Episode 3: The Plot Thickens
Devices are a quite different story from drugs. While an adhesive bandage is used to treat a medical condition, it’s difficult to imagine a case where it should receive the same scrutiny as an implantable defibrillator that shocks the heart. Hence regulators divide devices into three to four risk categories, ranging from Class I (low risk) to Class III (high risk) (the EU, Australia, and other nations further parse devices into Class IIa and IIb, while Canada uses Class I through IV). Low risk devices are generally those that do not support life and whose failure would not represent a major health hazard. High risk devices are those that support life and/or whose failure or misuse could cause death or major injury. In between are those devices that offer varying levels of risk, generally requiring physician oversight to deliver and operate, but where failure or misuse would not be catastrophic.
While we explore the methods by which devices are approved for sale, let’s also take a moment to put a harmful confusion to rest. In 2012, a certain consumer-products oriented service claimed that,
“For most implants and other high-risk devices brought to market, manufacturers do nothing more than file some paperwork and pay the Food and Drug Administration a user fee of roughly $4,000 to start selling a product that can rack up many millions of dollars in revenue. Often, the only safety “testing” that occurs is in the bodies of unsuspecting patients—including two of the three people whose stories are told in this report.”
This statement is in fact completely incorrect. While Class II devices generally do not require human testing prior to sale, they do require extensive bench testing to demonstrate equivalence to an existing approved product. This is the FDA’s 510(k) pathway, named for the particular line of the Code of Federal Regulations that guide the process.
A 510(k) clearance requires the manufacturer to demonstrate that a device is “substantially equivalent” to an existing product in its indication, technical features, scientific principles, and performance. If these facts cannot be demonstrated with analytical tests, then the product moves on to Pre-Market Approval (PMA) testing, which generally requires clinical trials.
Substantial equivalence is demonstrated via such means as material analysis, computational modelling, mechanical testing, electromagnetic testing, biocompatibility testing (of the product and any expected wear debris), usability testing, and a host of other engineering and biological tools. This testing frequently requires a multi-million dollar investment and one to three years; certainly this exceeds doing “nothing more than [filing] some paperwork”.
The 510(k) filing fee is $5,018 ($2,509 for small businesses filing their first product), whereas new devices under PMA require one to two trials and a $250,895 ($62,724 for small businesses) filing fee. Both pathways require an annual Establishment Registration Fee of $3,646 for each manufacturing site.
The drug and device “User Fees” (the payments for FDA review) are set annually via a simple analysis: the FDA estimates how many 510(k) and PMA filings they will receive in the coming year and how many Establishments will be active, and divides their review budget among the groups. The fees are revised annually–sometimes up, sometimes down–and published in the Federal Register.
There are no market exclusivities available for medical devices. This is in part due to the shorter development cycle for devices, meaning patents offer effective protection for innovation. Devices can also be protected via trade secrets on manufacturing techniques and copyrights on software, offering further protection.
Mobile medical apps have received increasing scrutiny in the last few years, largely because of whether they present themselves as offering diagnostic information. Is an app that helps diabetic patients track glucose levels a diagnostic tool, or simply an updated form of a notebook? What if it takes that data and suggests how much to eat or how much insulin to take? Is the phone itself now a medical device? If so, can you guarantee that the app on my iPhone will offer the exact same conclusions as the one on my Galaxy SV, or my iPad? Will today’s software update provide a consistent result? What about the next OS version; will it change things? These might individually be trivial questions, but the validation, verification, and documentation activities they lead to are labor intensive and may be cost prohibitive for a company. What if I take my phone to Canada and use the app; does Health Canada now need to evaluate the safety and efficacy of my app? Mobile services present a unique jurisdictional challenge, and the law simply hasn’t caught up yet.
At present, the FDA has indicated that it will use “enforcement discretion” (i.e., it will agree to not investigate certain cases, based upon their ambiguous nature or the low health and safety risk they present). Apps with simple data logging functions will generally not be subject to FDA review, but those with diagnostic functions, wherein a health intervention is selected for the user, will be regulated and require review prior to sale.
Next week: Science!