AF . No AF . . AF Alert True positives False positives PPV n = 21 n = 34 38% No AF Alert False negatives True negatives NPV n = 1 n = 99 99% Sensitivity Specificity 95% 74% Data are expressed as number of patients or percentage, as appropriate. AF, atrial fibrillation; NPV, negative predictive value; PPV, positive predictive value; S-ICD, subcutaneous implantable cardioverter-defibrillator. Open in new tab Table 2 Patient-based metrics of the S-ICD AF detection algorithm in the study population (n = 155) . AF . No AF . . AF Alert True positives False positives PPV n = 21 n = 34 38% No AF Alert False negatives True negatives NPV n = 1 n = 99 99% Sensitivity Specificity 95% 74% . AF . No AF . . AF Alert True positives False positives PPV n = 21 n = 34 38% No AF Alert False negatives True negatives NPV n = 1 n = 99 99% Sensitivity Specificity 95% 74% Data are expressed as number of patients or percentage, as appropriate. AF, atrial fibrillation; NPV, negative predictive value; PPV, positive predictive value; S-ICD, subcutaneous implantable cardioverter-defibrillator. Open in new tab Clinical implications Among patients with at least one correct diagnosis of AF by the S-ICD, 42% (9/21) had subsequent clinical implications. Five patients had de novo asymptomatic paroxysmal AF. Except for one patient, CHA2DS2-VASc score ≥2 led to the initiation of an anticoagulation therapy. No stroke or transient ischaemic attack occurred in the global population during follow-up. Four patients underwent electrical cardioversion and/or AF catheter ablation for recurrent symptomatic AF (an antiarrhythmic drug was initiated before AF catheter ablation in two of those patients). The overall number of AF alerts since implantation (75 ± 132 vs. 7 ± 39 AF alerts; P < 0.001), the daily duration of AF for the last 100 days (4.3 ± 7.3 vs. 0.1 ± 1.3 h/day; P < 0.001), and the number of days with at least one AF alert for the last 100 days (25 ± 35 vs. 2.1 ± 9.6 days; P < 0.001) were significantly higher in patients with true AF than in patients without any diagnosis of AF. Only the daily duration of AF for the last 100 days was significantly lower in patients with paroxysmal AF compared with patients with persistent/permanent AF (2.8 ± 5.3 vs. 6.4 ± 9.3 h/day; P = 0.02). Incidence of inappropriate shock related to AF or atrial flutter at the end of follow-up was 1.9% (3/155), two related to AF with rapid ventricular response and one related to atrial flutter waves oversensing. All inappropriate shocks related to AF or atrial flutter were preceded by at least 1 AF alert (1, 14, and 63 days before shock for each patient, respectively). The only false negative patient had symptomatic paroxysmal AF diagnosed on 24-h Holter monitoring and underwent AF catheter ablation after initiation of anticoagulation (CHA2DS2VASc = 2). In this patient, daily duration of paroxysmal AF on Holter monitoring was less than 6 min. Discussion Main findings The AF Monitor™ is a dedicated detection algorithm integrated in new-generation S-ICDs (A219 Emblem™), designed for early detection of AF. The main findings of this study can be summarized as follow: (i) more than a quarter of S-ICD recipients will have an AF alert at 6 months, with an incidence of de novo or recurrent AF of 8 ± 2.7%; (ii) the AF detection algorithm provides high patient-based sensitivity and NPV, at the expense of both low specificity and PPV; and (iii) the correct diagnosis of AF by the S-ICD has significant clinical implications in terms of therapeutic interventions and potential prevention of inappropriate shocks, for patients with or without a history of AF. Atrial fibrillation detection using cardiac implantable electronic devices Patients with an ICD are at high risk of AF and associated adverse outcomes.5,6 Atrial-based detection (either by pacemakers or ICD) is the gold standard for AF diagnosis, with nearly 100% sensitivity, specificity, PPV, and NPV.1,3,7 However, the implantation of an atrial lead for the specific purpose of AF detection is not recommended in the absence of pacing indication. Moreover, dual-chamber ICDs carry a higher risk of complications compared with single-chamber ICDs, without benefits on long-term outcomes.9 As a consequence, the ratio between single- and dual-chamber ICD implantations is approximately 70/30%.8 In this context, various algorithms have been developed to allow detection of AF from single-chamber ICDs, mostly based on intracardiac RR interval analysis.10,11 Those algorithms share common mechanisms with those integrated in ICMs, with the exception of a surface ECG-based rather than intracardiac electrogram-based detection. Altogether, previous studies found a sensitivity between 80% and 95% and a PPV between 70% and 80% for those algorithms, compared with continuous Holter monitoring.7,10–14 Performances of the subcutaneous implantable cardioverter-defibrillator atrial fibrillation detection algorithm Recently, the S-ICD has emerged has a safe and efficient alternative to the single-chamber ICD for sudden cardiac death prevention, in selected patients.15 As a consequence, the European countries S-ICD implantation rate has increased by a two-fold between 2015 and 2016, with a continuous growth.16 The AF detection algorithm of this device allows storage of a maximum of 250 AF alerts with a single-lead sECG of 44 s duration for each AF alert. Using a blinded analysis by senior electrophysiologists, we found a patient-based sensitivity of 95%, which is therefore similar to what has been previously reported in single-chamber ICDs and ICMs,7,10–14 and also similar to internal bench testing using a subset of data from the Physiobank public domain database.17 The only false negative in our cohort was a young patient with symptomatic paroxysmal AF, and the S-ICD algorithm might be underpowered by design to detect short and infrequent runs of AF, as well as it is for AF with regular ventricular response. Clinical implications of atrial fibrillation monitoring by the subcutaneous implantable cardioverter-defibrillator Atrial fibrillation, either symptomatic or not (i.e. subclinical), is a pandemic public health issue associated with adverse cardiovascular outcomes. Hopefully, a growing armamentarium has emerged in the past decades in order to effectively prevent stroke-related, heart failure-related, and other adverse outcomes.1–3 In our cohort, approximately half of the patient with a correct diagnosis of AF by the S-ICD had subsequent therapeutic adjustments, mostly anticoagulation initiation and AF catheter ablation. Atrial fibrillation catheter ablation has been shown to reduce mortality and heart failure hospitalizations in patients with an ICD and heart failure with reduced ejection fraction.19 However, the question of the benefit of anticoagulation in patients with subclinical AF (four patients in our study) detected by CIEDs remain unanswered.2,3 Inappropriate shocks related to oversensing or supraventricular tachycardia above the discrimination zone are one of the most frequent complications of the S-ICD.15 Even though it has not been fully validated in S-ICD patients, inappropriate ICD shocks carry a negative impact on prognosis and prevention of the latter has been shown to decrease the risk of adverse outcomes.20 In our cohort, the AF detection algorithm generated at least one AF alert (related to AF with rapid ventricular conduction, or oversensing of atrial flutter waves) before every inappropriate shock. Thoroughly monitoring AF alerts might therefore help anticipating and preventing inappropriate shocks related to atrial arrhythmias, although those data deserve to be validated in a larger sample of patients. Subcutaneous implantable cardioverter-defibrillator remote monitoring Remote monitoring of CIEDs allows early detection of AF, with a decrease of stroke- and AF-related adverse outcomes.4 However, the notification workload has been one of the limiting steps to the full-adoption of remote monitoring in clinical practice. The low specificity and PPV of the S-ICD AF detection algorithm underlines the necessity of a manual analysis of each AF alert to avoid diagnostic and treatment errors. As previously reported in single-chamber ICDs and ICMs,10,12,14 the majority of false positives in our cohort were due to PACs and PVCs. Automatically identifying and eliminating ectopic beats from the algorithm (e.g. morphology analysis for PVCs/NSVT, RR-plot analysis or P-wave detection for PACs13) might lead to fewer and more specific AF alerts, in order to increase both specificity and PPV. Increasing the AF alert duration threshold from 6 min to 24 h might also lower the false positive rate, knowing that patients with CIED-detected subclinical AF of less than 24 h duration might not benefit from long-term anticoagulation.2 Limitations This study has the following limitations. In the absence of a continuous monitoring method as a control for AF detection by the S-ICD, there might be a potential underestimation of false negative episodes/patients and therefore an overestimation of sensitivity. We believe this limitation has not a significant impact on the validity of our results, for several reasons. Firstly, the incidence of true AF in our cohort is slightly higher than previously reported in similar patients with a single-chamber ICD.5 Secondly, we found a similar patient-based sensitivity compared with recent data from Boersma et al.,17 using a public ECG database. Finally, as the gold standard for AF detection are CIEDs with a transvenous atrial lead, implanting such device would have been the undisputed control to evaluate episode- and patient-based metrics of the S-ICD algorithm. However, the S-ICD has been specifically designed to avoid transvenous access and its several important complications,15 so it would be highly questionable to implant simultaneously a transvenous device in addition to the S-ICD. Importantly, implanting an ICM as a control also would have led to a potential underestimation of false positive patients, as their own sensitivity for AF detection can be as low as 79%.7 Our study included a small subset of patients who were in atrial fibrillation or flutter at the time of S-ICD implantation (n = 9). This might have induced an overestimation of both the number of AF alerts and the number of patients with an AF diagnosis by the S-ICD. However, the number of AF alert was not different between patients in sinus rhythm at implantation and those who were not. We included a ‘real-world’ unselected population of consecutive S-ICD recipients for the specific purpose of evaluating the S-ICD performances and not de novo AF incidence. We believe therefore that the impact of such bias is limited. A dedicated study evaluating de novo AF incidence in S-ICD recipients is needed. Conclusion The S-ICD AF detection algorithm provides a high sensitivity for AF diagnosis, at the expense of both low specificity and PPV which can contribute to an increased remote monitoring notifications workload. In patients with AF correctly diagnosed by the S-ICD, clinical implications are substantial in terms of anticoagulation initiation, AF catheter ablation, and potential prevention of inappropriate shocks. Acknowledgements We gratefully acknowledge Ms Audrey Delestre for data collection. Conflict of interest: P.O. is a consultant for Abbott and Boston Scientific. P.J. is a consultant for Boston Scientific. L.C.R. is a consultant for Boston Scientific and MicroPort CRM. P.D. received research grant (through institution) and honoraria from Abbott, Boston Scientific, Medtronic, and MicroPort CRM. F.A. is a consultant for Boston Scientific, Medtronic and MicroPort CRM. P.M. received honoraria from Biotronik and Boston Scientific. Data availability The data underlying this article will be shared on reasonable request to the corresponding author. References 1 Mairesse GH , Moran P , Van Gelder IC , Elsner C , Rosenqvist M , Mant J et al. ; ESC Scientific Document Group. Screening for atrial fibrillation: an European Heart Rhythm association (EHRA) consensus document endorsed by Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS), and Sociedad Latinoamericana de Estimulacion Cardiaca y Electrofixiologia (SOLAECE) . Europace 2017 ; 19 : 1589 – 623 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Jones NR , Taylor CJ , Hobbs FDR , Bowman L , Casadei B. Screening for atrial fibrillation: a call for evidence . Eur Heart J 2020 ; 41 : 1075 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Gorenek B , Bax J , Boriani G , Chen SA , Dagres N , Glotzer TV et al. ; ESC Scientific Document Group. Device-detected subclinical atrial tachyarrhythmias: definition, implications and management. an European Heart Rhythm association (EHRA) consensus document endorsed by Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS), and Sociedad Latinoamericana de Estimulacion Cardiaca y Electrofixiologia (SOLAECE ). Europace 2017 ; 19 : 1556 – 78 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Mabo P , Victor F , Bazin P , Ahres S , Babuty D , Da Costa A et al. A randomized trial of long-term remote monitoring of pacemaker recipients (The COMPAS Trial) . Eur Heart J 2012 ; 33 : 1105 – 11 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Schloss EJ , Auricchio A , Kurita T , Meijer A , Roger T , Tejima T et al. How common is new onset atrial fibrillation in single chamber ICD patients? Sub-analysis from the PainFree SST Study . Circulation 2015 ; 132 : A17946 . Google Scholar OpenURL Placeholder Text WorldCat 6 Biffi M , Ziacchi M , Ricci RP , Facchin D , Morani G , Landolina M et al. Can we predict new AF occurrence in single-chamber ICD patients? Insights from an observational investigation . Int J Cardiol 2017 ; 230 : 275 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Podd SJ , Sugihara C , Furniss SS , Sulke N. Are implantable cardiac monitors the “gold standard” for atrial fibrillation detection? A prospective randomized trial comparing atrial fibrillation monitoring using implantable cardiac monitors and DDDRP permanent pacemakers in post atrial fibrillation ablation patients . Europace 2016 ; 18 : 1000 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Bogossian H , Frommeyer G , Hochadel M , Ince H , Spitzer SG , Eckardt L et al. Single chamber implantable cardioverter defibrillator compared to dual chamber implantable cardioverter defibrillator: less is more! Data from the German Device Registry . Clin Res Cardiol 2019 ;doi:10.1007/s00392-019-01584-x. Google Scholar OpenURL Placeholder Text WorldCat 9 Defaye P , Boveda S , Klug D , Beganton F , Piot O , Narayanan K et al. ; on Behalf of the DAI-PP Investigators. Dual- vs. single-chamber defibrillators for primary prevention of sudden cardiac death: long-term follow-up of the Défibrillateur Automatique Implantable-Prévention Primaire registry . Europace 2017 ; 19 : 1478 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Deshmukh A , Brown ML , Higgins E , Schousek B , Abeyratne A , Rovaris G et al. Performance of atrial fibrillation detection in a new single-chamber ICD . Pacing Clin Electrophysiol 2016 ; 39 : 1031 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Sarkar S , Ritscher D , Mehra R. A detector for a chronic implantable atrial tachyarrhythmia monitor . IEEE Trans Biomed Eng 2008 ; 55 : 1219 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Hindricks G , Pokushalov E , Urban L , Taborsky M , Kuck KH , Lebedev D et al. Performance of a new leadless implantable cardiac monitor in detecting and quantifying atrial fibrillation. Results of the XPECT Trial . Circ Arrhythm Electrophysiol 2010 ; 3 : 141 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Pürerfellner H , Pokushalov E , Sarkar S , Koehler J , Zhou R , Urban L et al. P-wave evidence as a method for improving algorithm to detect atrial fibrillation in insertable cardiac monitors . Heart Rhythm 2014 ; 11 : 1575 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Ciconte G , Saviano M , Giannelli L , Calovic Z , Baldi M , Ciaccio C et al. Atrial fibrillation detection using a novel three-vector cardiac implantable monitor: the atrial fibrillation detect study . Europace 2017 ; 19 : 1101 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Boersma L , Barr C , Knops R , Theuns D , Eckardt L , Neuzil P et al. Implant and midterm outcomes of the subcutaneous implantable cardioverter-defibrillator registry. The EFFORTLESS Study . J Am Coll Cardiol 2017 ; 70 : 830 – 41 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Raatikainen MJP , Arnar DO , Merkely B , Nielsen J , Hindricks G , Heidbuchel H et al. A decade of information on the use of cardiac implantable electronic devices and interventional electrophysiological procedures in the European Society of Cardiology countries: 2017 report from the European Heart Rhythm Association . Europace 2017 ; 19 : ii1 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Boersma LVA , Mahajan D , Jones PW , Mittal S. AB05-02 Performance of a novel atrial fibrillation detection algorithm for use in patients with a subcutaneous implantable cardioverter defibrillator . Heart Rhythm 2016 ; 13 : S11 . Google Scholar Crossref Search ADS WorldCat 18 Vandenbroucke JP , von Elm E , Altman DG , Gotzsche PC , Mulrow CD , Pocock SJ et al. Strenghtening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration . Ann Intern Med 2007 ; 147 : W163 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Marrouche N , Brachmann J , Andresen D , Siebels J , Boersma L , Jordaens L et al. Catheter ablation for atrial fibrillation with heart failure . N Engl J Med 2018 ; 378 : 417 – 27 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Moss AJ , Schuger C , Beck CA , Brown MW , Cannom DS , Daubert JP et al. Reduction in inappropriate therapy and mortality through ICD programming . N Engl J Med 2012 ; 367 : 2275 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2020. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Atrial fibrillation detection by the subcutaneous defibrillator: real-world clinical performances and implications from a multicentre study
JF - Europace
DO - 10.1093/europace/euaa184
DA - 2020-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/atrial-fibrillation-detection-by-the-subcutaneous-defibrillator-real-Yes0cRTaKL
SP - 1628
EP - 1634
VL - 22
IS - 11
DP - DeepDyve
ER -