In April, the US Food and Drug Administration (FDA) released a warning to the public, advising them not to use unapproved medical devices or apps for diagnosing concussions.
These tools – which include smartphone apps marketed to sports coaches – could lead to an incorrect diagnosis, which in the case of head injuries is a serious problem. To date, only a few medical devices have been approved for diagnosing concussions, and none are targeted directly at the consumer. All require an evaluation from a healthcare professional.
“I want to be clear, there are currently no devices to aid in assessing concussion that should be used by consumers on their own,” said Jeffrey Shuren, director of the FDA’s Center for Devices and Radiological Health, on a release. “Products being marketed for the assessment, diagnosis, or management of a head injury, including concussion, that have not been approved or cleared by the FDA are in violation of the law.”
The FDA said it would contact any companies that were flouting these rules, and would take action if its concerns weren’t addressed. Given the potential severity of head injuries, it’s easy to see why this warning was necessary.
“Any product in the United States that diagnoses a patient is always going to need regulation from the FDA,” says Michael Murphy, president of Conatus Consulting, a medical device regulatory consulting firm. “So if you have an app that counts calories or something, that’s not a product that would get regulated, but an app that determines concussions would be. That’s a serious medical situation and if there were some flaw in the app that could result in a misdiagnosis, it could endanger a person’s life.”
He points out that, while this distinction seems relatively clear cut, some medical devices companies do get confused about which regulatory pathway they need to follow.
“In the US, companies have the opportunity to request an informal meeting with the FDA, where you can go in and talk to them about a product and they’ll tell you whether it needs to be regulated or not,” he says. “Again, that seems pretty straightforward, but it can be quite intimidating for new companies.”
Failures in oversight
After all, this is not the first time potentially dangerous medical devices have hit the market. According to FDA data, over the past decade there have been more than 80,000 deaths and 1.7 million injuries possibly linked to medical devices. These have included everything from faulty defibrillators to rupturing breast implants.
Last year, the International Consortium of Investigative Journalists (ICIJ) conducted a major investigation into the issue. As well as bringing to light some harrowing personal stories, they also highlighted a number of failures in the FDA’s oversight.
For instance, they found that breast implant companies were able to bury evidence of harm, by keeping injury reports hidden from the public. (When this practice ended in 2017, the number of injury reports soared.) They also found that today’s faster approval times may be associated with faultier devices.
The complexity lies in the fact that medical devices are essential products that can save patients’ lives. If the approvals process is too protracted or expensive, these products may face delays in coming to market or may not come to market at all. On the other hand, regulation is there for a very good reason.
Laws and loopholes
In the US, the vast majority of medical devices win approval under a provision known as the 510(k) pathway. In this case, the device simply needs to show it has ‘substantial equivalence’ to a previously cleared ‘predicate device’, and clinical studies aren’t required.
“The 510(k) programme is fairly straightforward and it doesn’t cost a huge amount of money to get through,” says Murphy. “As a comparison, trying to get an oncology test through the FDA requires what we call a PMA – pre market approval – and that cost is typically ten times greater than the 510(k) pathway. If you don’t have a predicate device, a lot of the time that would force the company to abandon the product because they don’t want to spend the time and money trying to get a PMA.”
Although the 510(k) programme has long been the gold standard for safety, in November 2018 the FDA announced that it plans to modernise the framework. For instance, it may start recommending that companies avoid predicates that are less than ten years old. It has also discussed closing a regulatory loophole that allows lab tests to bypass regulation.
The loophole in question, referred to as the Laboratory Developed Test (LDT) loophole, means diagnostics are not regulated in the same way as medical devices. Rather than requiring FDA approval, they are regulated under the Clinical Laboratory Improvement Amendments (CLIA), and their accuracy isn’t monitored.
The loophole exists because research hospitals often need to modify commercial tests. But it also applies to any company that develops and conducts a diagnostic test in its own lab – and that is where the controversy comes in.
In recent years, there have been some high-profile stories about what can happen when these tests go wrong. A 2015 FDA report mentioned 20 LDTs that had caused harm to patients, including a genetic breast cancer test that was wrong 20% of the time. Most notoriously, the blood-testing start-up Theranos was shut down – and its founder, Elizabeth Holmes, charged with criminal fraud – after its tests failed to perform as indicated.
The FDA has been trying to develop a new framework for LDTs since 2013, and in December 2018 it proposed some specific reforms. Under the new paradigm, around 10% of diagnostics would require premarket review and 40-50% would be regulated via a pre-certification programme. The FDA would also have the authority to ‘take action against any problematic tests’.
“It’s really fairly recently that the FDA decided that they needed to play a role in oversight of laboratory developed tests,” says Murphy. “They have stated publicly that they have evidence of erroneous results coming out of labs that had an effect on the patients, and their other claim is that the volume of this kind of testing has increased substantially over the years and a lack of oversight is potentially going to involve public safety.”
A controversial topic
Not everyone agrees that this change is necessary. As Murphy points out, many diagnostics companies believe CLIA’s oversight is sufficient (and he himself describes the FDA’s claims as “pretty sketchy”).
“One of the problems is people don’t believe there’s a problem, so they’re having a hard time agreeing to let the FDA come in and have oversight over testing that for the most part has been done very well,” he says. “My opinion is that they need to substantially improve the scope and veracity of the current CLIA programmes and not try to reinvent the wheel.”
In particular, he believes that CLIA ought to focus more attention on the quality and reproducibility of the tests, and that labs should be externally audited.
“What’s lacking is the ability of someone to go in and audit those laboratories and make sure they are following the same standards as diagnostic kit manufacturers,” he says.
This debate is sure to continue. While the goal is clear – balancing patient safety against continued innovation – LDTs are a rapidly evolving technology and there are no simple answers about the best approach to take.
In a broader sense, though, we know that illegally marketed or poorly tested devices can put lives at risk. As such, any tightening of standards could assuage people’s concerns.
“Part of our commitment to protect patients is to recognize that all devices pose benefits and risks, and to seek ways to minimise those risks. We’re seizing the chance to make a generational change in the framework for 510(k) reviews – one that advances its best features while making sure it keeps pace with modern innovation,” said the FDA’s Scott Gottlieb and Jeffrey Shuren in a statement.