Transvenous versus subcutaneous implantable cardiac defibrillators for people at risk of sudden cardiac death

Objectives

This is a protocol for a Cochrane Review (intervention). The objectives are as follows:

To evaluate the benefits and harms of using a transvenous implantable cardioverter‐defibrillator (TV‐ICD) compared to a subcutaneous implantable cardioverter‐defibrillator (S‐ICD) in patients at risk of sudden cardiac death.

Background

Description of the condition

Sudden cardiac arrest (SCA) is the abrupt end of cardiac activity resulting in cardiovascular decompensation, where the victim does not respond, has irregular breathing and exhibits no signals of circulation — this can result in sudden cardiac death (SCD) if no immediate corrective measures are taken (Buxton 2006; Al‐Khatib 2017). Globally, cardiovascular illnesses account for an estimated 17 million deaths annually, 25% of which are SCD (Mendis 2011). Sudden cardiac arrest can result from ventricular arrhythmias (VAs) such as ventricular tachycardia (VT) or ventricular fibrillation (VF). Without resuscitation, these VAs could lead to sudden cardiac death. Based on different cardiac disorders, the risks of VA and SCD differ from population to population; and can be further traced to specific family history and genetic variations.

SCD represents a major public health issue that accounts for an estimated 50% of all cardiovascular deaths where a minimum of 25% are first symptomatic cardiac events (Myerburg 2001; Fishman 2010; Goldberger 2011; Myerburg 2012). The annual incidence of SCA in the general population of USA and Europe is within the range of 50 to 100 per 100,000, accounting for approximately 300,000 annual deaths in the USA, with comparable figures in Europe (Fishman 2010; Priori 2015). The latest cardiovascular statistics in 2017 from the American Heart Association (AHA) estimated the total annual burden of out‐of‐hospital cardiac arrest at 356,500 (Mozaffarian 2016; Benjamin 2017). A further 209,000 in‐hospital cardiac arrests take place on an annual basis. The numbers for out‐of‐hospital cardiac arrest survivals remain unsatisfactory, as the overall projected survival rate is 10% (Merchant 2011). In the UK, an estimated 80% of the 70,000 SCDs in England and Wales in 2010 can be attributed to VA, with a survival rate for out‐of‐hospital VA cases averaging around 7% (NICE 2014).

People with cardiomyopathy, whether ischaemic or non‐ischaemic, are at increased risk of developing ventricular arrhythmias due to underlying structural and mechanical factors (e.g. increased inward sodium and calcium currents due to myocardial stretch) (Wu 2007). Sustained VA accompanied by an acute coronary syndrome presents more often as polymorphic ventricular tachycardia (VT) or ventricular fibrillation. Reports showed that VT or VF after primary percutaneous interventions at any time was associated with a substantially increased risk of death within 90 days. Forty‐eight hours after hospital admission, VT or VF were more associated with increased risk of death than VT or VF diagnosed within 48 hours of hospital presentation (Volpi 1989). Several reports displayed that patients with a history of myocardial infarction (MI), spontaneous nonsustained VT (NSVT), and a depressed left ventricular ejection fraction (LVEF) below 40% showed an increased risk of sudden death (Al‐Khatib 2017).

Moreover, hypertrophic cardiomyopathy (HCM) displays a greater risk of developing SCA or SCD; they are particularly common in individuals aged below 40 (Maron 2000). HCM is considered as the main cause of cardiac arrest in individuals displaying effort or exertion, as exertion intensifies the risk of life‐threatening ventricular arrhythmias. In prior asymptomatic individuals, sudden death can be the first result of the disease. A judgement on substantial risk of future VT or VF with this condition can be derived by looking into the history of prior cardiac arrests in the individual’s family (Maron 2000).

Ventricular arrhythmias can also be affected by genetic syndromes that include the long‐ and short‐QT syndromes, Brugada syndrome, idiopathic VF, and catecholaminergic polymorphic VT (Moss 2002; Schimpf 2003; Viskin 2003; Zareba 2003; Brugada 2004; Goel 2004; Goldenberg 2006). These syndromes are usually present in the absence of any structural heart disease and can result in sudden cardiac arrest. It is generally approved that individuals with prior cardiac arrest or temporary loss of consciousness are at high risk for recurrent arrhythmic events, even though disagreements on risk factors for sudden death in individuals with these syndromes still remain (Al‐Khatib 2017).

Description of the intervention

In prevention of sudden death, several clinical trials have assessed the advantages and disadvantages of implantable cardioverter‐defibrillators (ICDs) in multiple patient populations at risk of SCD, counting those with previous MI and heart failure caused by either coronary artery disease or non‐ischaemic cardiomyopathy (NICM) for primary and secondary prevention. Currently, the pharmacological therapy that improves survival for patients with VA are beta blockers (e.g. metoprolol succinate, carvedilol) with a minimal role for other drugs including antiarrhythmics (Al‐Khatib 2017). In addition, defibrillation therapy has been proven to be highly efficient in improving survival of patients with VA (Epstein 2012). To date, the bulk of evidence from high‐quality randomized controlled trials (RCTs) for several patient populations has supported transvenous implantation of ICDs.

ICDs are a recognized therapy that has been widely used for preventing sudden death from VA for more than 30 years. ICDs monitor the heart and provide electrical energy to the heart during episodes of ventricular arrhythmias, thus ending the life‐threatening event and resetting to sinus rhythm. They have been efficacious in reducing mortality in people with tachycardia, fibrillation, and other arrhythmic anomalies (Kuck 2000; Young 2003; Josephson 2004). The very first guidelines that discussed the use of ICDs in clinical practice were based on the Antiarrhythmics Versus Implantable Defibrillators (AVID) study, a large, multicentre trial that showed a reduction in all‐cause mortality in cardiac arrest survivors (AVID Investigators 1997). Hence, a Class I recommendation was established by the American College of Cardiology (ACC)/American Heart Association (AHA) for the use of ICDs in secondary prevention in people meeting enrolment criteria for AVID (Poole 2014).

Since their introduction in the early 1990s, conventional ICDs initially depended on transvenous leads for detecting cardiac signals and defibrillation (Yee 1990). For more than 30 years, transvenous ICD (TV‐ICD) systems have been the most reliable therapy. Significant data from high‐quality RCTs advocate its utilisation in different patient populations, which include patients with ventricular tachycardia, survivors of cardiac arrest, patients with structural heart disease and patients with significant left ventricle dysfunction. TV‐ICDs can also detect bradycardia pacing in addition to antitachycardia pacing, which can end various VTs painlessly (Al‐Khatib 2017). Besides their lifesaving capacities, transvenous sensing and defibrillation leads carry both infective and mechanical complications (Burke 2015). The leads in TV‐ICDs are usually inserted in the subclavicular area under fluoroscopy guidance requiring venous access, which sometimes results in complications (Yee 1992). Lead‐related complications of inserting transvenous lead(s) can be pneumothorax, haemothorax, and cardiac tamponade. In addition, systemic infections can cause harmful side effects and even lead to increased morbidity and deaths (Bardy 2010).

To reduce some of the complications of inserting transvenous leads a new subcutaneous, extra‐thoracic implantable defibrillator system designed to avoid the need for venous access was recently introduced (Bardy 2010). The subcutaneous cardioverter‐defibrillator (S‐ICD) consists of a single lead placed on the sternum, which removes the requirement to place the lead in or on the heart; this simplifies the implant procedure by utilising anatomical landmarks instead of fluoroscopy imaging, which accordingly produces less exposure to radiation (Bardy 2010). There may also be lower risk of infections like infective endocarditis because of the extravascular position of the S‐ICD (Weiss 2013). Lead failure, which can be either lead dislodgement or lead fracture resulting from mechanical stress associated with heart motion, body motion, and patient anatomy, is still a major obstacle in the use of TV‐ICDs over a long duration. Lead failure can either cause inappropriate shocks or result in impeding the necessary therapy. Since the S‐ICD lead is not exposed to significant mechanical stress, the procedure can increase lead longevity. A reduction of lead extraction procedures — often indicated because of lead fractures or infections and usually associated with substantial morbidity and mortality — is another advantage offered by S‐ICDs (Burke 2015). Finally, the S‐ICD has displayed a strong level of specificity in identifying ventricular arrhythmias. Misidentification of supraventricular arrhythmias or noise can lead to inappropriate shock therapy. These inappropriate shocks can substantially decrease quality of life (QoL) (Burke 2015).

There are limitations to the various designs of the S‐ICD system. Unlike the TV‐ICD, the S‐ICD is not able to provide bradycardia pacing and antitachycardia pacing (ATP) because of the lack of an endocardial lead except from 30 seconds on‐demand post‐shock pacing. Hence, in cases where chronic pacing capabilities are needed, the S‐ICD is not appropriate for patients that need either bradycardia, anti‐tachycardia (ATP), or cardiac resynchronization pacing (Al‐Khatib 2017). Consequently, the absence of bradycardia pacing in the S‐ICD might lead to more bradycardia‐related events such as syncope or even death.

Nonrandomised studies have shown the effectiveness of S‐ICDs with the ability to successfully detect and convert VF during defibrillation threshold and terminate spontaneous sustained VT that may happen after follow‐up. In the EFFORTLESS (Evaluation of Factors Impacting Clinical Outcome and Cost Effectiveness of the S‐ICD) registry, 472 enrolled patients reliably showed that the complication‐free rate was 94% at 360 days but also reported a 13.1% inappropriate shock rate at 3 years' follow‐up (Lambiase 2014). In another study of 314 patients, the 180‐day complication‐free rate was 99%, and the success of VF termination with first shock was more than 90% (Weiss 2013). All spontaneous episodes of VT/VF recorded in 21 patients (6.7%) were successfully converted and there were no lead failures, endocarditis or bacteraemia, tamponade, cardiac perforation, pneumothorax, or haemothorax associated with the S‐ICD.

At present, it is not clear if the positive characteristics of the S‐ICD offset its disadvantages. Although the S‐ICD is perceived to be equally appropriate as the TV‐ICD in numerous clinical scenarios, there is still a significant inconsistency in S‐ICD usage due to lack of literature and absence of strong experience with the new device. Furthermore, the S‐ICD was accepted for use on the basis of prospective trials in the absence of control groups. Moreover, no randomised trials comparing S‐ICDs with TV‐ICDs have been performed, and only a small number of retrospective and case‐control studies that have presented comparisons of efficacy and complications of recipients. Therefore, to establish the S‐ICD’s role and identify it as an adjunctive or primary therapy in patients who are at risk of SCD, an RCT is required that compares the S‐ICD to the TV‐ICD in patients in primary and secondary preventions.

How the intervention might work

All patients will be on beta‐blockers which have an anti‐arrhythmic effect. Beta‐adrenoceptors are coupled to Gs proteins which activate adenylyl cyclase to form cAMP from ATP. Increased cAMP activates a cAMP‐dependent protein kinase (PK‐A) that phosphorylates L‐type calcium channels, which causes increased calcium entry into the cell. Increased calcium entry during action potentials leads to enhanced release of calcium by the sarcoplasmic reticulum in the heart. PK‐A also phosphorylates sites on the sarcoplasmic reticulum, which lead to enhanced release of calcium through the ryanodine receptors. Beta blockers would therefore decrease intracellular calcium. In addition, beta blockers such as propranolol block myocardial fast sodium channels. ICDs work through a different method by ending the chain reaction that is formed by the abnormal pacemaker cells. During episodes of arrhythmia, the heart rate is dangerously fast but can be instantly corrected by both transvenous and subcutaneous ICDs, through delivering a shock to the heart to restore a normal heartbeat, thereby stopping the progression of the arrhythmia (Yee 1990).

A transvenous ICD (TV‐ICD) is typically inserted in a pre‐pectoral approach: the generator is located in the subcutaneous tissue of the upper chest, superficial to the pectoralis major muscle. The transvenous ICD lead is inserted via a vein (subclavian, axillary or cephalic) into the right ventricle and across the tricuspid valve. Based on the condition of the heart, either one or two leads will be inserted in the heart for either pacing impulses or defibrillating shocks to the heart as necessary. To achieve ideal connectivity, the leads are attached to the heart wall once they are set in place. Issues with access to the heart through the vascular system and recurrent issues with transvenous leads called for the development of an S‐ICD. Unlike a TV‐ICD device, the lead of the pulse generator in an S‐ICD is implanted subcutaneously above the breastbone (Lambiase 2014). Both ICDs aim to end serious ventricular arrhythmias and reduce SCD or arrhythmia‐related complications. Thus, a decrease in sudden deaths from ventricular arrhythmias is expected despite differing implantation (Bardy 2010).

Why it is important to do this review

Numerous studies have assessed safety and efficacy of S‐ICDs compared to TV‐ICDs (Burke 2015; Quast 2018). It has been observed that the S‐ICD is a new alternative that minimised intravascular lead complications compared to the conventional transvenous ICD system. Recently published meta‐analyses have shown that lead‐related problems decreased with S‐ICD implantation; but S‐ICD was comparable to TV‐ICD for non‐lead‐related complications and inappropriate therapy. In summary, the data showed that S‐ICD is associated with decreasing lead‐related complications and ICD shocks (Basu‐Ray 2017; Chen 2019). None of them, however, described the long‐term performance of the S‐ICD regarding its impact on quality of life and associated costs.

The ACC and European Society of Cardiology (ESC) guidelines give Class IA recommendation for S‐ICD implantation, without the requirement of bradycardia, anti‐tachycardia pacing (ATP), or cardiac resynchronization therapy, for managing patients with ventricular arrhythmias; and Class IIa recommendation for S‐ICD implantation for the prevention of sudden cardiac death (Priori 2015; Al‐Khatib 2017). Approximately 55% of patients in routine clinical practice needing an ICD are possible candidates for a subcutaneous device (Al‐Khatib 2017). To maximise clinical outcome and cost/benefit ratio, it is fundamental to choose candidates that can benefit the most.

There is still not enough clinical data with a longer follow‐up for clinical implementation of the S‐ICD that shows improvement of survival or prevention of sudden death in a manner that is similar to traditional TV‐ICD. Two large clinical trials currently in progress aim to determine the potential use of S‐ICDs as an alternative method for certain populations or as a preferred strategy. The PRAETORIAN trial will be the first randomised study that compares both devices 'head to head' (NCT01296022), with first results expected to be available in 2020 (Olde 2012). The ATLAS S‐ICD trial, which initiated enrolment in 2017, will also compare single‐chamber TV‐ICDs with S‐ICDs (NCT02881255).

Results of higher‐quality evidence, similar to the pending results of the above‐mentioned large randomised controlled trials, might appear to be inconsistent with current guidelines and recent published systematic reviews. Hence, we intend to assess the benefits and harms of S‐ICDs compared to TV‐ICDs in patients at risk of sudden cardiac death and evaluate certainty of the evidence by means of a Cochrane Review of randomised controlled trials.

Objectives

To evaluate the benefits and harms of using a transvenous implantable cardioverter‐defibrillator (TV‐ICD) compared to a subcutaneous implantable cardioverter‐defibrillator (S‐ICD) in patients at risk of sudden cardiac death.

Methods

Criteria for considering studies for this review

Types of studies

We will include only randomised controlled trials addressing the question of interest, irrespective of the language of publication. We will include studies published as full texts, as abstracts only, and unpublished data. Cluster RCTs are not eligible as they do not support the nature of this clinical intervention. Cross‐over trials are not feasible and are considered not ethical in this setting.

Types of participants

We will include studies of adults (aged 18 years and older) who are eligible for an ICD implantation for primary or secondary prevention related to being at risk of developing ventricular arrhythmias and sudden cardiac death (SCD). Individuals at risk include those with prior MI and heart failure due to either coronary artery disease or non‐ischaemic dilated cardiomyopathy, hypertrophic cardiomyopathy, congenital heart disease, family history of SCD, or genetic arrhythmia syndrome (including the long‐ and short‐QT syndromes, Brugada syndrome, idiopathic VF, and catecholaminergic polymorphic VT).

For studies that include only a subset of eligible participants we will try to obtain the data for the relevant subset only. If that fails, we will include studies with a maximum of 25% of ineligible participants. We will explore the effect of including such studies with a sensitivity analysis.

Types of interventions

The intervention of interest is subcutaneous implantable cardioverter defibrillator (S‐ICD), either single‐ or dual‐chamber ICDs, in addition to optimal medical theapy.

The control of interest is transvenous implantable cardioverter defibrillator (TV‐ICD), either single‐or dual‐chamber ICDs, in addition to optimal medical therapy.

We will include trials that compare subcutaneous implantable cardioverter defibrillator (S‐ICDs ‒ intervention group) to transvenous implantable cardioverter defibrillator (TV‐ICD ‒ control). We will include studies where standard care (optimal medical therapy) was given to participants equally in both the intervention and control group.

We plan to conduct a stratified/subgroup analysis to explore whether any heterogeneity is explained by whether the prevention is primary versus secondary; if yes, we will pool primary and secondary prevention trials.

Types of outcome measures

Reporting one or more of the outcomes listed here will not be an inclusion criterion of the review. Where a published report does not appear to report one of these outcomes, we will access the trial protocol and contact the trial authors to ascertain whether the outcomes were measured but not reported. We will include in the review, as part of the narrative, relevant trials which measured these outcomes but did not report the data at all, or not in a usable format.

We will include studies with a short‐term follow‐up of less than 24 months and long‐term follow‐up time of at least 24 months, since we can obtain sufficient and reliable results for the mortality‐related outcomes of interest (e.g. SCD, all‐cause mortality).

Mortality may be measured and reported as both dichotomous and as time‐to‐event. We will aim to collect and analyse both types of measurements. We will abstract the log (hazard ratio) and its variance from trial reports for time‐to‐event survival data to evaluate whether survival time is longer in one of the devices. Also, we will extract the raw data necessary to calculate the risk ratio to assess whether mortality is lower with one type of device.

As trials might have included different QoL instruments, we will include any of those as long as they have been validated. For each one of the included instruments, we will identify their minimal clinically important difference (MCID) from the literature. We will aim to analyse these findings according to the way they are reported (whether continuous or dichotomous). The primary analysis will be that of the continuous outcomes.

We plan to analyse adverse events as dichotomous outcome, since we are interested in the number of people with at least one of the adverse events. We also plan to report on the number of individual adverse events per treatment arm.

We plan to assess cost‐effectiveness narratively by quality‐adjusted life‐year (QALY).

We plan to analyse data for both short‐term (less than 24 months) and long‐term (24 months and more) follow‐up.

Primary outcomes

• Sudden cardiac death (SCD)

• All‐cause mortality

• Adverse events of using TV‐ICD or S‐ICD, including non‐lead‐related complications (device infection, haematoma, pneumothorax, pericardial effusion), inappropriate shocks, device failure; and lead‐related complications

Secondary outcomes

• Cardiovascular mortality

• Health‐related quality of life

• Hospital re‐admissions

• Cost‐effectiveness

Search methods for identification of studies

Electronic searches

We will identify trials through systematic searches of the following bibliographic databases.

• Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library

• MEDLINE (Ovid, from 1946 to present)

• Embase (Ovid, from 1974 to present)

The preliminary search strategy for MEDLINE (Ovid) (Appendix 1) will be adapted for use in the other databases. We will apply the Cochrane sensitivity‐ and precision‐maximising RCT filter to MEDLINE (Ovid) (Lefebvre 2011); and an adaptation of it to Embase.

We will search all databases from their inception to the present, and we will impose no restriction on language of publication or publication status.

We will also conduct a search of ClinicalTrials.gov (www.ClinicalTrials.gov), the WHO International Clinical Trials Registry Platform (ICTRP) Search Portal (apps.who.int/trialsearch) and ISRCTN registry (www.isrctn.com) for ongoing or unpublished trials.

We will use the following keywords when searching the clinical trials registers: ‘subcutaneous ICD’; ‘transvenous ICD’; ‘conventional ICD’; ‘dual‐chamber ICD’; ‘single‐chamber ICD’; ‘out‐of‐hospital cardiac arrest’; ‘sudden cardiac death’.

We will also search for regulatory data from EMA (www.ema.europa.eu/ema) and FDA (www.fda.gov/Drugs/InformationOnDrugs).

We will use the European Association for Grey Literature Exploitation (EAGLE) and National Technical Information Service (NTIS) databases to search the grey literature.

Searching other resources

To identify additional studies, we will check reference lists of all included studies and any relevant systematic reviews for additional references. We will also examine any relevant retraction statements and errata for included studies. We will contact authors for missing data and ongoing trials; and search relevant manufacturers’ websites for trial information. These websites include:

• www.medtronic.com;

• www.abbott.com; and

• www.bostonscientific.com.

Data collection and analysis

Selection of studies

Four review authors (GI, MC, MS, AK) will work in pairs to screen the titles and abstracts of all studies found by the searches, in duplicate and independently. We will obtain the full texts of studies included by at least one of two reviewers. Pairs of reviewers will then screen the full texts for eligibility in duplicate and independently according to the eligibility criteria. They will use a standardized and pilot‐tested full‐text screening form and conduct calibration exercises. They will compare results and resolve any disagreement by discussion; and if there is no consensus, they will enlist the help of an arbitrator (MR). We will identify and exclude duplicates and collate multiple reports of the same study so that each study rather than each report is the unit of interest in the review. We will record the selection process in sufficient detail to complete a PRISMA flow diagram and ‘Characteristics of excluded studies’ table (Liberati 2009).

Data extraction and management

Four review authors (GI, MC, MS, AK) will work in pairs to extract data, in duplicate and independently. They will use a standardized and pilot‐tested data collection form and conduct calibration exercises. They will compare results and resolve any disagreement by discussion; and if there is no consensus, they will call upon another author (MR) to arbitrate.

We will extract the following characteristics and data from included studies.

• Methods: study design; total duration of study; number of study centres and location; study setting; date of study.

• Participants: number of patients randomised; lost to follow‐up/withdrawn; analysed; general characteristics (age, gender, LVEF, New York Heart Association (NYHA) class, diagnosis, primary prevention, secondary prevention, aetiology of heart disease, renal function).

• Interventions: intervention; comparison; concomitant medications.

• Outcomes: primary and secondary outcomes; time points.

• Notes: funding for trial; notable conflicts of interest of trial authors; ethical approval.

One review author (GI) will transfer data into the Review Manager 5 (RevMan 5) file (Review Manager 2014). We will double‐check that data is entered correctly by comparing the data presented in the systematic review with the study reports. A second review author (MC) will spot‐check study characteristics for accuracy against the trial report.

Assessment of risk of bias in included studies

Four review authors (GI, MC, MS, AK) will work in pairs independently to assess risk of bias for each study, in duplicate and independently. They will resolve any disagreements by discussion or by consulting another author (MR).

For RCTs, we will use the Cochrane ‘Risk of bias’ tool outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). The tool addresses the following.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessment

• Incomplete outcome data

• Selective outcome reporting

• Other bias

We will grade each potential source of bias as high, low or unclear and provide a quote from the study report together with a justification for our judgement in the ‘Risk of bias’ table.

We will summarize the risk of bias judgements across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the ‘Risk of bias’ table.

When considering treatment effects, we will take into account the risk of bias for the studies that contribute to that outcome (Egger 1997).

Assessment of bias in conducting the systematic review

We will conduct the review according to this published protocol and report any deviations from it in the ‘Differences between protocol and review’ section of the systematic review.

Measures of treatment effect

We will analyse dichotomous data as risk ratios (RRs) and time‐to‐event data as hazard ratios (HRs), with 95% confidence intervals (CIs). When data are available for both RR and HR, we will run analyses for the two effect measures. For continuous data, we plan to use the mean difference (MD) when pooling data from studies that used the same tool to assess the outcome of interest, and the standardised mean difference (SMD) when pooling data from studies that used different tools. SMD will be interpreted as follows: SMD less than 0.40 for small variations; SMD between 0.40 and 0.70 for moderate variations; and SMD greater than 0.70 for large variations. We will narratively describe skewed data reported as medians and interquartile ranges.

Unit of analysis issues

For studies with repeated outcome measurements, we plan to analyse all data at short‐term (< 24 months) and long‐term (≥ 24 months) follow‐up. We will perform a separate subgroup analysis of studies reporting the longest follow‐up data, divided into two subgroups (< 24 months vs ≥ 24 months).

In case of multi‐arm studies, a way to overcome a unit‐of‐analysis error for a study that could contribute multiple, correlated comparisons is to combine all relevant experimental intervention groups of the study into a single group, and to combine all relevant control intervention groups into a single control group if possible. A study comparing the intervention to multiple‐arm control would be included in the meta‐analysis by combining the participants of the intervention group and the combined control group in the usual way. For dichotomous outcomes, both the sample sizes and the numbers of people with events can be summed across groups. For continuous outcomes, means and standard deviations can be combined (Higgins 2019).

Dealing with missing data

We will contact investigators in order to verify key study characteristics and obtain missing outcome data where possible (e.g. when we identify a study as abstract only). Where possible, we will use the RevMan 5 calculator to calculate missing standard deviations using other data from the trial, such as CIs. Where this is not possible, and the missing data are thought to introduce serious bias, we will explore the impact of including such studies in the overall assessment of results by a sensitivity analysis (Akl 2013; Ebrahim 2013; Ebrahim 2014; Guyatt 2017).

Assessment of heterogeneity

Clinical heterogeneity

We will assess clinical heterogeneity by comparing study characteristics including participants, experimental intervention, control intervention, outcome measurements and time points of outcome measurements.

Statistical heterogeneity

We will inspect forest plots visually to consider the direction and magnitude of effects and the degree of overlap between CIs. We will use the I² statistic to measure statistical heterogeneity among the trials in each analysis but acknowledge that there is substantial uncertainty in the value of I² when there are only a few studies. We will also consider the P value (P = 0.10) from the Chi² test. We will interpret the I² statistic using the following criteria (Higgins 2017).

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity;

• 50% to 90%: may represent substantial heterogeneity;

• 75% to 100%: considerable heterogeneity.

If there is substantial or considerable heterogeneity across studies, we will investigate potential causes via prespecified subgroup analysis (see ‘Subgroup analysis and investigation of heterogeneity’ below).

Assessment of reporting biases

If we are able to pool more than 10 studies, we will create and examine an inverted funnel plot to explore possible study biases for the primary outcomes. If not, the power of tests becomes too low to distinguish chance variation from real asymmetry (Higgins 2017).

Data synthesis

We will conduct a random‐effects model to pool treatment effects considering heterogeneous populations and different delivery of interventions. We will present all results with 95% CIs. We will conduct all analysis according to the guidance provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017); and implement it in RevMan 5 software (Review Manager 2014).

We may not do meta‐analysis due to considerable heterogeneity that we cannot explain, or outcomes reported differently among studies. In this setting, we will narratively summarize the main findings and results of those included studies.

'Summary of findings' table

We will create a ‘Summary of findings’ table using the following outcomes: sudden cardiac death; all‐cause mortality; adverse events (device infection; haematoma; pneumothorax; pericardial effusion; inappropriate shocks; device failure; and lead‐related complications); cardiovascular mortality; health‐related quality of life; hospital re‐admissions; and cost‐effectiveness (Table 1).

1. 'Summary of findings' table ‒ draft.

S‐ICD vs. TV‐ICD for the treatment of sudden cardiac death
Patient or population: patients at risk of sudden cardiac death Setting: hospital and community Intervention: S‐ICD Comparison: TV‐ICD
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with control	Risk with treatment
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio; OR: Odds ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

We will rate the certainty of evidence following the GRADE approach, and its factors (study limitations; inconsistency; imprecision; indirectness; and publication bias as it relates to the prespecified outcomes) (Guyatt 2011a; Guyatt 2011b; Guyatt 2011c; Guyatt 2011d; Guyatt 2011e; Guyatt 2011f; Guyatt 2013a; Guyatt 2013b; Guyatt 2013c). We will use methods and recommendations described in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions using GRADEpro software (Schünemann 2019 and GRADEpro GDT respectively). We will justify all decisions to downgrade the certainty of evidence using footnotes.

Two review authors (GI, MC), working independently, will make judgements about certainty of evidence, with disagreements resolved by discussion or by involving a third author (EA). We will justify, document and incorporate judgements into reporting of results for each outcome.

Subgroup analysis and investigation of heterogeneity

We plan to carry out the following subgroup analyses to investigate heterogeneity.

• Mean age (< 65 years versus ≥ 65 years)

• Aetiology of heart diseases (ischaemic heart diseases versus non‐ischaemic cardiomyopathy versus inherited arrhythmia syndromes)

• Patients with left ventricular ejection fractions (LVEF) ≤ 35% versus > 35%.

• Gender (male versus female)

• Follow‐up duration (< 24 months vs ≥ 24 months) of studies’ subgroups.

We will conduct subgroup analyses for all primary and secondary outcomes.

We plan to conduct a subgroup analysis to explore whether any heterogeneity is explained by whether the prevention is primary versus secondary; if yes, we will pool primary and secondary prevention trials separately.

We will use the formal test for subgroup differences in Review Manager 5 (Review Manager 2014), and base our interpretation on this. If there is not sufficient information about prespecified subgroups in reports, we will try to contact the corresponding author and request the necessary data.

Sensitivity analysis

We plan to carry out the following sensitivity analyses, to test whether key methodological factors or decisions have affected the main result.

• Including only studies with a low risk of bias: we will assess for the overall risk of bias by low risk in at least four of the following five domains (random sequence; allocation concealment; blinding of outcome assessment; incomplete outcome data; and selective reporting).

• Excluding studies in which missing data cannot be adjusted for.

• Excluding trials with more than 25% of ineligible participants if we fail to acquire data related to the subset of eligible participants.

Reaching conclusions

We will base our conclusions only on findings from the quantitative or narrative synthesis of included studies for this review. We will avoid making recommendations for practice and our implications for research will suggest priorities for future research and outline what the remaining uncertainties are in the area.

History

Protocol first published: Issue 5, 2020

Notes

Glossary of terms

Implantable Cardioverter Defibrillator (ICD) system: a device (also called a pulse generator) and leads. An ICD system is implanted to monitor your heart rhythm and help treat dangerously fast or slow arrhythmias. Transvenous: implanted through a vein. Subcutaneous: implanted under the skin. Defibrillation: procedure in which a fast heart rate (i.e. ventricular fibrillation) is restored to a normal rhythm by delivering an electrical shock. Ventricular fibrillation (VF): a very fast, irregular heart rhythm caused by abnormal electrical signals starting from several areas of the ventricle. In VF, the ventricle beats so fast that it pumps very little blood to the body. A heart in VF may beat more than 300 beats per minute. Without immediate medical attention, VF can be fatal. Defibrillation is the only way to treat VF once it occurs. Ventricular tachycardia (VT): a fast rhythm caused by abnormal electrical signals coming from the ventricle. The rapid rate of 120 to 250 beats per minute may produce dizziness, weakness and eventual unconsciousness. VT may progress to ventricular fibrillation.

Appendices

Appendix 1. Preliminary MEDLINE (Ovid) search strategy

1 Defibrillators, Implantable/ (15767)
2 (defibrillator* adj5 implant*).tw. (14964)
3 ((ICD or ICDs) adj5 (subcutaneous or transvenous or implant*)).tw. (7413)
4 (S‐ICD or S‐ICDs).tw. (305)
5 (TV‐ICD or TV‐ICDs).tw. (42)
6 Electric Countershock/ (14465)
7 electric countershock*.tw. (72)
8 electric defibrillat*.tw. (101)
9 (electroversion adj3 (therap* or cardiac)).tw. (4)
10 (cardiover* or cardioconver*).tw. (18680)
11 (resynch* adj3 (therap* or treatment* or device*)).tw. (7325)
12 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 (40454)
13 randomized controlled trial.pt. (491031)
14 controlled clinical trial.pt. (93308)
15 randomized.ab. (456348)
16 placebo.ab. (201309)
17 clinical trials as topic.sh. (188647)
18 randomly.ab. (319177)
19 trial.ti. (205641)
20 13 or 14 or 15 or 16 or 17 or 18 or 19 (1242145)
21 exp animals/ not humans.sh. (4625829)
22 20 not 21 (1142411)
23 12 and 22 (4401)

Appendix 2. Abbreviations

•S‐ICD: subcutaneous implantable cardioverter‐defibrillator
•TV‐ICD: transvenous implantable cardioverter‐defibrillator
•VA: ventricular arrhythmia
•SCD: sudden cardiac death
•SCA: sudden cardiac arrest
•MI: myocardial infarction
•NICM: non‐ischaemic cardiomyopathy
•DCM: dilated cardiomyopathy
•HCM: hypertrophic cardiomyopathy
•ICD: implantable cardioverter defibrillator
•LVED: left ventricle end‐diastolic
•LVEF: left ventricle ejection fraction
•NSVT: non‐sustained ventricular tachycardia
•NYHA: New York Heart Association
•PTCA: percutaneous transluminal coronary angioplasty
•PVC: premature ventricular contractions
•VF: ventricular fibrillation
•VT: ventricular tachycardia

Contributions of authors

Ghida Iskandarani conceived the review protocol.

Mohamad Sabra helped draft the review protocol.

Minsi Cai helped draft the review protocol.

Assem Khamis helped draft the review protocol.

Elie A Akl helped draft and revised the review protocol.

Marwan Refaat helped draft and revised the review protocol.

Sources of support

Internal sources

• No sources of support supplied

External sources

• This project was supported by the National Institute for Health Research (NIHR), via Cochrane Infrastructure, Cochrane Programme Grant or Cochrane Incentive funding to the Heart Group. The views and opinions expressed herein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, the National Health Service or the Department of Health, UK

Declarations of interest

Ghida Iskandarani has no conflict of interest to declare.

Mohamad Sabra has no conflict of interest to declare.

Minsi Cai has no conflict of interest to declare.

Elie A Akl has no conflict of interest to declare.

Marwan Refaat has no conflict of interest to declare.

Assem Khamis has no conflict of interest to declare.

New