Remote monitoring for prevention of inappropriate implantable cardioverter defibrillator shocks: is there no place like home?

Although the implantable cardioverter defibrillator (ICD) has made a significant impact on prevention of sudden arrhythmic death, important limitations of this therapy remain. Chief among these is the risk of inappropriate shocks, which have been reported in 10–25% of patients in randomized trials and case registries over 2–4 years of follow-up. 1 Inappropriate shocks may cause pain, proarrhythmia, and even death in rare cases. Delivery of inappropriate shocks has also been associated with increased late mortality in several large ICD trials. Patients who experience multiple shocks may develop anxiety and other signs of psychological distress, and concern about this complication is commonly expressed by patients. The most common cause of inappropriate shocks is atrial fibrillation, followed by other supraventricular tachycardias, and lead-sensing abnormalities. The latter category includes conductor fracture, insulation failure, T-wave oversensing, and myopotential oversensing. 2 A recent study examined 990 patients implanted with a variety of ICD lead models between 1992 and 2005 and found lead failure rates of 15% at 5 years and 40% at 8 years, with a pattern of accelerating annual failure rates over time. 3 One-third of ICD lead failures in this study presented with inappropriate shocks. In patients dependent upon ventricular pacing, ICD lead failure may also lead to symptoms from inhibition of pacing or failure to capture. Concern regarding ICD lead failure has increased recently with the recall of a widely used lead. 4 Manufacturers' product performance reports suggest early annual lead failure rates of 0.5–1% of latest generation of ICD leads, while the failure rate of the Sprint Fidelis family is approximately two- to three-fold higher. A 2007 review of the US Food and Drug Administration's Manufacturer and User Facility Device Experience database found 679 reports of failed Sprint Fidelis leads, 33% of which were associated with inappropriate shocks. 4 Strategies to minimize the risk of inappropriate shocks include using high ventricular tachycardia/ventricular fibrillation rate cutoffs, multiple tachycardia detection zones, longer time to detection, anti-tachycardia pacing, supraventricular tachycardia discrimination algorithms, and medical therapies such as β-blockers and anti-arrhythmic drugs. 5 These strategies are less useful for the prevention of inappropriate shocks caused by ICD lead failure. For this purpose, algorithms have been devised to trigger device alert tones, device vibration, or automatic change in tachycardia detection criteria. 6 None of these algorithms, however, has been prospectively shown to reduce the risk of inappropriate shocks, and improved strategies to prevent inappropriate shocks are clearly needed. Spencker et al. report on their single-centre experience using an automatic wireless remote monitoring system for surveillance of ICD systems. 7 They retrospectively examined 54 patients who underwent ICD lead revision for the failure of the lead over a 6-year period. The 11 patients who had been undergoing daily remote wireless assessment were compared with the 43 patients who had been followed only with quarterly in-office visits. They found that fewer patients undergoing remote monitoring (27.3%) experienced ICD shocks for sensing failure prior to lead revision than those with standard in-office follow-up (46.5%). This trend reached statistical significance when symptomatic pacemaker inhibition was added to the endpoint (27.3% in remote monitoring group vs. 53.4% in the standard follow-up group, P = 0.04). In addition, the authors found that in the standard follow-up patients, evidence of non-sustained oversensing was detected by the ICD a mean of 16 days prior to the onset of symptoms, suggesting that inappropriate shocks could have been avoided in most of these patients, had they been followed with remote monitoring and contacted in a timely fashion. Although the report of Spencker et al. adds important new information on the potential value of automatic wireless remote monitoring systems, there are some limitations which must be considered. The study was a retrospective analysis of a small subset of patients being followed with remote monitoring, rather than a prospective, randomized comparison. Patients in the non-monitored group had older leads, which may have been more prone to cause shocks when failure occurred. Because the total number of patients followed with remote monitoring is not given, the yield of this strategy for the detection of lead failure is not known. Finally, most inappropriate ICD shocks are not caused by lead failure, but rather by supraventricular arrhythmias, and this study does not provide any evidence that home monitoring reduces risk of inappropriate shocks from this cause. Despite these limitations, the valuable contribution of Spencker et al. provides some of the first evidence that automatic wireless remote monitoring systems may reduce the risk of inappropriate ICD shocks, as was previously predicted. 2 Since the initial introduction of ICD remote monitoring systems in 2001, physicians have anticipated the potential benefits to patient care and convenience. 8 Remote monitoring systems have now been developed by each of the major ICD manufacturers and each employs a somewhat different technology and degree of automation. 9 There are two major potential uses for remote monitoring systems. The first is to substitute for some or most in-office visits, whether routine or patient-initiated to evaluate symptoms or perceived ICD shocks. The primary anticipated benefits of this function are patient convenience and perhaps reduction of device clinic workload, improved efficiency, and improved patient compliance with follow-up. 10 This function is not dependent on automatic wireless transmission of data, and is not likely to measurably improve the quality of care of ICD patients. The second potential function is to provide continuous monitoring for early detection of asymptomatic arrhythmias, generator and lead abnormalities, and signs of heart failure. It is this capacity which has the greatest potential to improve patient care and safety of ICD therapy, and several case reports have already shown that automatic wireless monitoring led to early detection of potentially life-threatening device complications. Although the benefits of remote monitoring may seem obvious and unalloyed, there are significant concerns that must be addressed prior to wholesale adoption of this technology. Remote monitoring systems pose a challenge to workflow in busy ICD clinics which must be accommodated. There is risk of significant ‘information overload’ for both physicians and allied professionals. One report examining early generation remote monitoring in 271 ICD patients found a 14% ‘false-negative’ rate, i.e. a remote report that did not alert the clinician to important, relevant information. 11 The systems cannot be used to reprogram devices, and manual pacing threshold testing cannot be performed. Although the quality and quantity of information from remote reports has improved with newer generation systems, it remains uncertain whether remote monitoring can substitute for in-office visits in all patients. It is also unclear whether the increased quantity of information provided by automatic remote monitoring will always improve care. Cost is another important consideration. A variety of payment models have been developed, with most vendors moving towards inclusion of remote monitoring in the initial cost of the ICD generator, increasing the purchase price by as much as several thousand US dollars. Other models use a monthly fee for the monitoring service. Regardless of how the accounting is done, the device hardware and costs of monitoring must be borne by some elements of the health care system. Implantable cardioverter-defibrillator therapy is already criticized in some quarters for being relatively cost-ineffective, especially in patients at the lower end of the primary prevention risk spectrum, and to justify adding the expense of remote monitoring to an already expensive therapy requires evidence of what utility is gained from the incremental expense. As an example, if one assumes that automatic remote wireless monitoring costs $300 US annually in incremental cost of device hardware and product support, that ICD leads have a 2% annual failure rate, one-third of which result in inappropriate shocks, and that remote monitoring can reduce this rate by 50%, then one finds a cost of $90 000 to prevent one case of inappropriate shocks. Clearly, this cost–benefit ratio could change significantly depending on the economic and clinical assumptions, as well as factoring in other benefits in outcome and quality of life. Fortunately, prospective studies are underway which should provide this information and thereby allow us to better assess the costs and benefits of remote monitoring in more specific terms. The TRUST study is enrolling 1000 patients who are randomly assigned in a 1:2 ratio to receive either usual follow-up or daily wireless remote monitoring with the Biotronik Home Monitoring system for 15 months. 12 Results are expected in mid-2009. The CONNECT study has a similar design using the Medtronic Carelink system, enrolling 2000 patients who will be assigned 1:1 to usual follow-up and remote monitoring. 13 This trial will include patients with biventricular ICDs, with results expected in late 2009. These two randomized trials, in addition to ongoing large prospective registries, should fill important gaps in our understanding of how remote monitoring systems may be used to improve the quality of care for ICD patients. The theoretical basis for their use is compelling. Whether their potential to improve the safety of ICD therapy, as suggested by the study of Spencker et al. , is fulfilled remains to be proven. Conflict of interest: none declared.