Presenter | see Discussant report
Introduction: Implantable cardioverter defibrillators (ICD) remote monitoring, using transmission of clinical and technical data and intracardiac electrograms to a central receiving station, attracts attention from cardiologists as a way to improve quality of care and reduce costs. However, we have limited information on the safety of the system.
Methods: The ECOST randomized, multicenter, non-inferiority trial examined the safety of long-term remote monitoring of ICD. Between January 2007 and April 2008, 433 patients were randomly assigned to remote monitoring follow-up (active group) versus standard care (control group). Patients assigned to active group underwent automatic daily remote monitoring and were seen in the ambulatory department at yearly intervals, unless an anomalous ICD function or an event of clinical concern was reported by remote monitoring, requiring an ambulatory visit. Patients assigned to the control group, were followed in the ambulatory department at 6-month intervals. The overall follow-up duration was 27 months. The primary objective was to confirm that the proportion of patients experiencing at least 1 major adverse event (MAE), including all-cause mortality, major cardiac adverse events and major device related adverse events, was not higher in the active than in the control group. Number of delivered shocks and battery longevity were compared in both groups. This trial is registered with ClinicalTrials.gov, number NCT00989417.
Results: The characteristics of the study groups were similar (88% male, 62±13 years, 53.6% primary prevention ICD indication, 30.3% dual chamber implants, 85.3% first implantation, 34.913.3% left ventricular ejection fraction). Over a follow-up of 24.2±7.3 months, 38.5% of patients in the active and 41.5% in the control group experienced at least 1 MAE (p<0.05 for non-inferiority). The total number of delivered shocks (193 vs. 657) was significantly reduced in the active compared to the control group (p<0.05). The total number of inappropriate shocks (28 vs. 283) was also reduced, with a number of patients suffering from inappropriate shocks lower in the active group (11 vs. 22) p<0.05. At the end of the study, the battery longevity was greater in the active group because of a number of capacitor charges significantly reduced (499 vs. 2081).
Conclusions: Our results are consistent with, and extend, previous findings by definitively showing that long term remote ICD monitoring is not less safe than conventional follow-up in terms of all cardiac or device related major adverse events and all cause deaths. Moreover it can significantly lower the number of delivered shocks.
Discussant | see Presenter abstract
We have to congratulate Professor Salem Kacet and the investigators of ECOST for having conducted this important trial with success
This is an important trial as the current clinical evidence to support remote monitoring of ICD patients is relatively poor. We essentially have the results of two recent controlled studies, TRUST and CONNECT that used surrogate markers to assess safety and efficiency: the number of total in-office visits in TRUST and time from clinical event to clinical decision in CONNECT. Within these limitations, the two studies were positive. As logically expected, the 1-year number of in-office visits per patient in TRUST was significantly greater in the standard care group as compared to home monitoring… but this is not really surprising! More interestingly in the two trials, the median time from event to evaluation or clinical decision was highly significantly reduced in the remote monitoring groups.
ECOST like EVATEL, is important as they are the first randomised trials to assess the impact of remote monitoring on the risk of major clinical events. As shown by Philippe Mabo and Salem Kacet in their presentation, the two studies have nearly similar design, nearly similar inclusion criteria and nearly similar endpoints. They differ slightly in the objectives: ECOST was primarily designed as a Safety study where the clinical events are described as major adverse events, when EVATEL was designed as both a Safety and Efficiency study.
ECOST has validated its primary objective of non-inferiority. Home monitoring is not inferior to standard care as regards to clinical safety. These data are consistent with those of TRUST that assessed clinical safety as secondary endpoint. The main strength of ECOST is to show that the effect is maintained over 27 months when TRUST was limited to 12-month follow-up. Contrary to TRUST, a trend to more favourable outcome in the Home monitoring group was observed in ECOST but the difference was not statistically significant and the curves tended to converge after one year. ECOST validated the hypothesis of non-inferiority when EVATEL did not. Is there an explanation for these apparently different results? As said before the two studies were nearly similar as regards to design, inclusion criteria and endpoints. There are actually some differences to point out as they may contribute to explain different results. ECOST was an industry driven trial when EVATEL was an institutional initiative. ECOST assessed one particular remote system, the Home monitoring system when EVATEL assessed all the systems available on the market. EVATEL can thus be considered as more representative of the real life of remote monitoring. Patient number was four times larger in EVATEL than in ECOST but the follow-up time was twice longer in ECOST than in EVATEL. Finally and due to different periods of time, secondary prevention indications were largely predominant in ECOST when there was a balance between primary and secondary prevention indications in EVATEL.
In practice, even if apparently different the results of the two studies appear to be consistent. Both studies show that remote monitoring of ICD patients is clinically safe in comparison with standard care. Both studies don’t provide evidence that remote monitoring may contribute to better prevent major clinical events. This is not really surprising and cannot be considered as disappointing. On the other hand, both studies show that remote monitoring is associated with a significant reduction in the risk of inappropriate shocks. The relative risk reduction was 37 and 52%, respectively. In addition, ECOST shows that Home monitoring is associated with a highly significant reduction in the risk of total shocks and of charged shocks. This finding may have important practical consequences in terms of device longevity and patient quality of life.
In conclusion, the results of ECOST are globally positive and support a wider use of Home monitoring to follow ICD patients. We can only regreat that a study named ECOST does not answer the pending question of cost and cost-effectiveness of remote monitoring.