During a House Health and Energy Subcommittee hearing on the U.S. Food and Drug Administration’s regulation of drugs, biologics, and medical devices on Wednesday, representatives questioned FDA directors.
Lawmakers said they are aiming to examine how the FDA prohibits doctors and patients from accessing tests that need to go through a lengthy pre-market approval process, and how the agency approaches assessments of artificial intelligence and machine learning-enabled devices. They also ask about third-party certification and other topics.
Some committee members also pressed FDA directors to improve communication with applicants.
Too slow for innovation
Health Subcommittee Chairman Brett Guthrie, R-Ky., said in opening remarks at a Congressional hearing Wednesday that some missteps caused “uncertainty among innovators.” on a bipartisan path.
“Repeated attacks on the expedited approval pathway from CMS, which is causing second-guessing among academics and payers alike, highlight that these approvals have not been proven or inferior to traditional assessment pathways will undermine public confidence.”
“The FDA cannot back down,” said Energy and Commerce Chairwoman Cathy McMorris Rodgers, R-Washington, but in 2023, “all three centers failed to meet their critical performance and goals.” recruitment despite funding reaching an all-time high.”
Stakeholders also reported that FDA staff are “disconnected and harder to work with than ever.”
“Everyone wants the FDA to succeed,” she added, but the committee wants to know what challenges the FDA is facing and why, and how Congress can help.
To attract more innovation, Dr. Jeff Shuren, director of the Center for Devices and Radiological Health, said the FDA launched the Breakthrough Devices Program that is currently designating 100 innovative medical devices each year. .
The program seeks to identify and provide patients and healthcare professionals access to devices that can improve treatment or diagnosis. The final guidance, issued in September, states that FDA may consider improving access to a device while determining whether it meets the first breakthrough standard.
However, most devices do not reach patients because there are a number of obstacles from idea to commercialization.
“It is appropriately called the valley of death,” he said.
A pilot program in which FDA consultants proactively and strategically engage to work with innovators and identify their challenges, such as reimbursement to payers, and Shuren said the response has been positive.
He told the committee that one company said it saved a year of development by participating.
Third-party AI certification required
According to Shuren, the FDA has also approved 900 AI/ML medical devices.
However, a third-party certification model is needed to spur innovation, and to do that, the agency needs help from Congress, he said.
Initially, the United States did not offer new technologies, and the agency made changes to attract innovation, said Shuren, CDRH director for 15 years.
But by 2023, he said the number of new technologies had “increased fivefold,” with 124 approvals (not including COVID-19 devices).
However, “For example, there are existing gaps related to the agency’s ability to effectively monitor the drug and medical device supply chain,” FDA witnesses said in written testimony their synthesis.
“Similarly, current regulatory pathways do not provide adequate flexibility for medical device manufacturers to adopt best practices for development, deployment or maintenance,” they said. continuous flow of AI/ML-enabled medical devices.”
To enable small innovators in the space, the FDA needs access to large data sets and is looking at federated models and learning from the Veterans Affairs agency, Shuren said.
To ensure medical device cybersecurity stays in check with security-by-design requirements, Shuren said, the FDA requires companies to design medical devices “in a way that allows it to can be patched”.
Balance LDT safety with innovation
Several committee members asked about the regulatory approach to laboratory-developed tests. The final rule, issued on May 6, affirms FDA’s position that LDTs are in vitro diagnostic products, regulated as medical devices under the Food, Drug, and Cosmetic Act Federal.
Guthrie questioned the ban on hospitals and patients; in an earlier statement, access to tests subject to “onerous requirements” under the new regime proposed by CDRH and Rodgers described the ruling as “an overreach by the executive” will affect how providers diagnose and care for patients.
“While the final rule is a small improvement over the proposed rule, it will still increase costs and reduce access to medical diagnostics and tests that provide information,” she said. important for doctors to treat their patients effectively.”
Rep. Diana DeGette, D-Colo., noted that LDT now uses sophisticated software that “informs critically important life decisions.”
While the final LDT Rule retains exceptions, as previous enforcement discretion factors phase out after four years, it lists important factors that could cause the diagnostic process to stall during the FDA review process. Guthrie cited 664 clinical inquiries in 2017-2021 and FDA policy of only responding in writing to applicants – no dialogue with applicants.
“That is unacceptable,” Guthrie said.
Notably, diagnostic tests and laboratory tests modified by new technology – such as the addition of AI or ML – will result in the need for pre-market review and compliance with regulatory requirements. FDA quality system requirements under the final rule.
Troy Balderson, R-Ohio, asked about facilitating the approval of AI/ML products that the FDA says it cannot accommodate and how the FDA can change that.
Shuren said the FDA has approved 900 AI/MNLs but shares that concern.
“Those specific functions fall under the category of lower-level diagnostics,” he explains.
“As we move into the current, AI generation [regulatory] framework is not fit for purpose, really designed around the hardware.”
Shuren says AI requires more post-market modeling so the FDA can track what’s happening to reduce the risk of bias.
He also said the amount of what will be produced using AI/ML will “skyrocket” and the agency will never have enough resources to review all of it.
“It’s unrealistic to expect that,” he said, and regulating the technology requires a third-party certification model. “But that will require help from Congress.”
Dr. Miller-Meeks accused CDRH of stifling innovation with the 500+ page rule and questioned why real-world evidence was sometimes used and sometimes ignored.
In a previous response, Shuren said that there are “problematic” tests that do not work well for patients and that they may not be treated properly.
“Today, if you are a cancer patient, whether you get the right treatment depends more on the lab you go to than on the biology of your tumor,” he said. “Unacceptable.”
To evaluate the models that underpin the algorithms, the FDA has the expertise but they don’t have the infrastructure, said Dr. Patricia Carvizzio, director of the FDA’s Center for Drug Evaluation and Research.
“We’re looking at a temporary fix,” but the FDA is looking at ways to upgrade for the future.
FDA asked about BOTTLE
Miller-Meeks also asked whether the FDA outsources certification to the Healthcare AI Alliance, saying it is not a diverse agency. She noted that Google and Microsoft are founding members, while the Mayo Clinic, which she said has more than 200 facilities implementing AI, employs some of its leaders.
“It doesn’t pass the smell test” and shows “clear signs of regulatory capture efforts,” she said.
Shuren responded that, although CDRH cooperates with CHAI as a federal liaison, as it does with other AI industry alliances, CDRH does not engage with the organization to evaluate applications.
“We also told CHAI that they need more representation in the health tech sector.” While his office may review CHAI’s work, he said, “they don’t work for us and we don’t work for them.”
Andrea Fox is a senior editor at Healthcare IT News.
Email: [email protected]
Healthcare IT News is a publication of HIMSS Media.
