Discussion:   
In this retrospective cohort study, an S-ICD extraction rate of 5.9% was identified over a median follow-up period of 4.4 [2.0-6.5] years with a median length of implantation of 39.6 [8.3-4.6] months. The most common indications for S-ICD extraction or removal were need for bradycardia pacing, infection, and inappropriate shocks due to oversensing. Both a history of smoking and a higher BMI were independently associated with S-ICD extraction.
Prior studies have investigated common complications leading to S-ICD extraction. A pooled analysis of patients from the S-ICD IDE study and EFFORTLESS registry cited an incidence of S-ICD extraction for pacing needs of 0.4% and an incidence of infection requiring extraction or revision of 1.7%.18 A secondary analysis on the PRAETORIAN trial showed that though S-ICDs caused lower rates of lead-related complications (1.4% in S-ICDs vs. 6.6% in TV-ICDs, p<0.001) and systemic infection (0% in S-ICDs vs. 1.2% in TV-ICDs, p=0.03), they were still shown to cause greater rates of pocket bleeding (22.2% in S-ICDs vs. 4.1% in TV-ICDs).18,19A large tertiary center study at the University Hospital Münster cited several complications of S-ICDs including oversensing with inappropriate shock delivery, need for pacing, hematoma, infection, hypermobility and ineffective shocks; however, those requiring S-ICD extraction were only in patients who had a need due to oversensing (incidence, 0.85%), new pacing requirements (0.57%), ineffective shock (0.28%), and infection (0.57%) with a total extraction rate of 2.85%. Other complications, such as pocket hematomas, device position change, and re-programming to correct oversensing, were corrected via revisions only.13 In April 2022, the University of Pennsylvania published a study which cited an incidence rate of S-ICD extraction rate of 12.9% with a mean dwell time of 20 months. Indications for extraction cited in this study included infection (incidence rate, 5.12%), improper device sensing (3.90%), pacing need (2.93%), heart transplantation (1.71%), patient discomfort (0.73%), and other less common device technical issues.14 These two studies by the University Hospital Münster and the University of Pennsylvania have shown a wide variability in rates of S-ICD extraction, demonstrating that current understanding of the nature of S-ICD extraction is limited.13,14 The S-ICD extraction rate found in the current study adds to this limited literature, demonstrating a low S-ICD extraction rate of 5.9% with appropriate patient selection. The present study also found that need for bradycardia pacing (1.08%), infection (1.34%), and inappropriate shocks due to oversensing (1.34%) were the most common specified indications for extraction, consistent with findings from prior studies. Other, less common, indications for extraction included chronic resynchronization therapy needs (0.54%), patient preference (0.81%), heart transplantation (0.54%), and failure to shock (0.27%) (Figure 1).
In addition, several studies have also attempted to identify risk factors associated with S-ICD extraction. The S-ICD Post Approval Study identified four risk factors of S-ICD-related infection: patients with previous ICD implantation (extracted for any reason), age ≤ 55 years old, left ventricular ejection fraction ≤ 30%, and patients with diabetes.20 Other studies evaluating risk factors for S-ICD extraction have shown that a history of repeat or multiple implantation procedures, perioperative fever, use of temporary pacing, diabetes requiring insulin, end stage renal disease requiring hemodialysis, crowded operating theaters, immunosuppression, corticosteroid use, the presence of central lines, and malignancy all confer a higher risk for infection associated with the new device.21-25 Results from the current study add to these prior data and suggest that a history of smoking, elevated BMI, and prior pocket infection are also associated with S-ICD extraction due to infection. While elevated BMI is a factor associated with S-ICD extraction for inappropriate shocks in the present study, it is to be noted that S-ICD implantation in obese patients can be safe and feasible if using appropriate technique.26
These results suggest that the S-ICD is usually well tolerated in appropriately selected patients, but care should be taken during S-ICD evaluation in patients with a history of smoking or elevated BMI. Further work must be done to determine whether this study’s findings regarding rates of, and indications for, S-ICD extraction are transferrable to other health systems.
Limitations:
The results of this study are limited by its retrospective nature and focus on a single healthcare system. The present study occurred in a large metropolitan center and could capture clinical and sociodemographic characteristics that differ from cohort profiles in other studies. Additionally, there is wide variability in provider preferences when it comes to opting for device implantation in patients who are considered borderline in candidacy for S-ICDs and extracting devices in the setting of local pocket infection rather than treating with antibiotic therapy.6 Future analyses may compare indications for S-ICD vs TV-ICD extraction to provide results which may shed light on relative benefits of one ICD system in specific patient populations. Furthermore, device technology improved dramatically during 2010-2022 and could have impacted device-related complication rates and provider preferences. Finally, contemporary studies examining patient outcomes and cost effectiveness have sparked an evolution of device guidelines and expansion of indications for S-ICD implantation, changing how S-ICDs have been implemented over the years.2, 3, 27 Other factors influencing device selection could be further analyzed including patient’s socioeconomic status, patient preferences, and cost.