Summary
Background
Remote monitoring (RM) is increasingly used to follow up patients with implantable cardioverter-defibrillators (ICDs). Randomized control trials provide evidence for the benefit of this intervention, but data for RM in daily clinical practice with multiple-brands and unselected patients is lacking.
Aims
To assess the effect of RM on patient management and clinical outcome for recipients of ICDs in daily practice.
Methods
We reviewed ICD recipients followed up at our institution in 2009 with RM or with traditional hospital only (HO) follow-up. We looked at the effect of RM on the number of scheduled ambulatory follow-ups and urgent unscheduled consultations, the time between onset of asymptomatic events to clinical intervention and the clinical effectiveness of all consultations. We also evaluated the proportion of RM notifications representing clinically relevant situations.
Results
We included 355 patients retrospectively (RM: n = 144, HO: n = 211, 76.9% male, 60.3 ± 15.2 years old, 50.1% with ICDs for primary prevention and mean left ventricular ejection fraction 35.5 ± 14.5%). Average follow-up was 13.5 months. The RM group required less scheduled ambulatory follow-up consultations (1.8 vs. 2.1/patient/year; P < 0.0001) and a far lower median time between the onset of asymptomatic events and clinical intervention (7 vs. 76 days; P = 0.016). Of the 784 scheduled ambulatory follow-up consultations carried out, only 152 (19.4%) resulted in therapeutic intervention or ICD reprogramming. We also found that the vast majority of RM notifications (61.9%) were of no clinical relevance.
Conclusion
RM allows early management of asymptomatic events and a reduction in scheduled ambulatory follow-up consultations in daily clinical practice, without compromising safety, endorsing RM as the new standard of care for ICD recipients.
Résumé
Contexte
La télésurveillance des défibrillateurs automatiques implantables (DAI) se généralise. Les études randomisées ont démontré son utilité, mais les données manquent sur son utilisation en pratique courante avec différentes marques et des patients non sélectionnés.
Objectifs
Évaluer l’apport de la télésurveillance pour le suivi et l’évolution des patients appareillés d’un DAI en pratique courante.
Méthodes
Nous avons inclus rétrospectivement les patients appareillés d’un DAI dont la télésurveillance a débuté en 2009 dans notre centre, suivis par télésurveillance ou suivi hospitalier (SH) seul. Nous avons étudié l’apport de la télésurveillance sur le nombre de consultations programmées et en urgence, le délai entre la survenue d’un événement asymptomatique et l’intervention clinique ainsi que la pertinence clinique des consultations programmées. Nous avons également évalué la proportion d’alertes de télésurveillance traduisant un événement clinique pertinent.
Résultats
Au total, 355 patients ont été inclus (télésurveillance : n = 144 ; SH : n = 211 ; 76,9 % d’hommes ; 60,3 ± 15,2 ans ; 50,1 % de DAI en prophylaxie ; fraction d’éjection ventriculaire gauche moyenne 35,5 ± 14,5 %). La durée moyenne de suivi était de 13,5 mois. Dans le groupe télésurveillance, le nombre de consultations programmées était significativement réduit (1,8 vs 2,1/patient/année ; p < 0,0001) ainsi que le délai médian entre un événement asymptomatique et l’intervention clinique (7 vs 76 jours ; p = 0,016). Parmi les 784 consultations programmées, 152 (19,4 %) seulement ont été suivies d’un changement de thérapeutique ou de programmation du DAI. La majorité des notifications de télésurveillance (61,9 %) n’avait pas de pertinence clinique.
Conclusion
La télésurveillance permet une prise en charge précoce des événements asymptomatiques et une diminution des consultations programmées, en pratique courante, sans une moindre sécurité des patients, soutenant son utilisation large pour les porteurs de DAI.
Background
Remote monitoring (RM) is increasingly used to follow-up patients with implantable cardioverter-defibrillators (ICDs) . Large clinical trials have already demonstrated the safety and efficacy of this approach, however, most have focused on specific ICD brands and results might not be applicable to other brands. Experiences during randomized studies may differ from daily practice and accounts of clinical experience of RM outside randomized trial protocols are lacking. There are also limited data on the effectiveness of RM in the management of asymptomatic events and the proportion of ICD notifications that represent clinically relevant events is relatively uncharacterized.
Aims
The present study aimed to assess the benefits of RM in daily clinical practice, for all RM systems. We sought to investigate the impact of RM on the number of scheduled ambulatory follow-up consultations as well as the number of urgent unscheduled consultations. We also looked at the effectiveness of RM in the management of asymptomatic events and analysed data records to evaluate the proportion of RM notifications that are clinically relevant.
Methods
Study design
We performed a retrospective study comparing RM versus traditional hospital only (HO) follow-up for ICD recipients. All patients implanted with an ICD in the year 2009 in our institution were included in the study, regardless their follow-up: RM or HO. In addition, patients who had previously been implanted with an ICD that was subsequently activated for RM in 2009 were also included. The study population was then divided into two groups, based on the mode of follow-up: RM or HO. For the RM group, follow-up started at the time of RM activation and for the HO group, at the time of ICD implantation. Follow-up ended in June 2011. Exclusion criteria included: follow-up at another centre, death before first outpatient consultation and ICD removal before first outpatient consultation. All brands and types of ICD were included, except in the RM group where no patients had Sorin ® ICDs manufactured by the Sorin Group (Milan, Italy) because the Smartview™ RM system was not available at the time of study.
In the RM group, patient ICDs were able to transmit two types of RM notifications: scheduled reports and RM alerts. Scheduled reports were transmitted regularly for Latitude™ (Boston Scientific ® , Marlborough, MA, USA), Carelink™ (Medtronic ® , Minneapolis, MN, USA) and Merlin™ (Saint Jude Medical ® , St. Paul, MN, USA) recipients and daily, as per manufacturer protocol, for Home Monitoring™ (Biotronik ® , Berlin, Germany) recipients. It was then at the discretion of the clinician to invite patients for ambulatory follow-up consultations based on the data in these reports, with consultations typically happening once or twice a year and the first consultation occurring usually three months after implantation. RM alerts, on the other hand, were transmitted in response to abnormal events detected by the ICD (typically for unusual cardiac activity or ICD dysfunction) and these would always trigger urgent clinical review, whether or not the patient was symptomatic at the time of the event. It was left at the discretion of the clinician to define thresholds for alert transmission.
For patients in the HO group, ambulatory follow-up was scheduled with routine consultations every six months irrespective of clinical condition. Patients who experienced symptoms or who received an electrical shock from their ICD could also request additional urgent consultations. The main difference for patients in this group was that there was no mechanism for an asymptomatic event to be brought to the attention of the clinician before the next scheduled ambulatory follow-up consultation, which could potentially put the patient’s life at risk, particularly in the context of ICD dysfunction.
Data collection
Data was collected retrospectively from implantation records, ambulatory follow-up notes and hospitalisation reports. RM data was extracted from the RM database at our centre and from RM databanks online. RM notifications were classified either as reports or alerts, the former being routine transmissions and the latter being in response to abnormal events detected by the ICD, either from abnormal cardiac activity or ICD dysfunction.
Events recorded in both the RM and HO groups included: episodes of arrhythmia (supraventricular or sustained ventricular arrhythmias, irrespective of whether these triggered ICD therapy or not); inappropriate triggering of therapy (shock delivery or antitachycardia pacing); device related events (elective ICD replacement indicator, device reset); lead related events (noise, impedance, pacing and sensing threshold issues as well as lead perforation) and ICD programming issues (oversensing, undersensing, pacemaker re-entrant tachycardia, loss of cardiac resynchronization therapy [CRT] or phrenic nerve stimulation). Inappropriate therapies secondary to supraventricular arrhythmias were not considered as separate events and heart failure alerts (mean ventricular rate, percentage of CRT, premature ventricular contractions per hour, OptiVol ® [Medtronic Inc.]) were reported but not considered in our analysis as their relevance and clinical consequences have not yet been fully validated.
For both RM and HO groups, events were defined as symptomatic in cases resulting in ICD shock delivery or if the patient requested an urgent consultation for which they were seen the same day, otherwise, they were defined as asymptomatic. For instance, if a clinician requested an urgent clinical review following an RM alert, this was defined as asymptomatic so long as the patient had not experienced any symptoms and their ICD had not delivered a therapeutic shock. Patients experiencing asymptomatic events in the HO group would typically be diagnosed at the next routine ambulatory follow-up consultation. Outcomes of each asymptomatic event were recorded for both groups. Some resulted in intervention, such as ambulatory ICD reprogramming, changes in medication regime or scheduling of invasive procedures on the ICD and some did not. For asymptomatic events resulting in intervention, the time between the event onset and intervention was evaluated and recorded.
Endpoints
Our goals were:
- •
to assess the time between onset of asymptomatic events detected by ICD and clinical intervention, for cases where intervention was required;
- •
to investigate the impact of RM on the number of scheduled ambulatory follow-up consultations and urgent unscheduled consultations;
- •
to investigate the outcome of scheduled ambulatory follow-up consultations for patients in the RM and HO groups and;
- •
to assess the clinical relevance of RM notifications.
Statistical analysis
Statistical analysis was performed using SAS software (version 9.2, SAS Institute Inc., Cary, North Carolina). Continuous variables were expressed in terms of their mean and standard deviation values in cases where they followed a normal distribution and in terms of their median and interquartile range in cases where they did not. Qualitative variables were expressed in terms of frequency and percentage. Comparisons between groups were performed using the χ 2 and Fisher exact tests for qualitative data, and nonparametric Wilcoxon test for quantitative data with calculation of exact P values. A P -value < 0.05 was considered statistically significant.
Results
Study population
Out of the 478 eligible patients, 123 were excluded for one or more of the following reasons: follow-up at another centre ( n = 109), ICD removal before first outpatient consultation (heart transplantation: n = 1; device infection: n = 1), death before first outpatient consultation (cardiogenic shock: n = 1) and missing data ( n = 11, in cases where ambulatory follow-up notes or hospitalisation reports were missing). Data from the remaining 355 patients was analysed (RM: n = 144; HO: n = 211). Baseline characteristics of the study population are summarized in Table 1 . The mean overall follow-up time was 13.5 months, with a median of 363 days and IQR of [268–436] for the RM group vs. 460 days and [330–589] for the HO group; P < 0.0001, the difference between these figures being a result of the delay between ICD implantation and RM activation (median 89 days IQR [2–301]).
| RM group ( n = 144) | HO group ( n = 211) | P | |
|---|---|---|---|
| Age – years | 57.6 ± 14.8 | 62.2 ± 15.2 | < 0.01 |
| Male gender – n (%) | 118 (82) | 155 (73) | 0.06 |
| Left ventricular ejection fraction – % | 34.4 ± 14.6 | 36.3 ± 14.4 | 0.23 |
| NYHA functional class – n (%) | 0.36 | ||
| I-II | 116 (87.2) | 170 (82.9) | |
| III-IV | 17 (12.8) | 35 (17.1) | |
| Ischemic heart disease – n (%) | 85 (59.0) | 107 (50.7) | 0.13 |
| Comorbidities – n (%) | |||
| Diabetes mellitus | 38 (26.8) | 50 (23.7) | 0.54 |
| Hypertension | 59 (41.3) | 88 (41.7) | 0.93 |
| Chronic kidney disease | 13 (9.1) | 30 (14.2) | 0.14 |
| History of stroke | 12 (8.4) | 21 (10.0) | 0.62 |
| Atrial fibrillation or flutter | 35 (24.3) | 84 (39.8) | < 0.01 |
| Patients receiving ICD for secondary prevention – n (%) | 72 (50.0) | 105 (49.8) | 0.97 |
| Primary events prompting ICD implantation – n (%) | 0.95 | ||
| SCD due to VF or sustained VT | 42 (58.4) | 66 (62.8) | |
| Syncope with inducible sustained VT | 13 (18.1) | 17 (16.2) | |
| SCD without documented arrhythmia | 7 (9.7) | 9 (8.6) | |
| Syncope in high-risk cardiomyopathy | 4 (5.6) | 6 (5.7) | |
| Syncope and low LVEF | 6 (8.3) | 6 (5.7) | |
| ICD brand – n (%) | < 0.0001 | ||
| Biotronik | 75 (52.1) | 11 (5.2) | |
| Boston Scientific | 5 (3.5) | 64 (30.3) | |
| Medtronic | 54 (37.5) | 33 (15.6) | |
| Sorin | 0 | 59 (28.0) | |
| St Jude Medical | 10 (6.9) | 44 (20.9) | |
| ICD type – n (%) | |||
| Single chamber | 69 (47.9) | 68 (32.2) | < 0.01 |
| Dual chamber | 50 (34.7) | 83 (39.3) | 0.43 |
| CRT | 25 (17.4) | 60 (28.4) | 0.02 |
| Medications – n (%) | |||
| Beta-blocker | 131 (90.9) | 191 (90.5) | 0.69 |
| ACE-inhibitor or angiotensin-antagonist | 123 (85.4) | 177 (83.9) | 0.54 |
| Platelet agglutination inhibitor | 97 (68.8) | 132 (63.2) | 0.28 |
| Anticoagulant | 46 (32.6) | 74 (35.4) | 0.59 |
| Antialdosterone | 44 (31.2) | 60 (28.7) | 0.62 |
| Amiodarone | 21 (14.9) | 41 (19.6) | 0.26 |
| Digoxin | 9 (6.4) | 9 (4.3) | 0.39 |
| Diuretics | 72 (50.0) | 121 (57.3) | 0.19 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
