Since the first implantable cardioverter defibrillator (ICD) was introduced in 1980, ICDs provide lifesaving therapy for patients at risk for ventricular arrhythmias [1, 2]. Transvenous ICD (TV-ICD) systems have been the first-line therapy since their introduction in the early 1990s. However, besides their lifesaving capacities, the transvenous leads carry their own risk. Lead-related complications and systemic infections are severe side effects causing significant morbidity and mortality [3].

The subcutaneous ICD (S-ICD) was designed to reduce lead-related complications and reducing the risks associated with systemic infections and device extractions, by creating an extra-thoracic, implantable, defibrillator system without the need for venous access [4]. Consisting a single lead placed on the sternum, the S-ICD is not able to provide chronic pacing except from 30 s on demand post-shock pacing. In the absence of chronic pacing capabilities, the S-ICD is not suitable for patients requiring either bradycardia, anti-tachycardia (ATP), or cardiac resynchronization pacing.

In the past years, several studies have shown similar safety and efficacy of the S-ICD [5, 6]. A recent meta-analysis compared observational data of outcomes in patients implanted with TV-ICD and S-ICDs [7], but currently no randomized data is available. The ongoing trial, a Prospective, Randomized Comparison of Subcutaneous and Transvenous Implantable Cardioverter Defibrillator Therapy (PRAETORIAN) trial, will be the first randomized study comparing both devices head-to-head (ClinicalTrials.​gov NCT01296022), of which the first results are expected in 2020 [8]. In addition, the randomized trial Avoid Transvenous Leads in Appropriate Subjects (ATLAS S-ICD), which started enrollment last year, will also compare single-chamber TV-ICDs with S-ICDs (ClincialTrials.​gov NCT02881255). In the absence of randomized data, the latest ESC guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death now give the S-ICD a class IIa recommendation for patients without need of bradycardia, anti-tachycardia pacing (ATP), or cardiac resynchronization therapy [9]. The recently published AHA/ACC/HRS guideline for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death give a S-ICD class I recommendation for patients at high risk for infection or without adequate venous access, without an indication for pacing or ATP [10]. As clear guidelines on when to select a subcutaneous ICD (S-ICD) over a transvenous ICD (TV-ICD) are lacking, the aim of this review is to provide a pooled overview of the clinical outcomes and discuss device selection considerations in order to aid physicians in selecting the optimal device for the individual patient.

A previously published systematic review and meta-analysis included five matched head-to-head comparison studies [7]. Since the publication of this meta-analysis, several new matched comparisons have been published including one with more patient years (PYs) of follow-up than the previous meta-analysis [11‐13]. Therefore, in this review, we added the most recently published studies to the existing meta-analysis and pooled the results for clinical outcomes. The outcomes were device-related complications, lead complications, device infection, appropriate shocks (AS) and inappropriate shocks (IAS). We present an incidence rate ratio (IRR) as this methodology corrects the number of events for the duration patients have been followed as studies had widely ranging follow-up durations. We obtained the IRR from the random effects model. Of the included studies, two of the seven only evaluated short-term in hospital outcomes [13, 14]. The heterogeneity, variation in study outcomes between studies, is assessed by using I² statistic, with I² > 50% indicating significant heterogeneity.

Seven studies were included in this analysis [11‐17]. The total number of included PYs was 1840 PYs in the S-ICD group versus 2288 PYs in the TV-ICD group. Figure 1a shows the device-related complication rate of all seven studies comparing TV-ICDs with S-ICDs. The pooled IRR of all-cause complications was 0.90 (95% CI 0.58–1.42), non-significantly in favor of the S-ICD. Five of the seven studies reported lead-related complications (Fig. 1b) and all five showed a decrease in favor of the S-ICD. The pooled IRR was 0.15 (95% CI 0.06–0.33), in favor of the S-ICD with 0% heterogeneity between studies. All seven studies compared infections in TV-ICDs versus S-ICDs (Fig. 1c). The pooled IRR was 2.00 (95% CI 0.95–4.22), suggesting more infection-related complications in S-ICD patients.

IAS were reported by six of the seven studies (Fig. 1d). The pooled IRR was 1.17 (95% CI 0.77–1.79), non-significantly in favor of TV-ICDs. All studies reporting the cause of IAS showed that those in TV-ICD, patients were primarily for supraventricular tachycardia and those in S-ICD due to T-wave oversensing (TWOS). Five out of the seven studies reported AS and four showed an increase in AS in TV-ICD patients [11, 12, 15‐17]. The pooled IRR was 0.68 (95% CI 0.42–1.03), non-significantly in favor of S-ICDs (Fig. 1e).

The S-ICD has proven to be a good alternative for TV-ICD implantation. With randomized data underway, the observational data demonstrates that the S-ICD is associated with reduced lead complications, but this has not yet resulted in a significant reduction in the total number of complication compared to TV-ICDs. Since the introduction of the S-ICD, important steps have been made to facilitate widespread adoption. New technologies such as a communicating leadless pacemaker with ATP capabilities and SMARTPASS for IAS reduction will make the S-ICD an even more attractive alternative, and new strategies for both anesthesia and defibrillation testing may soon reduce the logistical complexity of S-ICD implantation.