EHRA expert consensus statement and practical guide on optimal implantation technique for conventional pacemakers and implantable cardioverter-defibrillators: endorsed by the Heart Rhythm Society (HRS), the Asia Pacific Heart Rhythm Society (APHRS), and the Latin-American Heart Rhythm Society (LAHRS)

Abstract

With the global increase in device implantations, there is a growing need to train physicians to implant pacemakers and implantable cardioverter-defibrillators. Although there are international recommendations for device indications and programming, there is no consensus to date regarding implantation technique. This document is founded on a systematic literature search and review, and on consensus from an international task force. It aims to fill the gap by setting standards for device implantation.

Introduction

Scope of the document

According to the 2017 European Heart Rhythm Association (EHRA) white book, pacemaker (PM) implantations have increased by 20% and implantable cardioverter-defibrillator (ICD) by 44% over a 10-year period in the European Society of Cardiology (ESC) member countries.¹ Therefore, there is a growing need to train device implanters. Proper training in implantation technique results in improved patient outcome, as demonstrated by lower rates of in-hospital^2,3 and long-term⁴ complications in procedures performed by electrophysiologists compared to other specialties. The core curriculum for the heart rhythm specialist⁵ and the EHRA certification exam provide a framework for required training duration, procedural volume, and knowledge assessment, but do not specifically address proficiency in device implantation technique. An EHRA survey on preferred tools and techniques for implantation of cardiac electronic devices showed considerable heterogeneity in practices across Europe.⁶ There is, therefore, an unmet need to define best practice for device implantation. The aim of this document is to provide up-to-date recommendations for optimal implantation technique of standard PMs and ICDs, based upon published evidence and expert consensus, in order to provide guidance for training and clinical practice, and ultimately to improve patient outcome. Specific domains, such as leadless pacing, conduction system pacing, cardiac resynchronization therapy (CRT), sub-cutaneous ICDs, and device therapy in children and patients with congenital heart disease are not covered here.

Methodology

This EHRA consensus document is based on a systematic literature search and review. M.B. searched PubMed, Embase, Web of Science, and Cochrane electronic databases for full-text original articles and case reports in English language published between 1 January 2000 and 20 March 2020 relating to transvenous PM and ICD implantation. Additional references were hand-searched and retrieved by checking reference lists by H.B. and C.S. From 4108 retrieved articles uploaded to the Covidence platform (Melbourne, Australia), 477 articles were finally selected by H.B. and C.S. (for details see Supplementary material online, Appendix).

Recommendations are based upon strength of evidence and consensus as outlined in Table 1. All recommendations were subjected to voting and required at least 75% agreement to reach consensus for green and red hearts.

General considerations

Preoperative preparation

As a first step, it is important to re-evaluate the indication and to perform a risk-benefit analysis for every patient. This requires a full medical history, review of laboratory results and imaging studies, and knowing current guidelines. A list of elements to consider is shown in Table 2.

As a prerequisite for cardiac implantable electronic device (CIED) implantation the patient should be free of active infection and afebrile for >24 h. Antibiotic prophylaxis should be given in all cases (e.g. cephazolin 1–2 g or flucloxacillin 1–2 g within 1 h of incision, or, in case of allergy or high probability of resistant pathogens, vancomycin 15 mg/kg within 90–120 min). Specific measures to prevent CIED infection are covered in detail in a recent EHRA consensus document.⁷

Patients on anticoagulation and/or antiplatelet drugs require special considerations, as pocket haematoma increases the risk for infection.^7,8 In patients with non-rheumatic atrial fibrillation (AF) and a low risk for thrombo-embolic events (CHA₂DS₂VASc score < 3) perioperative interruption of anticoagulation is an option.⁹ In all other patients, uninterrupted oral anticoagulation (aiming for international normalized ratio <3.0 or <3.5 in case of a mechanical valve prosthesis) is preferable to heparin-bridging (especially if using low molecular weight heparin), as it reduces the risk of haematoma.^9–11 In the BRUISE-CONTROL 2 study,¹¹ patients on a direct oral anticoagulant with a CHA₂DS₂VASc score of ≥2 had no difference in haematoma with or without interruption of therapy (doses including the morning dose were taken in the latter group). The conclusion of the study is that either management strategy is reasonable and should be based upon clinical judgement. Perioperative management of anticoagulants and/or antiplatelet drugs¹² is shown in Table 3.

Operative environment

The operative room/catheterization or electrophysiology laboratory must be ventilated with >15 air changes/h (ideally 20–25 changes/h).^13,14 The utility of laminar airflow on surgical site infections remains controversial.^14,15 Operators should ensure that all required hardware and equipment is available before the intervention (see Table 4). Checklists and surgical ‘time-out’ are useful to avoid errors. Anaesthesia support should be readily available on the premises due to potentially life-threatening complications, to assist if necessary with analagesia or haemodynamic/respiratory support.

Staff, training, and qualifications

Team members should be proficient with sterile techniques and practice. Poor adherence to these standards results in higher rates of surgical site infections.¹⁶ The operator should have EHRA level 2 or national certification, equivalent expertise, or be adequately supervised if in training. Procedural volume is also important to maintain expertise; operators or centres with <50 implantations/year have been shown to have higher complication rates.^17,18

In addition to the operator, at least one non-scrubbed nurse or allied professional proficient with supporting device implantation and procedural sedation should be present in the room, but the presence of unnecessary personnel should be avoided. All personnel should be trained with radioprotection and the operator should strive to minimize radiation exposure.^19–21 The operator, or another physician present on site, should be able to perform emergency pericardiocentesis and cardiopulmonary resuscitation manoeuvres.

Specific operative steps

A checklist covering the different operative steps is provided in the Supplementary material online, Appendix, as are links to online video tutorials, which provide practical tips on performing selected techniques. Measures to prevent infection have been recently covered in a consensus document⁷ and will not be repeated here. Surgical hand preparation with antimicrobial soap and water or alcohol-based hand rubs should be performed according to the World Health Organization global guidelines for prevention of surgical site infection (see Supplementary material online, Appendix).¹⁴

Left-sided implants are usually favoured due to this being the non-dominant side [and also due to lower defibrillation thresholds (DFT) for ICDs]. The decision should be individualized and account for specific requirements (see Table 2).

Incision and pocket creation

After prepping the skin, local anaesthesia is mostly performed using 1% lidocaine (maximum dose 4.5 mg/kg or 300 mg) which may also be used semi-diluted to administer a larger volume over a greater area,²² or mixed with a long-acting local anaesthetic e.g. bupivacaine 0.25% or ropivacaine 0.5%. The use of epinephrine-containing local anaesthetics has been associated with a higher incidence of haematomas and should be avoided.²³ Deep sedation or general anaesthesia may be required in agitated patients, or in specific instances (e.g. lead tunnelization or submuscular pockets). A horizontal incision inferior and parallel to the clavicle, or an oblique incision slightly medial or along the deltopectoral groove may be performed according to operator preference, and have shown to yield similar scar healing (see Supplementary material online, Video S1).²⁴ The oblique incision may provide better access for cephalic vein cutdown, but being relatively lateral, the pocket should be prepared more medially with enough spacing to avoid conflict with the shoulder.

Many operators prefer to create the pocket at the beginning of the procedure for new implants, as local anaesthesia may be more effective at this stage, and bleeding can be monitored during the intervention. It is important to perform dissection down to the subfascial plane, even when a generous amount of adipose tissue is present (see Supplementary material online, Video S2). With the relatively small size of current generators, the majority of operators chose to implant PMs and ICDs in a subcutaneous (subfascial prepectoral) pocket. A submuscular pocket may be considered in thin patients who are at risk for pocket erosion, for cosmetic reasons, or in case of Twiddler’s syndrome. It is important to respect anatomical planes (i.e. between the pectoralis major and minor muscles) and not to create an intramuscular pocket to avoid undue tissue damage and bleeding (see Supplementary material online, Video S3). Submammary^25,26 and axillary²⁷ pockets have also been performed for cosmetic reasons but are rarely used.

Venous access

Due to variations in patient anatomy and for the reasons outlined below, implanters should be proficient with cephalic venous cutdown, axillary and subclavian vein puncture. It is useful to place an ipsilateral intravenous line in case a venogram is required; this does not increase the risk of infection.²⁸ The use of micropuncture kits allows vascular access with a small 21 gauge needle as opposed to the standard method, which requires an 18 gauge needle and can be used to decrease severity of complications due to inadvertent arterial or pleural puncture.

Cephalic venous cutdown

The cephalic vein is the first approach in 60% of centres according to the EHRA survey.⁶ After distal ligation, the vein may be incised followed by direct introduction of the lead (see Figure 1) or a puncture may be performed with introduction of a sheath (see Supplementary material online, Video S4). In case of difficulty inserting the lead or if more than one lead is implanted, using a guidewire and introducer sheath is helpful.

Cephalic vein access has the advantage of avoiding pneumothorax and reducing the risk of lead dysfunction compared to subclavian puncture [odds ratio (OR) 0.25, 95% confidence interval (CI) 0.13–0.51, P < 0.001].²⁹ Successful cannulation is reported in approximately 60–80% of patients,^30–33 but may reach >90% with the help of hydrophilic guidewires or use of retro-pectoral veins.^34–36 Hydrophilic angulated guidewires are useful for crossing venous valves and to steer access towards the subclavian vein when the lead tracks towards the arm (asking the patient to lift the shoulder while steering the lead or wire can also help). Introduction of up to three leads using introducer sheaths is possible in the great majority of patients, but in rare instances, a separate puncture for venous access is required.³⁶ Bleeding complications are not significantly different compared to subclavian access.^29,37 Supra-clavicular course of the vein is a rare variant that should be recognized by fluoroscopy and by palpating the lead over the clavicle.^38,39 In this rare instance, alternative routes should be used to avoid discomfort or lead complications.

Intra-thoracic subclavian vein puncture

Subclavian puncture was initially popular due to its high success rate (approximately 95%^32,40,41) but carries a risk of pneumothorax (approximately 1–2% of patients^40,42,43) and lead failure due to subclavian crush (see Figure 2).^{29,32,44–46} Friction with the clavicle, subclavius muscle or costo-clavicular ligament may also restrict lead manipulation and positioning. Subclavian access has been associated with a higher risk of bleeding complications in patients receiving antiplatelet drugs.⁴⁷ Other rare complications include arterio-venous fistula,⁴⁸ transient phrenic nerve palsy due to local anaesthesia,⁴⁹ and thoracic duct injury. For these reasons, it is recommended that intrathoracic subclavian vein puncture is not used as a first-line approach, but as a bailout technique in case other routes have failed, or for venous access medial to occlusions.⁵⁰ The puncture should be as lateral as possible to avoid lead crush (see Supplementary material online, Video S5). It may be performed using only bony landmarks (below and slightly lateral to the most prominent part of the clavicle for the puncture site, directing the needle towards the sternal notch) or under fluoroscopy aiming towards the middle of the clavicular head. A venogram may be useful to target the vein and reduce the risk of complications. For dual-chamber devices, separate punctures may reduce the risk of bleeding and of lead dysfunction due to crowding, but this has not been proven and may expose the patient to increased risk of pneumothorax compared to a single puncture, especially in case of difficult access or risky anatomy [e.g. low body mass index (BMI) or chronic obstructive pulmonary disease].

Extra-thoracic subclavian and axillary vein puncture

The axillary vein becomes the extra-thoracic subclavian vein as it crosses over the inferior border of the first rib. These segments are considered analogous for the purposes of this document. Axillary vein puncture is being increasingly adopted following the introduction of new techniques which have improved its safety and efficacy, as well as evidence of superior success rates compared to cephalic venous cutdown^30,32,33 and improved outcome compared to subclavian vein puncture, especially with regard to reducing the risk of lead failure.^32,40,46 The risk of bleeding is not different compared to cephalic access.³⁷

A variety of techniques have been described using anatomic landmarks,^51–53 a venogram,^30,32,54,55 a guidewire from the antecubital vein as a roadmap,⁵⁵ fluoroscopic bony landmarks (most often the outer border of the first rib)^{31,33,54,56–58} or ultrasound,^41,59–64 with approximately 95% success. A useful technique is to access the axillary vein using a 35° caudal fluoroscopic view and to aim the region overlying the outer border of the first rib (see Figure 3 and Supplementary material online, Video S6), which has been shown to be successful in 96% of cases.⁵⁸ This exposes the subclavian space and the anterior outline of the lungs, thereby eliminating the risk of pneumothorax if the technique is properly performed. The same view may be used with a venogram if the vein is not accessed using fluoroscopic landmarks only. A shallow approach using this technique also reduces X-ray exposure to the hands as well as flexion stress on the lead (which may be an issue with a steep angle for ‘walking’ the puncture needle on the first rib).

Ultrasonography is likely to be increasingly adopted as it reduces X-ray exposure to the operator (see Supplementary material online, Video S7), has a self-learned success rate of 97.7%,⁶⁵ and its intra-operative use is facilitated by the advent of wireless devices.⁶⁰ It also allows to concomitantly perform pectoral nerve block.⁶⁶

In some cases, narrowing of the axillary vein may occur due to compression by haematoma induced by inadvertent arterial puncture or bleeding from the muscle.

Troubleshooting venous access

In patients previously implanted with leads, venous stenosis of >75% is found in 6–21% of patients and vein occlusion in 6–26%.^67–69 Superficial collateral veins may be visible on the patient’s chest, but do not necessarily indicate obstruction,⁷⁰ and most patients have no complaints. In patients requiring an upgrade, it is very useful to perform a venogram prior to skin incision to confirm patency using the same view as for venous access (e.g. 35°caudal tilt) to plan the procedure. Peripheral venography may however overestimate occlusion with collaterals, as more selective dye injection via a dilator can reveal flow through the lesion in approximately two-thirds of these cases.⁷¹ Hydrophilic 0.035″ guidewires, hydrophilic catheters, and long sheaths are very useful to cross stenoses. Hydrophilic wires should be handled carefully to avoid perforation. More medial punctures^50,72,73 or venoplasty^69,71,74 may be performed for occlusions. For upgrades, contralateral venous access with subcutaneous tunnelization to the pocket may be performed, e.g. with a trochar and a chest tube.⁷⁵ Other more invasive options are lead extraction and the ‘inside out’ technique⁷⁶ which need to be performed in specialized centres.

In patients with pre-existing leads which are to be abandoned or explanted, the retained wire under lead insulation technique can be used to gain venous access if the old lead is able to be pushed a few centimetres into the vein.⁷⁷

Alternative routes for venous access as a bailout solution include the internal/external jugular vein,⁷⁸ and iliofemoral access.^79,80 Leadless PMs, subcutaneous ICDs, or epicardial leads are options which should be considered in case of venous access issues. Lead extraction may also be an option to gain venous access (either using extraction sheaths via a superior approach, or a femoral approach with a retained guidewire in the lead lumen).

The azygous vein may be unintentionally cannulated by the guidewire or the lead. This can be recognized by the posterior course of the vein just above the right main bronchus (Figure 4).

A persistent left superior vena cava (PLSVC) is encountered in up to 0.5% of patients.⁸¹ This structure usually regresses to become the Marshall vein/ligament, and drains into the coronary sinus. Ventricular lead placement can be challenging but can be facilitated by shaping the stylet with a large J curve (Figure 5). It is useful to search for an innominate vein which is sometimes present (but may be small) and can facilitate implantation from the usual route. Right-sided implant may be necessary, but a right superior vena cava may be absent in a minority of cases. Right sided venography is advisable to check for presence of a right-sided superior vena cava, before attempting to switch sides, should a PLSVC be identified.

Right ventricular lead placement

The ventricular lead is usually placed before the atrial lead (unless AAI pacing is considered) in order to provide backup pacing and to avoid dislodgment. Anatomic sites for right ventricular pacing are shown in Figure 6. According to the 2013 EHRA survey,⁶ half of the centres declared the right ventricular apex (RVA) as their preferred lead position, followed by the interventricular septum and the outflow tract in 47% and 3% of centres, respectively. A manually curved stylet is often used to cross the tricuspid valve (it is also possible to prolapse the lead off the right atrial wall into the ventricle by manipulating the stylet). Inadvertent cannulation of the coronary sinus or its branches is suggested by ‘tracking’ of the lead and absence of ventricular premature beats, and can be confirmed in the left anterior oblique (LAO) view.

Right ventricular apex

Leads have traditionally been placed at the RVA due to ease of positioning and presence of trabeculae for anchoring passive leads. Concern has arisen, however, regarding long-term deleterious effects of RVA pacing on left ventricular function, heart failure and death.^82,83 In addition, it has been shown that perforation is independently associated with an RVA position for active-fixation leads (OR 3.37, 95% CI 1.17–9.67, P = 0.024).⁸⁴

As a general rule, with extendible helix leads, fluoroscopic markers should be used for confirming deployment (and not number of turns). Over-rotation may result in lead destabilization, perforation, or damage to the lead. This writing group recommends a number of practical precautions to avoid complications when implanting any type of lead (see Table 5).

The lead may be positioned slightly proximal to the apex so as to point in a slightly inferior orientation. The right anterior oblique (RAO) 30° view is useful to expose the long axis of the heart and better outline the apex (see Figure 7). The LAO 40–60° view should also be checked to ensure that the lead is not inadvertently placed in a coronary sinus tributary or in the left ventricle via a septal defect (or arterial puncture), as this can be missed in the postero-anterior view.^85,86

Lead stability should be checked by withdrawing the stylet to the level of the inlet and buckling the lead by pushing it. Sufficient slack should be left to form a ‘heel’. Temporary pacing at maximum output should be performed while palpating or observing under fluoroscopy the left hemi-diaphragm to rule out phrenic capture. Implantation technique is shown in Supplementary material online, Video S8.

Right ventricular septum

The right ventricular septum (RVS) has been advocated as a more physiological alternative to the RVA to reduce adverse effects of pacing-induced dyssynchrony, with equivocal results.^82,83 Right ventricular systolic function⁸⁷ and tricuspid regurgitation^87,88 are also similar to apical pacing. Nevertheless, an advantage is that perforation may be avoided. Electrical parameters are comparable to apical pacing.^89,90

A caveat is that many reports have shown that leads intended for the RVS are often positioned on the anterior free wall or antero-septal groove (which forms a natural recess into which the lead wedges).^91–95 This site may possibly result in left ventricular dysfunction,^92,93 perforation,⁸⁹ and in rare instances in myocardial infarction, due to the lead being screwed into the left anterior descending artery.^96–98 It is therefore important to target a true septal position.

Studies which have validated final lead position by computed tomography (CT) or echocardiography have demonstrated that the LAO view alone is insufficient to confirm septal pacing and that an RAO view to visualize the long axis of the heart is also necessary.^91,94,95 Likewise, the angle of orientation of the lead in the LAO view is not indicative of a septal location,^95,99 as it may simply be modified by adjusting lead slack.¹⁰⁰ Furthermore, the LAO 40° may not adequately visualize the heart in its short axis, which is often better represented by the LAO 60° view.¹⁰¹

A fluoroscopic landmark which can be used for positioning septal leads is shown in Figure 8.⁹¹

Ventriculography in the RAO 20–30° view can accurately depict the contour of the right ventricle in its long axis for precise targeting of the RVS.⁹⁹ Although this may not be practical to perform on a routine basis, it may be useful in patients with dilated or hypertrophied ventricles, which can alter the cardiac silhouette. One should also evaluate lead tip movement, which is more ample with fixation to the free wall compared to the RVS.

A number of electrocardiographic (ECG) criteria for confirming lead position have been proposed, but are limited by improper validation of the pacing site,^102,103 or lack of comparison to anteroseptal/anterior free wall pacing.¹⁰⁴ All studies which have used echocardiography,¹⁰⁵ CT scans,^94,95 or electro-anatomical mapping¹⁰⁶ for validating pacing sites have shown that ECG criteria are unreliable indicators of RVS pacing.

The lead may slide on the RVS using a two-dimensional curved stylet. In order to facilitate positioning the lead, a posterior distal bend can be made (see Figure 9) to shape a three-dimensional (3D) stylet as initially described by Mond et al.^91,107

Positioning may either be done using the ‘pullback’ technique from the pulmonary artery (see Supplementary material online, Video S9) or directly on the RVS. By combining the 3D stylet and the fluoroscopic landmarks described above in the RAO 30° and LAO 40–60° views, around 95% of the leads can be correctly placed on the RVS.^91,94 The target area is the septomarginal trabeculation/moderator band (see Figure 6), which form a natural buttress to stabilize the lead and also harbours the right bundle branch¹⁰⁸ and may also help preserve electrical synchrony.

Only active fixation leads should be used for RVS pacing as this allows positioning in smooth-walled regions. If the lead is not stable, it will usually drop with stylet withdrawal. It is important to adjust lead slack to avoid rocking of the lead tip, which may result in dislodgement (as will too little slack, see Supplementary material online, Video S9).

Right ventricular outflow tract pacing

The same stylet shape and fluoroscopic views (RAO/LAO) can be used for implanting leads on the right ventricular outflow tract (RVOT) septum as for the RVS. This may however be more challenging, with heterogeneous lead positions,^109–111 due to the smaller target area of this funnel-shaped structure. Furthermore, the RVOT septum is free-standing, except for its inferior part, i.e. without direct contact with the interventricular septum,¹¹² and it is thin-walled, tapering from 3–5 mm down to 1–2 mm in its subvalvular segment,¹⁰⁸ raising concern for perforation. There is little evidence that this pacing site is beneficial, although long-term lead performance has been shown to be satisfactory.^113,114

Right atrial lead placement

Right atrial leads with dual-chamber devices have been consistently associated with a 1.5–2× increased risk of complications compared to single chamber VVI-systems.^17,115–120 The main issues are lead dislodgement and perforation.¹²¹ Unless the patient has a VDD system (and does not require atrial pacing), an atrial lead is necessary to maintain atrioventricular synchrony and to confirm atrial arrhythmias. An atrial lead also greatly facilitates follow-up of arrhythmic episodes with ICDs, and tended to reduce inappropriate shocks in a meta-analysis of six trials randomizing single- vs. dual-chamber ICDs (OR 1.46, 95% CI 0.97–2.19; P = 0.07), without, however, impacting mortality.¹²²

Right atrial appendage

The right atrial appendage (RAA) comprises the entire trabeculated anterolateral triangular part of the right atrium. It is demarcated posteriorly by the crista terminalis, which separates it from the smooth-walled vestibule. Wall thickness is 1–2 mm, with the lateral appendage being particularly thin (see Figure 6); histology shows one or few myocytes thickness between pectinate muscles with some areas being devoid of myocardium.^123,124 One should therefore avoid placing leads on the lateral wall of the RAA and right atrium to avoid perforation, which may result in tamponade, or in rare instances, right-sided pneumothorax and pneumopericardium.^125–128 The tip of the RAA points leftwards to lie over the aortic root,¹²⁹ which can result in life-threatening laceration of the aorta by active-fixation leads.^{124,130–133} It is therefore probably advisable to place the right atrial lead in the anterior RAA (see Figure 6).

Patients may complain of parasternal pain during implantation, which may indicate perforation despite normal electrical parameters (and is confirmed if the pain disappears after repositioning the lead). Passive J-shaped leads have been shown to have a lower risk of pericardial complications than active leads,^121,134 but may be slightly more prone to dislodgment.^134–136 Active J-shaped leads do not confer any major advantage compared to straight leads.^137,138

The medial RAA is also in proximity to the RVOT,¹²⁴ which can result in far-field R-wave (FFRW) oversensing. This can lead to inappropriate mode switching in >20% of patients,¹³⁹ to proarrhythmia with atrial antitachycardia pacing,¹⁴⁰ or issues with ICD rhythm discrimination algorithms. FFRW amplitude should therefore be carefully assessed at implantation (ideally before deploying the helix as it is not affected by fixation) and should be <20% compared to the near-field atrial electrogram. This is best accomplished by direct visualization of the intracardiac electrogram, rather than relying solely on a digital reading of the P-wave amplitude. FFRW amplitude can be reduced by placing the lead in the anterior RAA (pointing towards the operator in a postero-anterior view—see Supplementary material online, Video S10). It is useful to check lead position in the RAO/LAO views, as the ‘windscreen wiper’ movement may be more visible than in the postero-anterior projection. After fixating the lead, stability should be verified by withdrawing the stylet to the superior vena cava and pushing the lead. Sufficient slack should be left to prevent straightening of the lead, as this may result in suboptimal electrical parameters. It should be borne in mind that the amount of lead slack can significantly change near- and far-field sensing amplitude due to changes in atrial dipole orientation with respect to the activation vector (both in the frontal and horizontal planes).

Alternative pacing sites

In an attempt to reduce AF, alternative pacing sites have been studied. These include Bachmann’s bundle, the coronary sinus ostium area, and even dual site atrial pacing. Meta-analyses have failed to show any advantage of progression to persistent/permanent AF or lead-related complications compared to RAA pacing.^141,142 The interatrial septum may avoid perforation of the free wall, but proximity to the aortic root with risk of perforation should be borne in mind.¹³¹ Due to these considerations and the added complexity of lead placement, this writing group does not recommend atrial septal pacing as a first-line approach. It may however be useful as an option in case of issues with RAA lead placement (e.g. scarring after cannulation for cardiac surgery).

Electrical testing

Acute capture thresholds should be ≤1.5 [email protected] ms, with sensing amplitudes ≥1.5 mV for the atrium and ≥4mV for the ventricle, and lead impedances should be within normal specified limits (usually 400–1200 Ohms). It is important to not simply record measurements, but to observe waveforms (e.g. presence of FFRW on atrial leads, as mentioned above). A negative electrogram can result from reversal of the crocodile clips on the lead¹⁴³ (with elevated capture thresholds) or may indicate lead perforation.¹⁴⁴ With active-fixation leads it is important to confirm presence of current of injury (COI) (see Figure 10).

Some degree of ST elevation on the intracardiac electrogram can be encountered even before helix deployment, simply due to contact with the myocardium. To indicate adequate fixation, the elevation should increase compared to baseline and be clearly visible.^145–148 Capture thresholds can be high immediately after fixation with a COI, but usually decrease over the following minutes (making it worthwhile to recheck thresholds rather than immediately repositioning the lead). Sensing may also be affected by COI and should be rechecked. The significance of slew rate for adequate lead fixation is uncertain.

Electrical testing is usually straightforward for ventricular leads but may be challenging for atrial leads (see practical tips in the Supplementary material online, Appendix).

Lead and generator fixation

Haemostatic sutures using resorbable braided sutures around lead insertion sites are useful to control bleeding. However, care should be taken not to tie directly to the lead body as this may damage the insulation (see Supplementary material online, Video S11). The suture sleeve should be pushed up to the puncture site and non-absorbable braided suture used to secure the lead to the muscular plane, coaxial to lead insertion to avoid kinking. It is important that the sutures be made directly to muscle including the fascia and not to subcutaneous fat. In addition to being friable, the subcutaneous plane is mobile and may result in lead dislodgment in obese patients (particularly in women with large breasts which exert a downward pull and shift this plane). It is important to use an anchor knot technique (i.e. first tie a non-slip knot on the muscle and transverse to fibre orientation) under the sleeve and then tie a loop in the groove of the suture sleeve (see Figure 11). This is the technique which offers most traction resistance¹⁴⁹ and also prevents dislodgement from ‘ratchet’ or ‘reel’ syndrome^150–152 resulting from loosening of the tie due to tissue shrinkage if the muscle and sleeve are taken together. A second suture may be placed in the same manner or simply around the sleeve to secure it to the lead. Sutures should never be placed directly on the lead body and should not be too tight as this may cause lead damage. Two or more moderately tight ties will secure the lead better than one very tight tie, with less risk of lead damage. Stability should be checked by gentle traction on the lead.

Due to differences in auto-initialization methods among manufacturers, it is recommended to manually interrogate devices before implantation.¹⁵³ This also allows to establish wireless communication which is useful to check proper lead connection and parameters. The lead pins should be wiped of blood with a dry swab before being connected to avoid future ‘freezing’ in the header.

The pocket may be irrigated with saline to flush out blood clots and debris. Excess lead length should be coiled in the pocket, with care being taken to avoid kinking (which may result in lead damage), and placed under the generator to avoid inadvertent damage when the pocket is reopened. The header should ideally be oriented towards the incision (see Figure 12, left), as this will facilitate access to the leads during generator change. Fixation of the generator in the pocket is optional (including for ICDs).

Pocket and skin closure

The pocket should be closed by separate stitches (2–4 usually suffice) using resorbable braided suture (see Figure 12, right and Supplementary material online, Video S12). This avoids migration of the generator and the leads, as well as reduces surface wound tension in case of pocket haematoma. There are several ways of closing the skin. For example, a running stitch using resorbable 3–0 braided suture can close the subcutaneous plane followed by a subcuticular running stitch using 4–0 monofilament absorbable suture. Barbed sutures obviate the need for knots, but have not shown any significant advantage compared to standard sutures,¹⁵⁴ and are more costly. Skin glue has been compared to sutures in two randomized studies and was found either to have no significant advantage¹⁵⁵ or more early adverse events (9.3% vs. 6.0%; P = 0.02).¹⁵⁶ Some operators use staples, but this has not been evaluated against sutures for PM and ICD implantation, and requires that the patient return to have them removed.

Management of perioperative complications

Although CIED implantation has some potentially fatal complications, procedure-related death is exceptionally rare (0–0.1%).^17,157–159 Perioperative mortality is mainly due to comorbidities (e.g. heart failure). Perioperative complications are listed in Table 6.

Electronic health records or registries which capture complications are useful to keep track of these events for benchmarking or audits.

Perforation and pericarditis/tamponade

The incidence of clinically relevant lead perforation is difficult to determine due to variable definitions, but has been reported to range from 0.09% to 1.5%.^{17,84,120,135,160–165} Lead perforation usually manifests itself as an acute (<24 h) or sub-acute (<1 month) complication, but may rarely occur late, and even be diagnosed years after implantation.^166,167 The consequences of lead perforation may be acute pericarditis, pericardial effusion or tamponade,^{17,84,120,160–162} constrictive pericarditis,¹⁶⁸ pleural effusion, haemothorax,¹⁶⁹ pneumothorax,^125,127,170 lung perforation,¹⁷¹ abnormal electrical parameters, and diaphragmatic/intercostal stimulation,¹⁷² but it may also occur without clinical manifestations.¹⁷³

Intra-operative perforation can result in vagal symptoms, chest pain, high capture thresholds, diaphragmatic capture, and inversion of the COI¹⁴⁴ (rarely observed). Tamponade is strongly suggested by haemodynamic compromise (the differential diagnosis being a vagal reaction) and an immobile cardiac silhouette with fluoroscopy (the latter is sometimes present without effusion). Tamponade should be confirmed by echocardiography and treated with emergent pericardiocentesis. In a study including 968 patients who had an echocardiogram performed routinely before and <24 h after CIED implantation, a new mild pericardial effusion (≤10 mm in diastole) was found in as many as 8.3% of patients, with moderate (11–20 mm) and large (>20 mm) effusions in 0.4% and 1.5% of patients, respectively.¹⁶² Effusions are asymptomatic in 94% of patients,¹⁶² and therefore go unnoticed in most cases. If pain due to pericardial irritation is present, anti-inflammatory drugs can be administered. Patients with mild or moderate effusions should be monitored closely with continued surveillance after discharge to rule out recurrence/worsening of effusion. In case of haemodynamic compromise or large effusions, pericardiocentesis should be performed,¹⁶² and considered in those with moderate effusions which do not regress quickly, especially if the patient requires antiplatelet drugs and/or anticoagulation. Whether patients who require pericardiocentesis should undergo lead revision is debated. In a retrospective series of 48 patients with definitive perforations, 10 patients had pericardiocentesis for tamponade and were treated conservatively without lead revision, with recurrence of effusion over follow-up in only one patient.¹⁷⁴ However, conservative management of perforation was associated overall with increased complications, mainly tamponade over follow-up in patients who initially had no/mild effusion (most of whom were on antiplatelet drugs and anticoagulants).¹⁷⁴ Lead revision should therefore be carefully evaluated in patients with perforation. Most cases of lead repositioning/replacement are uneventful.¹⁷⁴ However, perforations which are overtly transmural should be treated in centres with cardiac-thoracic surgical standby. The increased risk of infection due to early re-intervention may be mitigated by the use of an antibacterial envelope.¹⁷⁵

Risk factors for lead perforation are shown in Table 7. Although several studies have reported increased incidence of perforation with active fixation leads,^163,164,176 a recent large study showed no difference compared to passive leads.¹⁶¹ Nevertheless, passive atrial J-leads^121,134 or VDD systems (in case atrial pacing is not necessary) can be considered in patients deemed to be at high risk of perforation, although atrial undersensing may be observed in about 10% of patients.¹⁷⁷

Arrhythmias

Audible pulse signals from continuous ECG monitoring during an implant procedure are useful for immediately identifying arrhythmias. Transient traumatic atrioventricular block can occur during right ventricular lead positioning in patients with underlying left bundle branch block. The implanter should anticipate this complication and have cables ready to provide rescue pacing (even prior to lead fixation). Backup transcutaneous pacing should also be available. The block usually resolves within a few minutes, but in some cases may persist for several hours.

Non-sustained ventricular tachycardia is frequent during right ventricular lead placement, and especially during lead manipulation and pullback of the lead from the RVOT. In rare instances, sustained ventricular tachycardia or ventricular fibrillation is induced and may require overdrive pacing or cardioversion/defibrillation. Atrial flutter/fibrillation may also be induced during lead placement.

Preparing pre-positioned defibrillation pads on the patient is strongly recommended to prevent disrupting the sterile field to apply paddles emergently for defibrillation or transcutaneous pacing.

Analysis of arrhythmias is limited by absence of a 12-lead ECG for most standard procedures. Entrainment manoeuvres performed via the pacing system analyser (PSA) can be used to diagnose 1:1 atrioventricular arrhythmias as in the electrophysiology laboratory.

Pneumothorax

The incidence of pneumothorax ranges between 0.4% and 2.8%, and is dependent on the venous access chosen for implantation.^{17,42,43,135,160,178–180} Pneumothorax should be suspected if air is aspirated while advancing the needle, although this is not always the case. Fluoroscopy may visualize the pneumothorax if it is large. A routine chest X-ray should be performed within 24 h in all patients. In case of suspected pneumothorax, a chest X-ray should be performed immediately and repeated after several hours or the next day, as the complication may not be initially visible. A CT scan may also be performed if the X-ray is negative but the complication is suspected.¹⁷⁰ In case of a small apical pneumothorax, conservative management may often suffice, with surveillance until resolution of the pneumothorax. In most other instances, a chest tube should be inserted.

Risk factors for the development of a pneumothorax include: age >80 years, female sex, low BMI, chronic obstructive pulmonary disease, and subclavian vein puncture.^42,43,178 Cephalic vein cutdown or axillary vein puncture⁵⁸ should be preferred to avoid this complication, and the use of a ‘Micro-Puncture’ needle can minimize damage to the lung.

Pocket haematoma

The incidence of pocket haematoma ranges between 0.2% and 16.0% depending on definition and factors such as anticoagulation regimen.^{10,11,17,118,119,181} It is associated with an approximately nine-fold increased risk of infection.⁷

Conservative management should be favoured if possible due to 15-fold risk of infection with reintervention.²⁸ However, in cases of wound dehiscence or skin erosion, severe pain, arm swelling, brachial plexus or arterial compromise, a surgical revision should be performed without delay. The procedure should be performed with the strictest precautionary measures by an experienced implanter and use of an antibacterial envelope should be considered. Needle aspiration is contra-indicated as it provides incomplete evacuation and may result in infection by seeding the pocket.

Pocket haematoma should be avoided by optimal perioperative management of anticoagulation and antiplatelet drugs (see Section Preoperative preparation and Table 3). In addition, good surgical technique with minimal tissue trauma, respecting anatomical planes and with meticulous attention paid to haemostasis should be the norm. Haemostatic resorbable sutures may be placed at venous entry sites (non-resorbable sutures may hamper future lead extraction, if required). Use of a suction drain has been described,¹⁸² but is controversial due to concern for infection. There is little evidence that haemostatic agents are effective.¹⁸³ Compression using sandbags, tapes or vests^184–186 can be useful to avoid haematoma.

Local and systemic CIED infection

CIED infection is a serious and potentially life-threatening complication. The incidence is reported to be 0.6–3.4%.⁷ Detailed description on diagnosis, treatment and prevention can be found in a recent EHRA consensus document.⁷

Lead dislodgement

Lead dislodgment is reported in 1.2–3.3% of implantations.^{3,45,135,160,180,187,188} Most dislodgments occur before discharge.¹⁸⁷ The incidence is significantly higher for atrial leads than for right ventricular leads.^45,180,187 There is a trend towards fewer dislodgments with active fixation leads compared to passive leads.¹³⁵ Diagnosis is usually confirmed by the chest X-ray after abnormal electrical testing, but micro-dislodgments may not be apparent. Dislodgement usually requires revision, but may be unnecessary if lead function is not compromised or the lead is not absolutely required and there are no adverse effects, such as arrhythmia or lead chatter. If repositioning is necessary, it should be deferred if possible for a few weeks to minimize pain and risk of infection.⁷

Other complications

Inadvertent arterial puncture should be identified before insertion of the sheath by observing pulsatile flow, bright red blood, or guidewire position. Simple compression is usually sufficient, but a vascular closure device or surgical revision may be necessary if the sheath was introduced.¹⁸⁹

Air embolism can cause haemodynamic compromise and is potentially lethal.^190,191 It is favoured by snoring or deep breathing and can be avoided by the operator temporarily blocking the sheath with a finger and introducing the lead during a breath-hold/expiration, or by using sheaths with haemostatic valves. Fluoroscopy shows air in the right ventricle and pulmonary artery. Patients may be asymptomatic or complain of dyspnoea. Patients should be placed in a Trendelenburg position, administered 100% oxygen, and if necessary, the air should be aspirated (e.g. with a pigtail or Judkins right diagnostic catheter).

Pneumopericardium may occur as a consequence of atrial lead perforation with right-sided pneumothorax.¹⁹² The diagnosis may be missed on the chest X-ray and may require a CT scan.¹⁷⁰ The issue resolves with drainage of the pneumothorax.

Acute deep vein thrombosis of the access vein may rarely occur within days of implantation,¹⁹³ or after months¹⁹⁴ or years and may in rare instances be associated with pulmonary embolism. The symptoms usually resolve with administration of heparin, followed by oral anticoagulation.

Postoperative management

A chest X-ray (postero-anterior and lateral) should be performed within 24 h in all patients after lead implantation to rule out pneumothorax and document lead position. A 12-lead ECG should be recorded and the device fully checked before discharge.

Patients may be mobilized freely once they have recovered from sedation, and same-day discharge is feasible in selected patients.¹⁹⁵ There is no evidence that limitation of arm movement after implantation avoids dislodgment, as securing by fibrosis requires months. A study randomizing standard care (restriction of abduction and lifting of weights for 6 weeks) vs. shoulder exercises, found significantly less shoulder pain at 1 month in the latter group (33% vs. 5%, P = 0.02).¹⁹⁶ For this reason, use of arm slings should be avoided.

The patient should be informed regarding post-operative care, ideally with written instructions. The wound should be properly covered by a dressing for 2–10 days. The patient may shower if he has a waterproof dressing, or otherwise after about a week and clean the upper body with a moist towel in the meantime. Routine wound inspection is not superior to patient-initiated consultation.¹⁹⁷ Patients should be seen in-office at 1–3 months, as delaying to >12 weeks is associated with adverse outcome.¹⁹⁸ Remote monitoring can be useful for early detection of technical issues.¹⁹⁹

Specific considerations for ICD implantation

Implantation of ICDs follows the same steps as for PMs However, specific aspects deserve consideration. Left-sided implantation should be favoured due to risk of increased DFT and total mortality with right-sided access (although co-morbidities requiring right-sided access, such as haemodialysis or cancer may have confounded outcome).^200–202 Defibrillation testing is already covered in the 2015 Heart Rhythm Society (HRS)/EHRA/Asia Pacific Heart Rhythm Society(APHRS)/SOLAECE expert consensus statement on optimal ICD programming and testing,²⁰³ without new data which requires modifying these recommendations.

Signal quality

Atrial and ventricular signal quality is of particular importance for rhythm discrimination algorithms to work properly. For example, atrial undersensing during a supra-ventricular tachycardia will fulfil the V > A criterion with misdiagnosis of ventricular tachycardia. Low R-wave amplitude may result in undersensing of ventricular arrhythmias and increases risk of T-wave oversensing due to automatic sensitivity gain levels starting as a percentage of the sensed amplitude. Sensed R-wave amplitude measured by the PSA may not correspond to that of the generator, due to differences in filter settings and regression of COI. It is therefore useful to check signal amplitude detected via the generator before closing the pocket.

The coil of integrated bipolar leads are part of the sensing circuit and records atrial signals in as many as 11% of patients if it lies close to the tricuspid annulus,²⁰⁴ with a risk of double-counting and inappropriate therapy. The ventricular electrogram should therefore be scrutinized for presence of low-amplitude atrial signals when implanting these leads.

As with any procedure involving abandoned leads, care should be taken to avoid chatter with pre-existing leads which can result in spurious arrhythmia detection or inhibition of pacing. This is particularly important with integrated bipolar leads which may be more prone to this phenomenon. In an in vitro setting, only metal-metal interaction resulted in chatter, which was prevented by expanded polytetrafluoroethylene coating of the coil.²⁰⁵ Coil coating also reduces fibrous adhesions which facilitate lead extraction.^206,207

It is advisable to implant ICD leads and generators from the same manufacturer to preserve labelled magnetic resonance imaging (MRI)-conditionality. In addition, manufacturer-specific signal processing and lead sensing may result in detection issues.²⁰⁸ Furthermore, elevation of subthreshold impedance values have been reported due to contact interface oxidation between lead pins and set screws that may be of different materials not tested for compatibility.²⁰⁹

Right ventricular lead position

Three studies with a total of 146 patients have randomized ICD lead position between the RVOT and RVA, and found no meaningful differences in DFT, with similar threshold measurements.^210–212 Two larger studies randomized RVS and RVA ICD lead position, and found no significant differences in outcome between groups. However, in the SPICE study⁹⁰ (which included 299 patients), the RVS group tended to have a higher rate of DFT > 25J (5.0% vs. 2.1%, P = 0.21) and a higher rate of ICD lead revisions (6.9% vs. 3.2%, P = 0.57). In the SEPTAL study,²¹³ which included 215 patients, the RVS group tended to have a higher incidence of total mortality (7.9% vs. 2.9%, P = NS). The main issue with these studies is that ‘mid-septal’ lead position was only verified in the LAO view, which is insufficient, as mentioned previously.^91,94,95 It is therefore likely that many leads were implanted in the anterior free wall and that a true septal position for ICD leads has not been properly validated. For these reasons, it may be preferable to implant ICD leads in the RVA, but alternative sites may be considered in case of suboptimal electrical parameters, or to avoid cardiac perforation associated with thin ICD leads¹²⁰ and certain patient characteristics (see Table 7).

Single-coil vs. dual-coil ICD leads

There has been a shift away from dual-coil ICD leads due to difficult and more risky lead extraction as a result of the proximal coil being fibrosed to the superior vena cava.^206,214 In a meta-analysis of 16 studies, there was a clinically irrelevant reduction in DFT with dual-coil leads (mean 0.81 J, 95% CI 0.31–1.30 J) and no difference in first shock efficacy.²¹⁵ Nevertheless, dual-coil leads may have a higher shock conversion rate of atrial arrhythmias (44.0% vs. 28.8%, P = 0.07 in the SIMPLE study²¹⁶) and offer additional morphology vectors for rhythm discrimination. Also, selected patients with right-sided ICD implants may benefit from dual-coil leads.²⁰⁰

DF-4 vs. DF-1 connector

The DF-4 standard is popular with implanting physicians as it facilitates lead connection. However, it offers fewer options compared to DF-1 leads, such as the requirement to implant a pacing lead when ‘downgrading’ from an ICD to a PM (e.g. during generator replacement in super-responders to CRT). There is also no possibility (without an adaptor) to implant stand-alone coils (e.g. in the azygos vein) or SQ-arrays in case of a high DFT, or to switch IS-1 connector pins with left ventricular leads during generator replacements in patients with functional recalled leads.^217,218

Specific considerations with generator replacement

The REPLACE registry¹¹⁹ provides prospective data in 1750 patients who underwent elective generator replacements. The rate of major adverse events at 6 months’ follow-up was 4.0% (generator replacement only) and 15.3% (with addition of a lead). Most complications were lead dislodgment or malfunction, haematoma and infection. Generator replacements are often performed by a junior physician alone, which is questionable given these high rates of adverse events.

The patient’s left ventricular ejection fraction should be known, as a generator replacement is an excellent opportunity for device upgrade (to an ICD or CRT^219–221) or if downgrade from CRT-D to CRT-P is considered in case of super-response or aging of the patient. Electrical parameters recorded over follow-up should be known, in order to compare them with current findings. In patients with very old leads, it is useful to check the connector standard of the lead (e.g. 5 mm or LV-1) and ensure that the required hardware is available. It is also advisable to determine if any leads are on recall or alert, as abandonment or extraction with replacement vs. continued use needs to be considered before the patient is scheduled for a procedure. Abandoned leads may either be capped or sectioned (care being taken to avoid extrusion of metallic components and also to avoid blood clogging the lumen, which may compromise future extraction). This may be achieved by pulling lead insulation over the sectioned tip or knuckling the extremity, and then fastening a tie. Abandonned leads should be fixated in the pocket to prevent them from migrating under the skin. In patients with a dual-chamber system who have developed permanent AF, alternatives are to either maintain a dual-chamber device to preserve MRI-conditionality, or to abandon the atrial lead.

Patients who are PM-dependent usually have an escape rhythm which emerges after some time spent in VVI 30 b.p.m. Otherwise isoprenaline (1–10 µg/min) will solve this issue in almost all cases and avoid need for temporary pacing.

It is recommended that an X-ray or high-resolution fluoroscopy be done prior to skin incision to evaluate lead damage, with particular attention being paid to the subclavian region, site of fixation sleeves, and along the leads, particularly in specific situations (e.g. conductor externalization of Riata leads). This will also yield information on presence of abandoned leads (which is relevant for MRI-conditionality) and position of the header and leads in the pocket. Independently of the initial incision, access should be performed about 1 cm lateral or superior from the current pocket (and never over the generator) to avoid damaging the leads and to provide a margin between skin closure and the pocket in case of pocket haematoma or superficial infection. The incision may be performed over the initial scar if this is desirable for cosmetic reasons, as long as the scar does not lie over the pocket.

When dissecting fibrotic tissue around the leads, electrocautery may be used in cut mode at low energy settings parallel to the course of the lead without prolonged applications at a fixed position. Improper use of electrocautery may result in thermal damage to the insulation, the risk of which can be reduced using a plasma surgery tool.²²² Electrocautery should be used with extreme caution over the generator (and should not come into contact with the metallic housing) or conductor coil in case of lead insulation failure, as it may induce ventricular arrhythmia.^223–225

Leads may be frozen to the header and can be freed using an orthopaedic drill^226,227 or scalpel²²⁸ (to access the tip from the rear of the header and push out the lead with the wrench), bone cutter²²⁹ or by injecting a solvent (pure ethanol) into the header with a needle.²³⁰ It is also possible to inject lubricant around the pin after removing the setscrew.²²⁹ It is important to avoid this issue by ensuring that lead pins are clean of blood before insertion in the header. The set screw may be frozen, which may require an Allen key to be unscrewed. If the set screw is stripped such that the wrench cannot provide the needed torque, it is sometimes possible to remove the silicone grommet and grasp and turn the set screw from the sides using a small clamp.

In some patients, lead damage may be visualized, and it may be difficult to implant a new lead (e.g. in case of vein obstruction or in a frail patient). It is useful in these instances to have repair kits with silicone sleeves and glue, and tools for lead splicing (e.g. Medtronic 5866–9 M Adaptor, which can be used for salvaging a fractured lead).^231,232

Conclusions

Tools and techniques for device implantation have evolved over the years, as has education in this field. Virtual reality simulators are increasingly used to train device implantation in a safe environment^233–235 and are useful adjuncts for education. This document provides a framework for standardizing PM and ICD implantation. Ultimately, by standardizing the procedure and avoiding poor techniques, the document should serve to improve patient outcome.

Consensus statement tables

Supplementary material

Supplementary material is available at Europace online.

Acknowledgements

The authors thank the EHRA Scientific Document Committee: Dr Nikolaos Dagres, Dr Serge Boveda, Dr Kevin Vernooy, Prof. Zbigniew Kalarus, Prof. Gulmira Kudaiberdieva, Dr Georges H Mairesse, Prof.Valentina Kutyifa, Prof. Thomas Deneke, Prof. Jesper Hastrup Svendsen, Dr Vassil B Traykov, Prof. Arthur Wilde, Prof. Frank R. Heinzel.

Conflict of interest: A.A. is a consultant to Boston Scientific, Backbeat, Biosense Webster, Cairdac, Corvia, Microport CRM, EPD-Philips, Radcliffe Publisher. He received speaker fees from Boston Scientific, Medtronic, and Microport. He participates in clinical trials sponsored by Boston Scientific, Medtronic, EPD-Philips. He has intellectual properties with Boston Scientific, Biosense Webster, and Microport CRM. H.B. has received speaker honoraria from Biotronik and Medtronic and has received institutional fellowship support and research grants from Abbott, Biotronic, Boston Scientific, Medtronic, and Microport. J.-C.D. declares honoraria for lectures, research and travel grants from Abbott, Biotronik, Boston Scientific, Medtronic, Microport. C.J.L. is a consultant to Medtronic and Philips Medical, and has received honoraria from Convatec and Abbott Medical, and research support from Boston Scientific. C.S. declares payment to his institution related to his activity as speaker fees, honoraria, consultancy, advisory board fees, investigator, committee member from Angiodynamics, Medtronic, Spectranetics, Biotronik, Liva Nova (Sorin) and Cook Medical and departmental or institutional research funding from Cook Medical. K.V. is consultant for Medtronic and Abbott. None of the other authors report any conflicts of interest.

References