William Maisel, M.D., calls implantable defibrillators and pacemakers the poster children of modern medicine. Amazing devices, life-sustaining devices, devices that we are lucky to have. However, they’re also very complicated devices. Beneath their cool titanium skins are computers, circuits and gizmos that most of us understand very little about. “The message isn’t that they are bad devices,” says Maisel, a practicing cardiologist who is the founder of the Medical Device Safety Institute at Beth Israel Deaconess Medical Center in Boston and assistant professor of medicine at Harvard Medical School. “The message is that modern medicine relies on sophisticated technology, and things sometimes go wrong. We need to monitor device performance so we can detect problems as soon as possible.”
JHC readers may not recognize Maisel’s name, but they will know some of the high-profile work he and colleagues have conducted in the area of medical device safety. For example, in March 2008, Maisel and two fellow researchers from the University of Massachusetts and the University of Washington demonstrated how hackers could extract private medical information from patients’ implantable devices. Worse, they showed how hackers could actually reprogram these devices without the patient’s knowledge.
Maisel has been concerned about medical device safety for about 10 years, when he first realized how difficult it is to get objective data about the performance of medical devices on the market. He has published articles about the topic, served as chairman of the FDA’s postmarket and heart device advisory panels, and participated in CMS’s Medicare Evidence Development and Coverage Advisory Committee. Since 2003, he has served as a consultant to the FDA’s Center for Devices and Radiologic Health. In May 2008, he testified on medical-device liability claims before the U.S. House Committee on Oversight and Government Reform. He agreed to share with JHC readers his concerns about the safety of medical devices on the market, and some potential ways to solve the problem.
Journal of Healthcare Contracting: How and why did you become interested in the issue of medical device safety?
William Maisel, M.D.: I’m a cardiologist and I specialize in the treatment of patients with arrhythmias. As a result, I implant pacemakers and defibrillators. Back in the late 1990s, there were a number of product recalls affecting them. That was notable to me. I set about trying to answer what I considered a very straightforward and easy question: How many patients were affected by these product recalls? How many were being harmed by safety issues? It turned out that getting answers to these simple questions was enormously difficult. It took over a year of research and hard work to get them.
It became increasingly clear to me that in order to improve safety, you have to have measures of safety. The type of information [needed to establish the safety] of devices – at least for pacemakers and defibrillators – was very difficult to quantify. As a result, I became a consultant to the FDA. Subsequently, I served as a member and chair of the FDA postmarket and heart device advisory panels, I have been involved in the drug-eluting stent issue and other cardiovascular device issues. But initially, pacemaker and defibrillator issues attracted my attention.
JHC: Can you describe the Medical Device Safety Institute? When was it established? How does it operate? Does it concern itself with drug safety as well as that of medical devices?
Maisel: It became clear to me that it is very difficult for a patient or physician to get unbiased, straightforward information about device safety and performance. Much of the research that is published is sponsored by industry. The reputation of the FDA – which has many hardworking and thoughtful people – has taken a little bit of a hit, so that patients and physicians don’t necessarily fully rely on the information coming out of that organization. It struck me that there was a vacuum, a lack of an impartial industry-independent organization that could provide information to physicians and patients about these sorts of issues.
So the Medical Device Safety Institute was formed in 2007. It is a part of Beth Israel Deaconess Medical Center, a nonprofit organization. It is funded primarily by philanthropy and non-industry grants. Because of the diversity of the types of knowledge that are needed to address the many issues out there, we’re relying primarily on consulting arrangements with people who have appropriate expertise. It’s conceivable the organization will blossom into one that has a larger, full-time staff.
JHC: Much press has been given to your concerns about the safety of implanted devices. Does the Medical Device Safety Institute concern itself with other devices as well?
Maisel: Implantables tend to be the highest profile devices, both because they have the greatest impact on patient safety and the greatest potential to improve patient care. They tend to be expensive, so often there are large amounts of money invested in them by many companies. But other devices are also important, such as telemetry units and infusion pumps in hospitals, which have a great impact on patient care, and which are becoming increasingly sophisticated. We are also interested in one-time-use devices, such as catheters or wires or other products used in procedures.
JHC: Do manufacturers of medical devices have a tougher or easier time establishing the safety of their devices prior to introducing them to the market than manufacturers of pharmaceuticals?
Maisel: The clinical trials [for drugs] are larger and very scientifically robust. They are randomized, meaning that [patients] either get the drug or a placebo; and that’s decided by a flip of a coin. The studies are double-blinded [meaning neither the medical staff nor the patient knows which of several therapies the person is receiving]. For medical devices, these things are difficult to do. Imagine a total artificial heart. You can’t “blind” someone during a study; the patient knows whether he or she got an artificial heart or not. The same is true with stents or insulin pumps. Patients know something is being done to them. So it’s very difficult to have scientifically robust studies of medical devices. But researchers do their best. Rather than using a placebo, they may [compare a new device] with the best medical therapy.
Another factor [contributing to the difficulty of establishing the safety of a medical device vs. that of a drug] is that after a drug comes on the market, it remains on the market unaltered for the rest of its life. Devices, on the other hand, go through very rapid changes because of the movement of technology. A device marketed five years ago likely isn’t on the market today. It’s been revised or upgraded, or a newer, smaller model has been introduced. So the life cycle of medical devices is much shorter than that of drugs. As a result, the duration of study required by the FDA for new medical devices is much shorter than that for drugs. There’s a much greater chance that a product that gets to market will have potential safety problems.
Another challenge is this: It’s difficult to determine when a medical device has caused a problem. Suppose someone has undergone surgery to have a device placed, and bleeding occurs. Is that a complication of the implanted device, the surgical procedure, or perhaps of a device used during the procedure? It can be challenging to sort out whether the medical device caused the problem.
Medical devices are amenable to bench testing. So some features can be tested on a desktop – either by computer simulation or in a lab. At the Institute, we plan to develop better bench testing for devices. The ideal outcome is to identify safety problems before devices ever get to market. That means developing robust testing to ensure their safety without lengthening the time it takes to bring them to market.
A paradigm of the type of work we do at the Institute was our report regarding defibrillator and pacemaker security [in which Maisel and colleagues from the University of Massachusetts and University of Wisconsin proved that hackers could steal medical information from, and even reprogram, implantable pacemakers and defibrillators from unsuspecting patients]. This is an example of an issue facing medical devices and burgeoning technology, related to the security and privacy of the information that these devices contain. There’s a potential for interaction among devices, as more patients get multiple implanted devices. And there’s the potential for interaction in the hospital environment, where telemetry interacts with some of these medical devices. We demonstrated that an implantable defibrillator has the potential to be susceptible to a security breach, and we were able to reprogram the device therapies in an unauthorized fashion. We did these tests on a bench top.
JHC: In your testimony to the House Committee on Oversight and Government Reform in May, you made the point that patients may fail to receive critical information about their medical device’s performance, and that they may be exposed to potentially faulty products despite the FDA’s approval process. Is it true that a manufacturer can continue marketing a device that has been found to have some kind of problem with its design or manufacturing process, without disclosing this to the public?
Maisel: That is true. There are high-profile examples of companies that have known of a malfunction but did not publicly disclose it. But this is a very complicated issue. There’s no such thing as a complicated device that will behave perfectly. Performance issues will inevitably arise. Companies should be rewarded for identifying and fixing them.
Not every performance issue has to be publicly reported directly to patients. For example, if a complicated device is known to have a 5 percent failure rate over a certain time frame, then not every malfunction needs to be publicly reported immediately. That said, companies should be obligated to report on a regular basis the overall performance of their devices. And the process of deciding which issues require public reporting and which do not could be improved. Manufacturers have an inherent conflict of interest when performance issues arise. That’s where we hope the FDA steps in and uses its good judgment to navigate through this issue.
The problem is, this process takes a long time. It’s difficult for the FDA to get all the information it needs to publicly disclose a performance issue in a timely fashion. It can take months or even a year. There have been cases in which product recalls have been issued more than a year after the manufacturer and the FDA knew of a problem. That time frame should be shortened.
JHC: Is it realistic to expect that it will be shortened?
Maisel: Yes. But it will require cooperation from industry and more resources for the FDA. And it requires good judgment insofar as focusing attention on the areas where performance issues repeatedly come up. But for manufacturers, there shouldn’t be a lot of judgment involved when deciding what needs to be reported to the FDA. There are well-defined criteria that list what manufacturers are required to report. For example, for medical devices, any adverse event or performance issue that harmed or that could have harmed a patient must be reported. More challenging is determining when an event is truly an anomaly, and when it is a bona fide problem with the product. That’s where the judgment piece comes in. That’s the gray area, where manufacturers can do a better job of publicly reporting rates of failure. But it’s difficult for the FDA, with its limited resources, to do anything more than take what the manufacturer gives them in terms of reports.
That said, there has been a paradigm shift at FDA. There used to be silos on the premarket side and the postmarket side. One group looked at the product until it was approved; then it fell under the jurisdiction of someone else, who would monitor it on the postmarket side. But they have developed more of a continuum today, so that the postmarket people are helping during the premarket evaluation to design studies that will be implemented soon after the device is on the market. And vice versa [that is, the premarket people lend their expertise to the evaluation of post market problems]. That definitely helps.
They’ve also done a better job of keeping track of what postmarket studies they’ve asked for, and which have been completed. Previously, they would ask for a postmarket study, but they wouldn’t always go back to make sure the study was done, or what the results of it were. So some studies were never started. Now the FDA has a public Web site with information about postmarket studies. There’s much greater accountability.
JHC: On Feb. 20, 2008, the United States Supreme Court decided, in Riegel v. Medtronic Inc., that once a medical device has premarket approval by the FDA, federal law bars many types of claims by users of medical devices that challenge the safety or effectiveness of the device, its design or its label. The plaintiffs’ claims were thrown out on the basis that they were “preempted” by the FDA’s approval of the device. Meanwhile, both houses of Congress are considering the Medical Device Safety Act, which would allow individuals to bring suit against manufacturers. Can you comment on this issue?
Maisel: To me, the idea that FDA approval would preempt manufacturers’ responsibility for ongoing product performance seems inappropriate. Many things happen after the FDA approves a device. The manufacturer still has a responsibility to make sure that its plants are operating appropriately and that devices are being built properly. That being said, just because a device has a malfunction doesn’t mean the manufacturer should be liable. Many devices are very complicated and they will malfunction, even if everything is done properly.
One argument made [in favor of FDA pre-emption] is that a jury of lay people aren’t sophisticated enough to sort out whether the manufacturer should be held liable or not. For me, the analogy I would use would be malpractice suits. Many malpractice suits go to trial, and many of these issues are complicated. But 95 percent of the time, juries find in favor of the physician; only one in 20 suits are actually found in favor of the plaintiff.
Certainly, litigation is expensive. But it also has implications for patient safety. The threat of a lawsuit will encourage manufacturers and companies to be a little more careful.
JHC: Given the recent spate of adverse events related to Chinese-made products (medical and otherwise), what is being done to ensure the safety of foreign-made devices? The FDA has launched its Beyond Our Borders program, which is an effort to establish an FDA presence overseas in at least five regions, beginning with China. Is that enough?
Maisel: This is an extremely challenging issue for the FDA. Many drugs and devices are made in other countries. It’s unrealistic to think the FDA can carefully monitor all of them to absolutely ensure their safety. The Beyond Our Borders program is conceptually a great idea. Rather than sit here and wait for products to come to us, we put tentacles out to the places where products are being built. But even manufacturing plant inspections can’t uncover complicated product problems. So the concept would be to get on the ground where products are being built, not only to inspect manufacturing plants, but to develop relationships with regulatory agencies in other countries. Not that the FDA doesn’t already have these relationships, but if they’re present in the country, they will be stronger.