In recent years, the rate of pacemaker malfunctions has decreased while the rate of implantable cardioverter defibrillator malfunctions has increased dramatically, according to an analysis by Dr. William H. Maisel and a group of investigators from the Food and Drug Administration.
Using data that device manufacturers are required to report to the FDA, the investigators were able to determine that from 1990 to 2002, 2.25 million pacemakers and 416,000 implantable cardioverter-defibrillators (ICDs) were implanted in the United States. During that same period, 8,834 pacemakers and 8,489 ICDs were explanted because of confirmed malfunctions (JAMA 2006;295:1901–6).
The annual rate of malfunctioning devices ranged from 1.4 to 9/1,000 implants for pacemakers and from 7.9 to 38.6/1,000 implants for ICDs. Overall, the ICD malfunction replacement rate was significantly higher than the pacemaker malfunction replacement rate.
The annual pacemaker malfunction replacement rate waned significantly during the years covered by the study. During the final year of the study the pacemaker malfunction replacement rate was 1.4/1,000 devices, the lowest rate of any year in the study.
In contrast, the ICD malfunction replacement rate trended down during the first half of the study, reaching its nadir between 1994 and 1998, but then the rate increased markedly from 1999 through 2001 before falling somewhat in 2002. The ICD malfunction replacement rate for the final 3 years of the study was 26.8/1,000 implants, a statistically significant increase of more than 3 times the ICD malfunction replacement rate for the mid-1990s and more than 10 times the rate of pacemaker malfunctions during that same period.
Device malfunction was directly responsible for deaths in 30 pacemaker patients and 31 ICD patients. Malfunctions of pacemakers and ICDs were equally likely to result in death, but patients with ICDs were 5.6 times more likely to die of device malfunction because of the smaller number of ICD patients.
Hardware malfunctions were by far the most common type of malfunction in both pacemakers and ICDs, comprising 80% of the total. Integral device software or “firmware” was the cause of malfunction 4% of the time; miscellaneous malfunctions including physical damage, foreign material contamination, manufacturing errors, and so on caused 12% of malfunctions; and the manufacturers were unable to determine the cause of the malfunction in 5% of the cases.
Battery/capacitor abnormalities together with electrical issues accounted for more than half the hardware problems in both pacemakers and ICDs. Battery/capacitor abnormalities and charge-circuit abnormalities each accounted for a significantly higher percentage of malfunctions in ICDs than pacemakers. On the other hand, hermetic-seal abnormalities affected a much greater proportion of pacemakers than of ICDs. In absolute numbers, hermetic-seal abnormalities were responsible for 1,082 pacemaker malfunctions and only a single ICD malfunction. This was due primarily to a single manufacturer's repeated problems in multiple pacemaker models in the early 1990s.
The estimates of malfunction numbers and rates included in the study are likely to be underestimates of the true rates. Any devices that were not explanted, not returned to the manufacturer, or found by the manufacturer to be working normally were not counted as malfunctioning. Furthermore, the investigators excluded biventricular pacemakers and ICDs from the analysis because relatively few of those devices were implanted during the study period.
But the biggest source of uncounted malfunctions probably came about because of the investigators' exclusion of malfunctions that involved pacemaker or ICD leads.
In an accompanying editorial, Dr. Bruce L. Wilkoff of Case Western Reserve University, Cleveland, noted that the device, its leads, the chosen programmed parameters, and the patient's response are all interdependent (JAMA 2006;295:1944–6). The malfunction rate of this entire system is probably the most important measure of reliability, he wrote, “but it is also the most elusive because no single measure captures all elements in a verifiable fashion.”
He acknowledged that lead dysfunction can be related to design issues, implantation technique, the patient's activities, and the patient's anatomy and physiology, and that lead malfunctions were appropriately excluded from the analysis.