In the absence of testing, ICD optimisation should always be done


Mark Kroll
Mark Kroll

There is a trend away from testing implantable cardioverter defibrillators (ICDs) at implant and this has been supported by two clinical trials, SIMPLE and NORDIC ICD.1,2 While these trials have significant limitations in terms of study design and statistical power, a thorough discussion of them goes beyond the scope of this commentary. Some clinicians have misinterpreted the results of these studies to believe that optimisation is no longer required. Such confusion could affect patient survival and that is the focus here; writes Mark W Kroll for Cardiac Rhythm News. He spoke on the subject at the Heart Rhythm Congress (4-7 October, Birmingham, UK).


Cathodic stimulation is used in pacemakers, as cathodes are very efficient at launching wavefronts. The goal of a defibrillation shock is the opposite; its goal is to kill wavefronts. The use of an anodic right ventricular coil lowers the defibrillation threshold (DFT) by 18%. Cathodes can defibrillate but with a 22% penalty in required energy.3 This simple scientific issue has been clouded by the fact that some manufactures have yet to update the default polarity of their ICDs and, ironically, refer to the anodal polarity as “reverse” polarity. This unfortunate situation arose from the early work on transvenous ICD systems when we mistakenly copied the cathodal polarity of transvenous pacemakers. Many major institutions solve this problem by directing that the polarity will always be set to anodal before the implant.

The polarity was not standardised nor reported in the SIMPLE study; it was simply set according to the normal practice of the centre. To be conservative, one must assume that the vast majority of implants were done with the polarity corrected away from the default cathodal polarity of the device used. Failing to correct a cathodal polarity more than doubles (2.3x) the proportion of high DFTs.4

Single vs. dual-coil leads

Defibrillation is optimised when the first phase of a biphasic shock has a duration of 3.5-5.0ms.5 (The duration of the second phase is not as critical, at least with anodal polarity shocks.) With a dual-coil hot-can system, a typical system impedance is 40Ω and the ICD capacitor is drained relatively rapidly.3 To keep the phase duration within the 3.5-5.0ms window, the capacitor voltage is optimally allowed to decay to 40% of its initial voltage. This waveform is used by two manufacturers and referred to as a “60% tilt.” Another two manufacturers have set their first phase tilt at 50% and are thus optimised for single coil leads. With a typical 60Ω resistance with a single coil system, this lower tilt will usually place the first phase duration in the 3.5-5.0ms window. A 50% tilt was used in the SCD-HEFT trial, and dual-coil lead patients had twice (2x) the first shock failure rate of the single-coil patients.6 One manufacturer, St Jude Medical, allows the direct programming of the phase durations independently of the system resistance and lead system.

In summary, 60% tilt is optimised for dual-coil implants while 50% tilt is optimised for single-coil systems. The penalties for ignoring this can be severe. Gabriels et al reported on different waveforms used with single-coil patients.7 In the tilt-based waveform arm the majority (2/3) of the patients had a 60% tilt waveform. This resulted in twice (2x) as many patients having unacceptable DFTs and increased the DFT by 29% (compared to optimal phase durations). (Optimal waveforms defibrillate with 22% less energy.)

When a dual-coil lead is used, the SVC coil should be kept out of the right atrium. Blood is the best conductor in the human body (besides urine) and a low SVC coil will preferentially draw current from the right ventricular coil and thus “steal” current from the ventricles. An innominate junction location will reduce the DFT by an average of 30% compared to a right atrial location which generally provides no benefit (except for high tilt waveforms which require the lower resistance).8 At least one manufacturer provides leads with different SVC coil spacings; this can be used to advantage regardless of the model of ICD implanted.


Table 1 shows the potential penalty from committing the two tempting implant sins: (1) leaving a default cathodal polarity, and (2) using a single coil lead with a high-tilt device. The result is a loss of nearly half of the stored energy (of the SIMPLE device) compared to an optimised implant in terms of equivalent DFTs.

ICD implants do not provide as much benefit as we like to believe. In secondary-prevention ICD patients, 42% of cardiac death is still sudden.9 The ICD patients with residual sudden cardiac death tend to be younger, with higher ejection fractions and with better NYHA functional class.10 Ironically, these “healthier” patients may be the very ones that are tempting to skip testing with.

Previously, system optimisation was performed when an acceptable safety margin could not be obtained. While the DFT could be different with a spontaneous ventricular fibrillation, at least the clinician knew that the ICD performed with electrically induced ventricular fibrillation. It is important to appreciate that without testing, optimisation should always be done.11 We now know enough about the science of defibrillation that simple good practice rules can be given as shown in Table 2. Optimisation is far more important in the absence of testing and should never be confused with testing itself.


Healey JS et al, Lancet 2015;385(9970):785-791

Bansch D et al, Europace 2015;17(1):142-147

Kroll MW, Swerdlow CD, J Interv Card Electrophysiol 2007;18(3):247-263

Baccillieri MS et al, J Interv Card Electrophysiol 2015;43(1):45-54

Kroll MW, Schwab JO, Fundam Clin Pharmacol 2010;24(5):561-573

Aoukar PS et al, HeartRhythm 2013;10(7):970-976

Gabriels J et al, Cardiology 2013;124(2):71-75

Gold M et al, Circulation 2006; 114:II_690

Anderson KP, J Interv Card Electrophysiol 2005;14(2):71-78

Mitchell LB et al, J Am Coll Cardiol 2002;39(8):1323-1328

Gold MR et al, Pacing Clin Electrophysiol 2009;32(5):567-569

Mark W Kroll is adjunct professor, Biomedical Engineering, University of Minnesota Minneapolis, USA. He has worked for Medtronic and St Jude Medical and has received speaking honoraria
from St Jude Medical