With subcutaneous ICDs (S-ICD) not offering the same range of pacing therapies as transvenous systems (T-ICD), it is unclear how many ICD recipients may be suitable for a S-ICD. In addition, the benefit from ICD therapy is dependent on non-sudden cardiac death (SCD) risk. The study aims were to  establish how many ICD recipients may be suitable for S-ICD implantation based on a range of selection policies, and  determine the impact of S-ICD use on the risk of non-SCD.Methods
We performed a retrospective study of all ICD implants over 5 years (2009–2013). We evaluated how widely the S-ICD could be utilised using the following indications: (a) Least liberal use – S-ICD used only in patients with inherited channelopathies and idiopathic ventricular fibrillation; (b) Intermediate use – S-ICD also used for SCD primary prevention in hypertrophic cardiomyopathy; (c) Most liberal use – S-ICD also used for SCD primary prevention in coronary artery disease (CAD) and non-ischaemic dilated cardiomyopathy (DCM) who have a narrow QRS complex (120msec). Given the potential impact of age on the benefit from S-ICDs, we stratified our results by implant age (70, 60 and 50 years). Non-SCD risk was quantified using the MADIT-II risk score (score 0–5, 5 (highest risk)). Mann-Whitney U testing was used to compare risk scores.Results
Over the 5 years, there were 402 implants (79% male, age 70±13 years). After exclusion of patients implanted with a CRT-D device or with a pacing indication, 219 remained for consideration of a S-ICD. Using our definition of least liberal use, intermediate use, and most liberal use, the number of patients considered suitable for a S-ICD was 18 (4.5%), 30 (7.5%) and 77 (19.2%), respectively. Age as a discriminator had a significant impact on the proportion of patients considered suitable for a S-ICD. This was predominantly due to a reduction in the number of patients with CAD and DCM being considered for a prophylactic device. Using the most liberal definition, the number of potential S-ICD candidates dropped from 52 (12.9%) in those 70 years old at implant, to 34 (8.5%) for 60 years, to 20 (5%) using 50 years as the cut-off.Results
The MADIT-II score was 0.22±0.1, 0.47±0.1 and 0.96±0.1, in the least liberal, intermediate and most liberal use S-ICD groups, respectively. The score was 2.07±0.1, 2.11±0.1 and 2.24±0.1 in the respective T-ICD groups. The mean MADIT-II score in the S-ICD group was significantly lower than that in the respective T-ICD group (p<0.0001, for all three selection policies).Conclusions
While up to one-fifth of patients currently receiving ICDs may be suitable for a S-ICD, this result is dependent on both which patient groups the S-ICD is used in and whether an age cut-off is applied. The cohort of patients suitable for a S-ICD also have a lower risk of non-SCD compared to those receiving conventional T-ICDs.