The Heart of the Matter

If there were a problem with a life-saving medical device implanted in your heart, someone would tell you, right? Tom Hatch thought so. He was wrong.

In March 1998, the FDA approved the valve—not via the 501(k) provision, but based on the filing of a supplement to the application for the original mechanical heart valve. The processes are legally distinct, but both result in the marketing of a device without full-scale trials. In fact, 98 percent of medical devices are approved without any clinical testing.

In theory, the agency has a process for identifying problem devices that do make it onto the market, a database of voluntary reports from manufacturers, health care providers, and patients. Unfortunately, research by congressional investigators and consumer advocates suggests that as many as 9 in 10 medical-device failures—including severe injuries and deaths—go unreported. One federal audit found that hospitals, doctors, and distributors failed to report some 5,000 device-related deaths to the FDA between fiscal years 1992 and 1995. Some of the failures that were reported weren't entered into the database for up to two years.

According to a City Pages analysis of the database, the FDA has received 411 reports of Silzone malfunctions. (Tom Hatch's valve was not found among the reports.) In 65 of the cases, the patient died. Thirty-eight of the database's reports mentioned thrombus, 27 included reports of TIA (small strokes that can be clot-related), and 18 involved strokes. By contrast, to date only seven deaths have been associated with the 300,000 recalled defibrillators that are the subject of Guidant's current scandal.

Raoul Benavides

Even with the FDA's early-warning system broken, the number of recalls of the most critical devices has skyrocketed in recent years, nearly tripling between 1992 and 2002. Heart devices are recalled most frequently. Manufacturers say the recall rate has increased because the complexity of the devices has, too. The fastest-growing category is so-called combination devices, which marry an old technology—the pacemaker, say—to a new one, like the ability to call a patient's doctor with data about the heart's function.

Virtually all recalls are voluntary, allowing manufacturers much more control over the details of the process. Overworked and understaffed, the FDA prefers this.

"It's so unusual for the government to actually do the recalls that they'd have to create a mechanism to do it," explains Public Citizen's Lurie. "Part of what happens when you have a voluntary recall is you get to set the terms of the recall. So let's say the government feels that the following five models of the device should be recalled. But you, the company, think that the fifth one has a lower risk, perhaps, and might be a big seller. So you say we're going to voluntarily recall four of them. And that leaves the FDA sort of saying, 'Well, do we want to make a big fuss about the fifth? We're getting most of what we want.' It makes it likely that the government will compromise and put up with it."

The companies, Lurie adds, can say they are being good corporate citizens. "And it allows them to keep the government out of the whole process in the course of which the government might learn additional information which the company might [prefer] never see the light of day."

Despite repeated warnings from Congress, the FDA still does not require device makers to notify recipients of recalls. For years after medical devices were first regulated, the agency did not require their makers to track who received them. In November 1986, the Bjork-Shiley heart valve was recalled after 389 reports of fractures in the struts holding the device in place; 248 people died as a result. There were 23,000 of the valves in use, but the manufacturer, Shiley, Inc., could identify only 14,000 recipients.

In 1990, Congress gave the FDA two years to establish a system for tracking patients with implantable devices. In keeping with its history of arm's-length regulation, the agency passed the buck, requiring device manufacturers to maintain databases only of patients who have the most crucial life-sustaining implants. When a device is recalled, manufacturers notify only doctors—a dramatically lower standard than when a stroller or a car part fails.

It's Lurie's opinion that the FDA ought to require companies to notify individual patients and should demand a say in what they are told. "The history of the device companies is not an honest one when it comes to this," he says. "What you really want is very careful control over the information that's provided and you want some assurance that that information will make its way to the patient, otherwise some doctors will make a paternalistic decision that this doesn't affect their patients and just kind of bury it."

Doctors, he claims, typically don't go through their old charts and call patients. "The doctors probably aren't delighted to be dealing with people who come back and say, 'You implanted this device in me and it turns out to be faulty,'" says Lurie. "At that point, since the doctor is an active participant in the administration of the device in a way that's different from in drugs, I think the doctor develops an investment in the effectiveness and safety of the device that mirrors the manufacturer."

Hatch agrees. "The only reason I contacted [St. Jude] was because I read that the valve had been withdrawn," he says. "Now, they implanted 11,000 [Silzone valves] in this country and I'll bet that there are 11,000 people who didn't see the business section of the paper in January of 2000. And I think that people should know about it."

« Previous Page
Next Page »