The Growing Mismatch Between Patient Longevity and the Service Life of Implantable Cardioverter-Defibrillators

ICD patient longevity

The cumulative survival of 1,280 patients who have received ICDs since 1987 at the Minneapolis Heart Institute is shown in Figure 1.The probability of a patient living 4, 5, and 6 years after ICD implantation is 79 ± 1%, 75 ± 1%, and 68 ± 2%. Over 40% of patients were still living at 10 years. The average ejection fraction was 34 ± 15%, and these patients had ischemic heart disease (73%), non-ischemic cardiomyopathy (13%), hypertrophic cardiomyopathy (5%), and other cardiac conditions. These data suggest that the majority of ICD patients would not require a replacement procedure for battery depletion if the service life of ICD models was at least 10 years.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Survival of patients after implantable cardioverter-defibrillator implantation. SE = standard error of the mean.

ICD pulse generator service life

The Multicenter Registry has gathered ICD pulse generator failure data since 1999. To date, of 1,107 ICD pulse generator failures, 90% were caused by normal battery depletion and 10% were caused by electronic or housing defects. The majority of pulse generators were manufactured by Guidant Inc. (St. Paul, Minnesota) (49%), Medtronic Inc. (Minneapolis, Minnesota) (31%), and St. Jude Medical Inc. (St. Paul, Minnesota) (19%). The average service life of these pulse generators was 4.3 ± 1.5 years (range <1 day to 8.9 years). Only 5% of pulse generators functioned for seven years. After excluding non-battery failures caused by electronic or housing defects, the average service life of 814 single-chamber pulse generators was 4.7 ± 1 year compared with 4.0 ± 1 year for 293 dual-chamber pulse generators (p < 0.0001). Thus, the service life of contemporary ICD models is far less than the expected patient longevity, and the shift to dual-chamber models has significantly shortened battery life. The growing proportion of dual-chamber implants may be inappropriate because patients who do not require bradycardia pacing support may be adversely affected by right ventricular stimulation (11).

Cost effectiveness

In the U.S., the generally accepted benchmark for economically attractive cost-effective care is $50,000 per life year. The cost effectiveness of ICDs depends, in part, on the population being studied and the frequency and cost of pulse generator replacement. A recent study by the Blue Cross Blue Shield Association (12) evaluated the cost effectiveness of ICDs in the Multicenter Automatic Defibrillator Implantation Trial (MADIT)-II population. The base case assumed a 7-year pulse generator lifespanand a 67% reduction in sudden cardiac death. Given these assumptions, the investigators estimated that the cost effectiveness of ICDs was $51,000 per quality-adjusted life year (QALY). Sensitivity analyses showed that cost effectiveness was $44,300 per QALY if the pulse generator was replaced every 11 years but decreased to $58,100 per QALY if the pulse generator was replaced every 5 years. Based on the average longevity of devices in the Multicenter Registry’s database, the base-case assumption overestimated the cost effectiveness of ICDs. Nevertheless, this study showed how important pulse generator longevity is to the cost-effective application of ICD therapy.

Effect of battery capacity on service life

The ICD pulse generators use proprietary lithium silver vanadium oxide batteries that are capable of delivering ampere-level current. One manufacturer (Guidant Inc.) has offered four ICD models with a choice of battery capacities, i.e., a standard battery (0.92 to 1.06 AmpHr) or a large battery (1.44 to 1.75 AmpHr). This allowed a multicenter comparison of the effect of different battery capacities on pulse generator service life (13). The results are shown in Figure 2.Large-capacity batteries increased average service life by 2.3 years. The number of therapeutic shocks did not seem to affect battery longevity in either group. Large-capacity batteries increased pulse generator size and weight by 5 cc and 11 g. The added battery longevity reduced annualpulse generator costs by $1,000 to $1,600 per patient. These data suggest that larger batteries can reduce costs by significantly prolonging the service life of ICDs with only a small increase in pulse generator size.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Longevity of standard and large-battery-capacity implantable cardioverter-defibrillator pulse generators. Non-battery failures were censored. SE = standard error of the mean.

Physicians tend to believe that patients would resist any increase in pulse generator size. However, a recent study by Wild et al. (14) found that 90% of patients thought that ICD and pacemaker longevity was more important than size. The vast majority of patients preferred a larger device that could reduce the number of potential replacement procedures. This preference was consistent across the spectrum of patients with a previous implant, those undergoing initial implantations, those returning for routine follow-up, and patients of various ages, gender, and body habitus. However, large pulse generators may increase the risk of pocket complications, including erosion and discomfort, and some patients may not be suitable candidates for such models.

Physicians are also concerned that a long-lived pulse generator may become technologically obsolete as newer models are introduced. Consequently, it is argued that patients may be denied the benefits of new features or functions. Such reservations, however, are not supported by clinical studies that quantify the risks and benefits of regular pulse generator replacement. Moreover, pulse generator replacement is associated with known complications, including infection and the potential for lead damage. More than one-half of all pacemaker and ICD infections occur after replacement procedures (15). An ICD infection often requires complex lead extraction that is associated with significant morbidity and mortality (16). Thus, pulse generator replacement is not desirable except in patients whose clinical circumstances warrant the risks and costs of upgrading to a new model.

The growing economic burden of ICD therapy is likely to accelerate as more ICD implants include CRT. These devices are more expensive and use more energy than standard ICDs. Over the past 5 years, the Minneapolis Heart Institute’s annual new and replacement ICD implantations have more than doubled from 148 in 2000 to 326 in 2004 (Fig. 3).In addition, the proportion of ICDs with CRT capability (ICD-CRT) has increased from nil in 2000 to 44% in 2004. During this same period, our hospital’s total ICD device supply costs increased from $2.6 to $7.0 million, and ICD-CRT models accounted for 55% of ICD device costs in 2004 (Fig. 4).The growth and shift to ICD-CRT models were driven by the results of clinical trials rather than changes in our practice or patient population. With the recent publication of the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) (7), we expect a substantial increase in the number of ICD-eligible patients, and device costs will grow accordingly.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

The implantable cardioverter defibrillator (ICD) implants at the Minneapolis Heart Institute during the past five years. CRT = cardiac resynchronization therapy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Cost of implantable cardioverter-defibrillator (ICD) pulse generators at the Minneapolis Heart Institute during the past five years. CRT = cardiac resynchronization therapy.

Unfortunately, as of this writing, only one manufacturer provides a large-battery-capacity ICD, and this option is available for one dual-chamber model. Thus, we do not have the choice of implanting large-battery-capacity single-chamber or CRT ICDs because they are not available. The ICD manufacturers have no incentive to provide long-lived pulse generators. Indeed, the opposite is true, namely frequent replacements increase sales and profits. Although rechargeable batteries exist, they are not being used. Whereas hospitals could benefit financially by using shorter-life ICDs, our analysis of the diagnosis-related groups suggest that institutions will have difficulty recovering the costs of ICD pulse generator replacements unless the replacement device is an upgrade, e.g., from a single-chamber to a dual-chamber model.

It is time to revise reimbursement methods to encourage manufacturers to produce and hospitals to implant longer-life devices. Several approaches could be adopted by payers and offered by manufacturers, including a program that would cap device costs for the life of an individual patient and reward ICD longevity. Patients and payers would benefit because the frequency of replacement surgery would decrease, and the cost-benefit of ICD therapy would improve. Further, it is more reasonable for Medicare to focus its cost containment efforts on the actual costs associated with delivering ICD therapy than to limit reimbursable indications by arbitrary coverage decisions. Whatever the solution, the problem is clear, i.e., the health economy is facing a staggering expense. The SCD-HeFT study results alone could double the number of ICD implants. The medical community, payers, and industry must address this critical issue in 2005. If we begin now, most patients could receive a lifetimeICD pulse generator by 2007. Although additional strategies will be needed to control the costs of ICD therapy, longer-life models can narrow the mismatch between patient and device longevity.