Pacemakers and implantable defibrillators malfunction far more often than manufacturers acknowledge, according to ongoing research in California.
Zian Tseng, a physician who researches cardiac devices at the University of California, San Francisco, realized as far back as 2006 that he wouldn’t necessarily know if the pacemakers and defibrillators he implanted in his patients failed. When a person with heart disease dies of a heart attack, it’s rare for anyone to look deeply into the cause. “It’s really the medical examiners that see those cases, the coroners,” Tseng said. “But the problem is that coroners and medical examiners often are busy with legal cases. The natural deaths almost never get investigated.”
So Tseng decided to investigate heart attacks as thoroughly as medical examiners investigate murders. For 35 months, he and his team tracked down everyone who died suddenly in San Francisco County, regardless of the cause. If the person’s body had a cardiac device, the team removed it for examination and performed an autopsy to identify the cause of death.
The results the team published in 2015 stunned Tseng: Half of the cardiac devices had some kind of defect, ranging from drained batteries to serious mechanical failures. Among defibrillator patients, 6.4% of those who died suddenly had a problem with their device.
Tseng quickly realized that cardiac devices could be malfunctioning much more often than the manufacturers indicated. “If you do a back-of-the-envelope calculation on the rates of failure based upon the overall pool of devices out there, it’s orders of magnitude higher than the device companies publish,” Tseng said. So he approached three major manufacturers—St. Jude Medical, Medtronic, and Boston Scientific—to tell them what he’d found.
“When I went to each of these device companies, I got a range of responses from ‘Wow, this could really help our [quality control] processes’ to outright abject denial that these were even issues,” Tseng said.
Tseng found Medtronic the most receptive of the companies he contacted. According to Tracy McNulty, a Medtronic spokeswoman, they offered Tseng technical support to continue his research. Medtronic was the only one of the three companies that agreed to comment for this story.
Tseng says that St. Jude Medical listened to what he had to say, but was less enthusiastic than Medtronic. A St. Jude spokesman initially responded to a request for comment, but did not provide one before press time.
Tseng says he and his research got a cold greeting at Boston Scientific. The company did not respond to repeated requests for comment. “They didn’t want a collaborator, let’s put it that way,” Tseng said. “That was the least constructive response.”
Manufacturers aren’t the only ones responsible for keeping tabs on medical devices. Once devices have been released into the market, the Food & Drug Administration (FDA) is in charge of making sure they’re as safe and effective as clinical trials indicated. If a doctor thinks her patient has been hurt or killed by a device problem, she can report it to the FDA. “There are two flaws with that,” Tseng said. “Number one, it’s voluntary. Number two, we only see the people who make it to the clinic, who are by definition alive.”
Joseph Ross, a Yale cardiologist who studies health-care delivery, agrees. “The way we are able to monitor medical devices…is quite limited, actually,” Ross said. That’s partly due to the lack of data. When a patient is given a medication, the exact drug and dose are reported to whoever ends up footing the bill. But when a patient gets a medical device, the bill is for the implantation surgery. There’s no record of what brand of device the patient got, and no way to track down a specific device in the future.
“The FDA recognizes limitations in its current medical device post market surveillance,” said FDA spokeswoman Stephanie Caccomo. She also said that the FDA is pursuing “big data” solutions to reveal patterns more quickly.
Better analysis would do a lot of good for the FDA database, which Ross called “a black file of data.” But it won’t change the fact that fatal problems are hardly ever reported to begin with because natural deaths are so rarely investigated.
That’s why Zian Tseng has continued to learn from the dead. His new data, released later this year, will be welcomed by researchers who have spent years trying to study device failure using the tiny handful of published cases. As Ross put it, “that’s a tip-of-the-iceberg type of signal.”