A recent study led by the Yale School of Medicine found that medical devices that were designed based on previously recalled models were more likely to also be recalled, potentially putting patients’ lives at risk.
Medical devices are everywhere – from thermometers and pulse oximeters to IV transfusion pumps and cardiac implants. Last year, 149 medical devices were subject to Class I recalls — a category in which continued use “will cause serious adverse health consequences or death.”
Unlike the approval process for drugs, more than 95% of new medical devices authorized by the Food and Drug Administration (FDA) are not based on new clinical data.
The Yale study, in collaboration with Harvard Medical School, analyzed a number of medical devices that eventually issued Class I recalls.
“If the FDA authorizes a device based on an unsafe device, a device that was pulled off the market because of a Class I recall, then the risk that that device is going to subsequently be recalled, is much, much higher than if that device had been authorized based on a device that had not been recalled,” said Dr. Harlan Krumholz, cardiologist and professor of medicine at Yale School of Medicine.
“I mean, that seems to make common sense, you’re basically putting this out on the market because of its similarity to a recalled device, then the risk of that device ending up being recalled is pretty high, much higher than average,” Krumholz said. “And ... you wouldn’t even think we’d need to show that. But we’re trying to present the evidence to lock down this case, that we should close this loophole.”
The study looked into 156 cases of FDA-authorized devices with Class I recalls from 2014 to 2021. The researchers found that 44% of the devices used predicates (a medical device in the U.S. that is used as a comparison to develop another, similar device for FDA authorization) with Class I recalls. In addition, 48% of the devices were subsequently used as predicates to authorize devices that later themselves would be subject to Class I recalls.
Kushal Kadakia of Harvard Medical School, the study’s co-author, offered an example: CareFusion’s Alaris Infusion Pump, a device that is used to control the delivery of IV fluids and medications for hospitalized patients, has been subject to multiple Class I recalls between 2019 and 2021, with hundreds of thousands of units affected. The device cites three previous designs to help it gain FDA authorization, two of which have also been subject to multiple Class I recalls between 2009 and 2014 (before Alaris' approval), and both of them were subject to active recalls affecting thousands of units that were still unresolved at the time Alaris was authorized in 2014. Both class-action and shareholder lawsuits are pending.
Also, Biomet’s Artificial Shoulder, a device used in shoulder replacement, was recalled nine years after it was approved “because of a higher than expected fracture rate due to the device’s design,” Kadakia said. “The device cites seven predicates, six of which have been subject to some recall of any kind, including two with Class II recalls prior to the new device's approval, and many others with recalls after the device's approval, with thousands of units affected.”
Kadakia said manufacturers have continued to use Biomet’s device as a predicate even after it was subject to a Class I recall in February 2017 and has since had three more Class II recalls. This “makes you wonder whether patients knew the regulatory history of the devices implanted in their shoulders,” said Kadakia.
“Our premises [is] that people don’t really realize what a big issue this is, that in fact, there are many devices that end up being used in tens of thousands of patients that are authorized without any new clinical data, and solely based on its similarity to a device that ultimately was pulled off the market and considered to have had life-threatening safety issues,” Krumholz said.
Legislation was introduced about a decade ago to try to address the loophole, but it never made it through Congress.
