This study has three main findings: 1) innovative EVDs were adopted relatively quickly in the Netherlands, and especially the S‑ICD has become a substantial part of device therapy in the Netherlands; 2) utilisation of S‑ICDs was highest among younger patients, while use of LPs was not higher in specific age groups; 3) EVDs were mainly implanted in high-volume centres with cardiothoracic surgery available, although S‑ICDs are increasingly being implanted in centres without cardiothoracic surgery on-site, potentially leading to a broader adoption than LPs in the Netherlands.

This study with nationwide registration data demonstrated that the use of EVDs has been increasing in the Netherlands since their introduction and that S‑ICDs in particular are currently used in a substantial proportion of patients. This study highlights that patients aged 10–40 years with a single-chamber ICD indication form a subgroup with a very high S‑ICD adoption. More than half of those patients received an S‑ICD. Studies from other countries show a similar adoption, including an almost identical distribution of S‑ICD use over age groups in a Japanese cohort [8, 9]. Implanters often choose S‑ICDs in younger patients due to their increased risk of lead-related complications because of their more active lifestyle and longer need for ICD therapy [10]. Also, younger patients rarely have a concomitant indication for antibradycardia, antitachycardia or cardiac resynchronisation therapy pacing. The adoption of LPs was not higher in patients of specific age groups. The finding that LP recipients were slightly younger than conventional single-chamber PM recipients should be put in perspective, as the studied PM population is relatively old compared with the vastly larger group of dual-chamber PM recipients [11]. We believe that among all PM recipients, older patients are generally favoured for LP therapy, due to the current absence of dual-chamber LPs and probably also due to the uncertain end-of-life strategy of LPs, either replacement or abandonment. The relatively higher adoption of S‑ICDs among female patients was also seen in a US database [8].

S‑ICDs were numerically mostly implanted in high-volume centres, but middle- and high-volume centres showed an equal adoption with S‑ICD implantations in approximately one in six patients with a single-chamber ICD indication. This difference in use related to implantation volume might be explained by a higher percentage of younger patients who more often have genetic or congenital heart disease for which patients are typically referred to specialised centres. However, an increasing adoption over time beyond centres with cardiothoracic surgical back-up was demonstrated. Most LP implantations were in high-volume centres, where the use of LPs was highest. The presumable explanation is that the high-volume centres include more centres with cardiothoracic surgery on-site, which is mandatory for LP implantation in the Netherlands. There were significant differences in the adoption of EVDs between regions in the Netherlands, with a two- or three-times higher use of EVDs in the Northern region compared with the Southern region. This may potentially be caused by the centre types in each region and whether centres engaged early on with this technology in clinical trials.

In the most recent years, it can be noted that the growth of EVD use has stopped in the Netherlands. This may be due to temporal causes with the covid-19 pandemic in 2020 and its logistic difficulties. However, the stagnation may potentially also be due to structural causes, such as higher costs of EVDs (approximately factor 3–4 higher). Further, LP implantation is restricted to centres with cardiothoracic surgery available, as perforations occur in ~1% of implantations and require emergency surgery in an estimated quarter of those [12,13,14,15]. On the other hand, a further increase in the use of EVDs is expected due to additional experience and broadening of the indication areas using novel EVDs. The main current limitation of S‑ICDs is their lack of pacing capabilities. However, in February 2023, the extravascular ICD (EV-ICD; Medtronic, Minneapolis, MN, USA), which is capable of antitachycardia pacing and back-up pacing through a substernal lead, received Conformité Européenne (CE) approval [16]. Further, a modular system consisting of a communicating S‑ICD and antitachycardia pacing-enabled LP is currently being studied in the first clinical trial [17, 18]. Cardiac resynchronisation therapy using a subcutaneous pulse generator and left ventricular LP is also being studied clinically [19]. The guideline-directed indication area for LPs is expected to expand too. The Micra AV, capable of atrioventricular synchronous pacing, was CE-approved in June 2020 and only the first cases in the Netherlands are described in this study. A clinical trial with a dual-chamber LP, the Aveir DR (Abbott, Medical Inc., Abbott Park, IL, USA), demonstrated adequate results and has already received approval from the Food and Drug Administration [20]. Lastly, the use of LPs may also be influenced by the emergence of conduction system pacing, which has different benefits than LPs. At the moment, patients may simultaneously be good candidates for leadless and conduction system pacing, and a choice should be made based on individual characteristics. Yet, novel LP designs that can pace the conduction system are being studied. To improve the data collection regarding all EVDs including novel devices, a Dutch nationwide registry with long-term follow-up of EVDs will be incorporated in NHR’s device registry [21].

This study was limited by the lack of data after 2020. Therefore, not all currently available EVD models are described. Also, within the conventional single-chamber pacemakers, we could not distinguish between AAI(R) and VVI(R) pacemakers. However, we do think that this group mainly consists of VVI(R) pacemakers as AAI(R) pacemakers are rarely implanted in the Netherlands [1]. Also, no distinction could be made between transvenous and abdominal CIEDs, although the large majority of conventional CIEDs described in this study are expected to be transvenous CIEDs. Lastly, the data presented are purely descriptive and therefore, do not infer causality.