Commissioner Scott Gottlieb of the U.S. Food and Drug Administration (“FDA”) stole the spotlight this past month when he delivered a speech discussing big promises from the agency regarding artificial intelligence (“AI”) in healthcare. “One of the most promising digital health tools is artificial intelligence, particularly efforts that use machine learning,” said Gottlieb when explaining that the FDA was “actively developing a new regulatory framework to promote innovation in this space.”
Indeed, this year the FDA has shown a lot of support for AI in healthcare by authorizing the marketing of several cutting-edge AI technologies—one of which was even permitted to be marketed without the requirement for additional clinician oversight.
In February, the FDA permitted marketing of clinical decision support software that uses algorithms to help neurovascular specialists arrive at answers more quickly and speed time to treatment for potential stroke patients.
Then in April, the FDA permitted marketing of the first device to use AI to detect a medical condition. Called IDx-DR, the device utilizes an AI algorithm to screen for diabetic retinopathy. This device is unique in that its results do not require additional review by a specialized clinician, which allows the test to be performed in a primary care setting. Like many devices in the digital health space, IDx-DR was reviewed through the FDA’s De Novo premarket review pathway, which is a way for new medical devices that present “a low to moderate risk to patients” and have no legally marketed predicate device on which to base a determination of substantial equivalence to be reclassified into Class I or Class II and avoid the need for a full Premarket Approval (“PMA”) application.
IDx-DR was also granted Breakthrough Device designation, which expedites the review of medical devices that provide for more effective treatment or diagnosis of life-threatening or irreversibly debilitating diseases or conditions. The FDA has consistently shared its intention to be flexible with digital health product developers whose software and devices do not fall neatly into the FDA’s well-established “product types.” Most recently, in May, the FDA permitted the marketing of OsteoDetect, an AI algorithm for aiding providers in the detection of wrist fractures by analysis of two-dimensional X-ray images.
These examples of successful utilization of the De Novo pathway and Breakthrough Devices program are likely to continue to encourage AI innovation by medical device and digital health companies by using these pathways to market as an FDA roadmap for marketing authorization of their own products.