Subcutaneous or Transvenous Defibrillator Therapy

Reinoud E. Knops*, Louise R. A. Olde Nordkamp, Peter-Paul H. M. Delnoy, Lucas V. A. Boersma, Juergen Kuschyk, Mikhael F. El-Chami, Hendrik Bonnemeier, Elijah R. Behr, Tom F. Brouwer, Stefan Kaab, Suneet Mittal, Anne-Floor B. E. Quast, Lonneke Smeding, Willeke van der Stuijt, Anouk de Weger, Koen C. de Wilde, Nick R. Bijsterveld, Sergio Richter, Marc A. Brouwer, Joris R. de GrootKirsten M. Kooiman, Pier D. Lambiase, Petr Neuzil, Kevin Vernooy, Marco Alings, Tim R. Betts, Frank A. L. E. Bracke, Martin C. Burke, Jonas S. S. G. de Jong, David J. Wright, Jan G. P. Tijssen, Arthur A. M. Wilde, PRAETORIAN Investigators

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review



The subcutaneous implantable cardioverter-defibrillator (ICD) was designed to avoid complications related to the transvenous ICD lead by using an entirely extrathoracic placement. Evidence comparing these systems has been based primarily on observational studies.


We conducted a noninferiority trial in which patients with an indication for an ICD but no indication for pacing were assigned to receive a subcutaneous ICD or transvenous ICD. The primary end point was the composite of device-related complications and inappropriate shocks; the noninferiority margin for the upper boundary of the 95% confidence interval for the hazard ratio (subcutaneous ICD vs. transvenous ICD) was 1.45. A superiority analysis was prespecified if noninferiority was established. Secondary end points included death and appropriate shocks.


A total of 849 patients (426 in the subcutaneous ICD group and 423 in the transvenous ICD group) were included in the analyses. At a median follow-up of 49.1 months, a primary end-point event occurred in 68 patients in the subcutaneous ICD group and in 68 patients in the transvenous ICD group (48-month Kaplan-Meier estimated cumulative incidence, 15.1% and 15.7%, respectively; hazard ratio, 0.99; 95% confidence interval [CI], 0.71 to 1.39; P=0.01 for noninferiority; P=0.95 for superiority). Device-related complications occurred in 31 patients in the subcutaneous ICD group and in 44 in the transvenous ICD group (hazard ratio, 0.69; 95% CI, 0.44 to 1.09); inappropriate shocks occurred in 41 and 29 patients, respectively (hazard ratio, 1.43; 95% CI, 0.89 to 2.30). Death occurred in 83 patients in the subcutaneous ICD group and in 68 in the transvenous ICD group (hazard ratio, 1.23; 95% CI, 0.89 to 1.70); appropriate shocks occurred in 83 and 57 patients, respectively (hazard ratio, 1.52; 95% CI, 1.08 to 2.12).


In patients with an indication for an ICD but no indication for pacing, the subcutaneous ICD was noninferior to the transvenous ICD with respect to device-related complications and inappropriate shocks.

Original languageEnglish
Pages (from-to)526-536
Number of pages11
JournalNew England Journal of Medicine
Issue number6
Publication statusPublished - 6 Aug 2020



Cite this