The U.S. FDA notice regarding device risk classification proposes to eliminate premarket notification requirements for several device types, a welcome development for industry. However, Amanda Johnston, senior attorney at Gardner Law PLLC, of Stillwater, Minn., told BioWorld that there is a question regarding the provenance of the document, as it fails to identify any FDA staff member. In addition, the notice omits several device types that are solid candidates for down-classification.
The 30-plus page notification addresses the subject of regulatory flexibilities offered by the FDA’s Center for Devices and Radiological Health (CDRH) to aid in the effort to thwart the COVID-19 pandemic. The document features several tables, including those for enforcement policies and devices with zero reports to the Manufacturer and User Facility Device Experience (MAUDE) database for the 10-year period ending in November 2020. There are also a few class II devices that are the subjects of a proposal to exempt from premarket notification requirements under Section 510(k) of the Food, Drug and Cosmetic Act, although there is also a series of class I devices for which the terms of the 510(k) program are inapplicable despite their status.
The policy is signed by Secretary of Health and Human Services (HHS) Alex Azar, an anomaly for FDA items in the Federal Register, which are typically signed by an official with that agency. HHS has been active recently regarding the undertakings of its constituent agencies, including the recent move to require that regulations be reviewed decennially with an eye toward repeal or revision.
Johnston said the lack of any documented sign-off from CDRH staff leads to questions about the extent of agency involvement, adding, “that’s something that’s important enough that we would want to know.” Conversely, the question may be moot, as the agency could allow the matter to disappear into regulatory history under the new administration, she noted.
Because the notice is signed by Azar, stakeholders could speculate that neither CDRH Director Jeff Shuren nor anyone else at the FDA is comfortable with the contents. With that said, Johnston pointed out that the document recites a U.S. Supreme Court case known for stripping preemption for 510(k) devices. She said the recitation of Medtronic v. Lohr, which emerged from the Supreme Court 25 years ago, is an oddity for a notice of this kind, because Lohr seems to have little immediate bearing on the question of risk classification.
Old information not necessarily outdated
The document provides information about the cost of regulatory filings and the elapsed time to market under the various FDA premarket regimes, although some of the recited data are anything but recent. Survey data from 2010 regarding total investment and FDA times for PMAs suggest an average cost of $94 million and 54 months, but Johnston said the 54-month estimate might still be fair for PMAs.
That figure is likely a total elapsed time, beginning with a pre-submission (pre-sub) meeting for the investigational device exemption, Johnston said. Since the beginning of the pandemic, however, “this all goes out the window, because the FDA’s time has been lagging for pre-subs,” she said. Johnston added that she has heard anecdotally that the FDA is telling sponsors that it cannot handle non-COVID pre-subs within the usual timelines.
On the other hand, total time to market also can be impeded by decisions made by device makers about one aspect or another of their device. Overall, Johnston said, it is not terribly evident that this is a conspicuous problem for the agency, given the time needed for R&D, not to mention the costly and snail’s pace nature of clinical trials.
Anomalies as administrative conveniences
Regarding class I devices that are subject to premarket notification requirements, Johnston said, “I don’t think they’re misclassified per se.” One explanation is that these devices are not backed by enough evidence for the FDA to leave them in class I without some mechanism to ensure data collection. “I think we’ll see some of these go away” in terms of premarket notification requirements as the data sets are rounded out and the agency is comfortable that special controls are unnecessary, she explained. This approach is also administratively less irksome for the agency than the down-classification process that would otherwise be invoked.
The notice also appears to argue that a class II device with a zero or near-zero presence in the MAUDE database is probably class I, and Johnston said, “there’s an argument to be made that there’s a connection there.” On the other hand, a lack of adverse events does not necessarily dictate risk classification, and she made the case that this could demonstrate that special controls often have the desired effect on device safety profiles.
Johnston pointed out a few seeming disparities, including that the document lists a proposal to classify respirator masks with antimicrobial agents as class I products while omitting regular surgical masks. These latter products, which appear in the 510(k) database as procode FXX, are seemingly a natural fit into a down-regulation move. There are other similar seeming incongruities, but Johnston said that regardless of the authorship of the policy, it is appropriate to remove at least some of these useless regulatory barriers, particularly given that the pandemic seems to have lost little momentum in recent months.