Summary
Defibrillator shocks, appropriate or not, are associated with significant morbidity, as they decrease quality of life, can be involved in depression and anxiety, and are known to be proarrhythmic. Most recent data have even shown an association between shocks and overall mortality. As opposed to other defibrillator-related complications, the rate of inappropriate and unnecessary shocks can (and should) be decreased with adequate programming. This review focuses on the different programming strategies and tips available to reduce the rate of shocks in primary prevention patients with left ventricular dysfunction implanted with a defibrillator, as well as some of the manufacturers’ device specificities.
Résumé
Les chocs délivrés par les défibrillateurs automatiques implantables sont associés à une baisse de la qualité de vie des patients implantés (dépression, anxiété), et peuvent avoir un effet pro-arythmogène. Les études les plus récentes ont même montré un lien entre les chocs électriques internes et la mortalité totale. Contrairement aux autres complications, la fréquence de ces chocs peut (et doit) être diminuée grâce à une programmation adaptée. Cette revue de la littérature se focalise plus particulièrement sur les stratégies disponibles pour réduire la fréquence des chocs chez les patients implantés en prévention primaire pour une dysfonction systolique ventriculaire gauche, ainsi que sur les spécificités des appareils des différents constructeurs.
Background
Implantable cardioverter-defibrillators (ICDs) significantly reduce all-cause mortality in the primary prevention of sudden cardiac death in patients at increased risk, especially those with a reduced left ventricular ejection fraction (LVEF). The Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II) first established a strong mortality benefit for the ICD in patients with a LVEF ≤ 30% and a previous myocardial infarction . The reduction in total mortality was 31% in patients implanted with a prophylactic ICD compared with patients randomized to optimal medical therapy. This effect was durable, with a 34% reduction at 8 years by multivariable analyses; six patients had to be implanted to prevent one death . The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) confirmed these results in patients with a LVEF ≤ 35% and ischaemic or non-ischaemic cardiomyopathy , finally leading to Class I recommendations in the current guidelines .
For the past 10 years, as the number of prophylactic ICD implantations has drastically increased, so have all the related complications: acute implant-related complications in up to 5% of patients (especially pocket haematoma, haemothorax, pneumothorax, cardiac tamponade); an almost 1% rate of 30-day mortality; lead dislodgement; pocket infection and lead-related endocarditis; lead and device failures; and inappropriate or unnecessary shocks .
The rate of inappropriate shocks during the first year of ICD implantation is reported to be 6–10%, affecting 11.5–17.4% of patients in major trials . Such shocks are usually classified as supraventricular tachycardia (SVT)-related or oversensing-related inappropriate discharges. The former are the most frequent, due to the absence or failure of device discrimination algorithms, leading to the misdiagnosis of the supraventricular origin of arrhythmia as ventricular, whereas the latter can be ICD-related (QRS, T-wave or P-wave oversensing, inappropriate programming), lead-related (isolation defect, fracture, dislodgement) or environmental (electromagnetic interference, myopotentials).
The prevalence of unnecessary shocks is more difficult to quantify, as one cannot predict what would have happened if a shock had not been delivered to a ventricular arrhythmia. These are mainly represented by shocks to non-sustained ventricular tachycardias (VTs), which represent the majority of detected VTs. They may also be ICD-related, as antitachycardia pacing (ATP) can induce or accelerate ventricular arrhythmias and cause unnecessary shocks. Finally, the question of whether sustained asymptomatic VT should be treated remains controversial.
It is now well established that shocks, appropriate or not, are associated with significant morbidity, as they decrease quality of life, can be involved in depression and anxiety, and are known to be proarrhythmic . Most recent data have even shown an association between shocks and overall mortality . These data are easily explained: patients with more severe cardiomyopathy and a poorer prognosis usually present with a higher ventricular arrhythmia burden and thus may experience more appropriate shocks; most inappropriate shocks are related to atrial fibrillation (AF), which is one of the main factors associated with a poor outcome in patients with heart failure. Shocks were actually thought to reflect a more advanced cardiac disease. The results of the recent MADIT trial to reduce inappropriate therapy (MADIT-RIT) demonstrated that changing the ICD programming to reduce the rate of ICD shocks might lead to a reduction in overall mortality of up to 55% . Even if the underlying mechanisms still remain unclear, the deleterious effects of shocks themselves, which can cause myocardial damage, may explain these findings. Shocks can be associated with an increased adrenergic response, inflammation and oxidative stress, which eventually may lead to an alteration of the patient’s haemodynamics .
As opposed to other defibrillator-related complications, the rate of inappropriate and unnecessary shocks can be decreased with adequate programming. This review focuses on the different programming strategies available to reduce the rate of shocks in primary prevention patients with reduced LVEF, implanted with an ICD.
Supraventricular tachycardia discrimination
Discrimination is used to define the ability of the algorithms for differentiating non-ventricular-detected events from true ventricular tachyarrhythmias. The early generations of ICDs had only rate detection. Once the ventricular rate was detected above a preprogrammed rate during a minimal prespecified duration, therapies were automatically delivered. The sensitivity (percentage of clinical ventricular arrhythmias accurately diagnosed as ventricular episodes) was high, but the specificity (percentage of accurately diagnosed ventricular arrhythmias among all ventricular episodes) was rather low, with a high-rate of inappropriate therapies. Algorithms were then progressively implemented in the software to improve specificity, while remaining safe (high sensitivity); they should be used routinely, but in patients with permanent complete atrioventricular block in which rapid ventricular rates can only have a ventricular origin.
Single-chamber algorithms
Three criteria are mainly used: the sudden onset criterion is able to discriminate sinus tachycardia from VT, as the former usually displays a gradual rate increase; the stability criterion can discriminate AF from VT, as the former usually displays irregular RR intervals; the morphology criterion is used to discriminate supraventricular from ventricular arrhythmia by comparing the similarity of the ventricular electrogram during tachycardia with a reference electrogram template acquired during a rhythm in which the supraventricular origin is certain.
These three main algorithms can usually be used in combination to improve discrimination accuracy. In a tiered-therapy cardioverter-defibrillator study, a programmed stability of 40 ms (interval maximum variation between detected interval and any of the three previous intervals) decreased detection of AF by 95–99%, with a sensitivity for true VT episodes of 100% . In the same study, a programmed sudden onset ratio of 87% (detected interval divided by the mean of the four preceding intervals) rejected 98% of sinus tachycardia episodes, while 0.5% of true VT episodes were misdiagnosed. A strategy of programming both stability and onset criteria has proven effective in decreasing the rate of inappropriate therapies by > 50% . Some manufacturers have since implemented morphology algorithms ( Table 1 ).
| Biotronik a | Boston Scientific b | Medtronic c | St. Jude Medical d | Sorin Group e | |
|---|---|---|---|---|---|
| Sudden onset | Fulfilled if the average of the last four DIs is X% shorter than the average of the previous DI (eight DI sliding window) Nominal: X = 20% | Available with onset/stability only Fulfilled if the average of the four consecutive DIs following sudden onset is X% shorter than the average of the four consecutive DIs preceding the two intervals before sudden onset Nominal: X = 9% | Fulfilled if the average of four DIs is < X% of the average of the four previous DIs Nominal: X = 81% | Fulfilled if the average of the last four DIs is > X% (adaptive) or < Y ms (fixed) shorter than one of the averages of four consecutive DIs within the last eight DIs Nominal: X = 18%; Y = 100 ms | Fulfilled if a DI is > X% shorter than the average of the four previous DIs Nominal: X = 19% |
| Stability | Fulfilled if the difference between a DI and each of the three previous DIs is < X ms Nominal: X = 24 ms | Available with onset/stability, and with rhythm ID for redetection only Fulfilled if a DI varies by > X ms from the average variation of the five previous intervals Nominal: X = 30 ms (rhythm ID) or 20 ms (onset/stability) | Fulfilled if the fourth DI is within X ms of the three preceding DIs Nominal: X = 40 ms | Fulfilled if the difference between the second longest and the second shortest interval during a window of X consecutive DIs is < Y ms Nominal: Y = 80 ms; X = eight | Fulfilled if a majority (> X%) of DIs within the last Y intervals vary by < Z ms Nominal: X = 75%; Y = eight; Z = 65 ms |
| Morphology | None | Available with rhythm ID for initial detection only Vector timing and correlation: SVT is indicated if three of the last 10 complexes are similar (> X%) to template; VT is indicated if eight of the 10 last complexes are not correlated Nominal: X = 94% | Wavelet: detection is suspended if three of the last eight complexes are similar (> X%) to template; detection is fulfilled if six of the last eight complexes differ (< X%) from template Nominal: X = 70% | Far-field morphology discrimination: SVT is indicated if Y out of Z DIs are similar (> X%) to template Nominal: X = 90%; Y = three out of Z (= 10 DIs) | None |
| Others | None | None | None | None | Long cycle occurrence: therapy is inhibited for 24 cycles if a DI is longer than the average of the four previous DIs by > X ms Nominal: X = 170 ms |
Dual-chamber algorithms
Addition of an atrial lead allows analysis of the atrioventricular relationship during arrhythmia. Each manufacturer uses its own approach, but the main principles are similar:
- •
if the ventricular electrograms are more frequent than the atrial electrograms (V > A), which is actually the case in > 90% of VTs, the ventricular origin is certain;
- •
if the number of detected ventricular electrograms equals those from the atrium (V = A), the supraventricular origin is more probable, although VT with a 1:1 retrograde conduction cannot be ruled out;
- •
if the atrial electrograms are more frequent (A > V), the supraventricular origin is also more probable, although dual-tachycardia should also not be ruled out (VT in a patient with AF, for instance).
Single-chamber algorithms and PR association analyses are then implemented to achieve a final diagnosis ( Figs. 1–5 ).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
