Safety of Implantable Cardiac Devices

To the Editor: Dr Maisel and colleagues¹ called into question the safety of pacemakers and implantable cardiac defibrillators (ICDs) and caused undue alarm among device recipients and physicians. They reported that during the past 11 years, 52 advisories affected more than 500 000 patients. This led to stories in the lay press that each of these devices poses significant risk to patients, which may have caused unfounded concern among patients and physicians.

Medtronic's primary concern during the design, manufacturing, and evaluation of these devices is patient safety and well-being. When a potential problem is detected, we perform a Health Hazard Analysis, which predicts the risk to patients and the number of affected devices. Because it is impossible to predict which specific devices will develop the potential problem, advisories include all people with devices that potentially are affected. Today's sophisticated devices offer a number of corrective solutions, such as simple reprogramming, downloadable software, or, in extreme cases, explantation.

During the 1990-2000 time period analyzed by Maisel et al, Medtronic had 12 advisories encompassing 126 908 patients worldwide. However, the total number of devices predicted to be affected was 1243 (<1.0%). Devices that were explanted because of advisories represented 0.6% of all devices under advisory and only 0.03% of the more than 2.1 million people who received Medtronic pacemakers/ICDs during that time period. We believe that our findings are representative of the industry. Thus, we believe that the actual number of explanted devices is closer to 3200 rather than the 36 187 estimated by Maisel et al. Their estimate of the cost of device advisories thus also may be grossly overstated.

In his accompanying Editorial, Dr Eagle² concluded that the regulatory system is deficient. His first recommendation ("longer periods of preapproval testing") implicitly assumes that all problems are detectable if only preapproval testing were lengthened. Failures often occur at such low frequency that they are evident only in a very large population. For instance, one recall was based on 20 malfunctions out of 144 000 devices.³ Furthermore, delaying approval denies people the benefits of new technologies. Until the early 1990s, ICDs were implanted via thoracotomy with a perioperative mortality of 4%. Technological advancements have reduced that mortality to less than 1%.⁴ This year, with about 80 000 ICDs being implanted, more than 2400 lives will be saved just by that technological advance.

His second recommendation calls for "independent, non-industry representatives" in the preapproval process. This is the role of the US Food and Drug Administration (FDA). The FDA has internal technical and clinical expertise and outside experts to support their review. Eagle did not present any data or arguments suggesting that the FDA is failing in this role.

His third recommendation urges inclusion of non-industry representatives in postapproval device tracking systems. Given that it is the current industry tracking systems that have identified all these problems, including one with 20 malfunctions out of 144 000 devices, the postapproval process cannot be faulted.

We support improvements to premarket and postmarket evaluations and useful changes to the regulatory process. This must be done with thoughtful consideration of the benefits and cost of such changes. Eagle's proposed "logical solution" would slow technological advances, worsen patient safety, and unjustifiably deprive people of life-sustaining benefits.