Complications Associated With ICD Replacement

Abstract & Commentary

By John P. DiMarco, MD, PhD, Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville. Dr. DiMarco is a consultant for Novartis, and does research for Medtronic and Guidant.

Synopsis: ICD generator replacement in patients with device advisories is associated with a substantial rate of complications, including death.

Source: Gould PA, et al. Complications Associated with Implantable Cardioverter-Defibrillator Replacement in Response to Device Advisories. JAMA. 2006;295:1907-1911.

In this study, gould and colleagues from the Canadian Heart Rhythm Society tracked complications associated with implantable cardioverter-defibrillator (ICD) generator replacements that were performed in response to recent device manufacturers' advisories or recall notices. The objective of the study was to document the frequency of and the reasons for replacements and of any complications associated with the replacements. Participants in the survey included 17 of the 21 ICD implant centers in Canada. These centers represent a catchment area that includes almost 93% of Canada's population. Questionnaires were completed by each center describing the number of patients they followed with devices covered by advisories or recalls in 2004 and 2005. The reason for any device replacement was recorded and complications of device replacement documented. Devices replaced for indications (upgrades, normal battery depletion, infection, etc.) other than the advisory were not included.

During the period of the survey the centers reported there were 2915 patients with devices under advisory. Of these, 533 patients (18%) underwent elective replacement, a mean of 26.5 months after the initial implant. The mean age of the patients was 64 years and 77% were male. Sixty-six percent had received their ICD for secondary prevention, 45% had previously received an appropriate shock, and 21% were pacer dependent. Replacement rates, however, varied widely between centers. Four of 17 centers replaced less than 10% of the devices they followed that were covered by an advisory notice, whereas, 5 of 17 replaced more than 30%.

Complications associated with device replacement were noted in 43 patients. Twelve of these, however, were classified as minor, including 9 medically-managed incisional infections. Major complications were noted in 31 patients (5.8%). Ten patients developed pocket infections that required system extraction, and 2 patients died as a result of either persistent infection or as a result of the extraction attempt. Re-operation after generator change was required in 21 patients (3.9%). The reason for re-operation was either a large hematoma requiring drainage, system malfunction, or site pain.

During the period of the study, only 3 spontaneous device malfunctions covered by the advisory notices were detected. These were not associated with clinical injury. Gould et al conclude that complication rates after device replacement may have been underestimated in the past. There is a substantial morbidity and detectable mortality associated with elective device replacement, even in high volume ICD implant centers. They urge consideration of these data when guidelines concerning device replacement in a recall or advisory situation are formulated.


ICD and, to a lesser extent, pacemaker malfunction were top medical news stories in 2005. The trigger for the increased media interest was the death of a young man with hypertrophic obstructive cardiomyopathy whose death was associated with an electrical short in his defibrillator circuitry. The device's manufacturer came in for considerable criticism when it was revealed that similar events had been reported earlier and were being tracked by the company. Although these malfunctions had been reported to the FDA, no direct public notification had been made. One of the companies' rationales for not issuing a public advisory was that their projection that the risk of replacement for patients in general would probably be higher than the overall risk of failure. This argument has been largely rejected by the physician community who feel that physicians and patients should make any decision about replacing an ICD or pacemaker based on individual patient characteristics rather than leaving the decision to the device manufacturer.

Recent proposed guidelines from the Heart Rhythm Society are that the decision to replace a device with a systematic malfunction should be made on an individual basis, with the final decision made by the patient and his doctor. The data reported here should be helpful to both patients and physicians when they consider any possible replacement. In this study, the risk of replacement was considerably higher than the risk of patient injury associated with device failure. A similar observation was made 10 years ago when a large number of patient injuries and deaths occurred during attempts to remove Telectronics ACCU-FIX atrial leads.

Decisions about whether or not to replace a device under a recall or advisory remain quite difficult. This paper gives us a better estimate of the real risk of replacement. Physicians will also be dependent upon manufacturers and regulatory agencies to provide accurate predictions of the real failure rates during use, and an accurate clinical assessment of the consequences should the device fail. Only with full access to and understanding of the necessary information, can physicians and patients make informed decisions.