Section 4: Cardiovascular Diseases

4.2e. Implantable Cardioverter/Defibrillator Devices (ICDs)

ICDs are used in the management of individuals with recurrent VT of VF that cannot be controlled with antiarrhythmic medications. ICDs treat symptoms of cardiac disease (e.g., VT/VF) that are notable because of their 1) unpredictability, 2) suddenness of onset, and 3) potential for rapid incapacitation of the individual.

The Working Groups on Cardiac Pacing and Arrhythmias of the European Society of Cardiology (Jung et al., 1997) suggest that the following issues be considered to help assess risk of death and injury as a result of ICDs:

i) Risk of recurrence

Data derived from individuals who have had cardioverter-defibrillator devices implanted to manage their ventricular tachyarrhythmias suggest that the risk of recurrence of a ventricular arrhythmia is highest in the first few months following discharge. Using those data, the recurrence of arrhythmias can be determined by examining delivery of appropriate shocks by the ICD. In the older models, “appropriate shocks are usually defined as shocks delivered during sustained VT or VF documented by ECG recordings or shocks delivered during a period of syncope or presyncope that resulted in restoration of consciousness” (Miles, 1997, p. 328). Newer devices provide telemetered electrograms during arrhythmia. Results from studies using telemetered electrograms suggest that appropriate shocks (i.e., shocks delivered for VT/VF) may occur in the absence of premonitory symptoms. Thus, the strictcriteria of defining appropriate shocks based on significant warning symptoms may underestimate the true number of appropriate shocks. Other studies indicate that up to 47 percent of VT episodes may be preceded by no symptoms (Maloney, Masterson, Khoury, et al., 1991; Marchlinksi, Buxton, and Flores, 1990; Steinberg and Sugalski, 1991). Data from the newer devices, therefore, may represent a more accurate assessment of the recurrence of VT/VF compared to data from the older models.

Results from a number of studies (see Table 6) reveal that approximately 50 percent of patients will experience an appropriate shock during several years of follow-up (Fogoros, Elson, and Bonnet, 1989; Grimm,Flores, and Marchlinski 1993; Levine, Mellitis, Baumgardner, et al., 1991; Tchou, Axtell, Anderson, et al., 1991). Fogoros et al. investigated the pattern of occurrence of shocks in individuals following implantation of an ICD from one month to 71 months. The actuarial incidence of any shocks in this investigation was 43 percent at six months, with an 81-percent cumulative incidence of shocks at 48 months. The actuarial incidence of appropriate shocks during the same time periods was 28 percent and 64 percent, respectively. Appropriate shocks were defined as shocks that were preceded by symptoms of severe lightheadedness, presyncope, or syncope, and which were followed by immediate relief of those symptoms or documented VT or VF. Actuarial incidence rates of appropriate shocks reported by Grimm et al. based on the same criteria as used by Fogoros et al. were 13 percent, 42 percent, and 63 percent at 1, 3, and 5 years of follow-up, respectively.

Table 6 Recurrence of Ventricular Arrhythmias in Individuals with Implantable Cardioverter Defibrillator Devices (ICDs)

ii) The likelihood that VT/VF episodes will result in loss of consciousness

The likelihood that episodes of ventricular arrhythmias will result in loss of consciousness is of primary consideration when assessing the driving risks of patients with ventricular tachyarrhythmias. Although the data are limited, those that are available suggest that a significant number of individuals experience lightheadedness or syncope prior to receiving a shock, and a number of patients experience loss of consciousness prior to receiving a shock (See Table 7). Importantly, the absence of syncope during one episode does not always predict absence of syncope during subsequent shocks.

The findings from studies examining predictors of risk of syncope prior to receiving an ICD shock are mixed. For example, age, antiarrhythmic therapy, and left ventricular ejection fraction (LVEF) failed to distinguish between syncope and non-syncope groups in studies by Axtelland Akhatr (1990), and Kou, Calkins, Lewis, et al. (1991). However, individuals with a LVEF of < 40 percent had a significantly higher risk of syncope than patients with a LVEF> 40 percent in a study by Bansch, Brunn, Castrucci, Weber, Gietzen, Borggrefe, Breithardt, and Block, (1998). Using Cox regression, a one percent increase in LVEF implied a decrease of syncope by two percent. Individuals with an inducible fast VT during programmed ventricular stimulation showed a 2.2 fold increase in risk. Other predictors of risk of syncope included chronic atrial fibrillation (3.6 fold increase in risk). In addition, once patients had a recurrence of their VT/VF, syncope during the first VT and a high VT rate were the strongest predictors of future syncope. Schoels, Sarason, Beyer, and Brachmann (1995) also report that patients with lower LVEF are more at risk for syncope than individuals with a higher LVEF. However, as noted by the authors, the predictive value of LVEF was low because of an overlap between LVEF in patients with and without syncope.

Table 7 Percentage of Patients Experiencing Pre-Syncope or LOC Prior to Receiving an ICD Shock for Recurrent Ventricular Arrhythmias

iii) The risk that such an event will cause a crash

There are few data on the risk of crashes in individuals with ICDs. Results from Bansch et al. (1998) reveal that  14.7 percent of the patients in that study had syncope following ICD implantation (n = 421). Most of the episodes of syncope occurred while patients were at rest (43.6 percent). Of those experiencing syncope, one syncopal episode occurred while the patient was driving, with a crash prevented by the front seat passenger.

iv) The probability that such a crash will result in harm to other road users, bystanders, or passengers

Based on the formula provided by the CCS Consensus Conference (1996), Jung et al. (1997) have estimated that allowing an individual with an ICD to operate a private motor vehicle to be associated with an annual risk of harm of individuals to other road users or harm to bystanders or passengers to be 1 in 45,000.

This estimate is based on the following calculations:

RH = TD x V X SCI x Ac
= 0.04 x 0.28 x 0.1 x 0.02
= 0.0000224 or 1 in 45,0000

(see Jung et al., 1997, p. 50 for a full description).