FDA seeks input on clinical evaluation of software as a medical device

Register now

The Food and Drug Administration has released draft guidance for public comment on the clinical evaluation of software as a medical device (SaMD), with significant implications for the growing number of digital health and mobile applications entering the healthcare marketplace.

Developed by the International Medical Device Regulators Forum (IMDRF) and endorsed last month by IMDRF’s management committee, the FDA draft guidance provides recommendations for medical device manufacturers on the methods of clinical evaluation and the level of clinical evidence necessary to support the use of SaMD.

“The draft guidance is intended to provide globally harmonized principles of when and what type of clinical evaluation is appropriate based on the risk of the SaMD,” states the FDA’s announcement in the Federal Register.

"Based on the significant impact SaMD has on clinical outcomes and patient care, a SaMD manufacturer is expected to gather, analyze, and evaluate data, and develop evidence to demonstrate the assurance of safety, effectiveness and performance of the SaMD,” states the draft guidance.

While this is a document developed by the IMDRF, when finalized the guidance will represent FDA's current thinking on the topic of the clinical evaluation of standalone software. As such, Bradley Merrill Thompson, an attorney at Washington-based law firm of Epstein Becker Green who counsels medical device companies on regulatory issues, believes this guidance is a big deal.

“Digital health is an exploding field, so this may affect quite a few products down the road,” says Thompson. “There are a huge number of mobile apps as well as stand-alone programs for PCs that are finding their way into the practice of medicine, and this guidance lays out the requirements for clinical evaluation.”

Typically, clinical evaluation is one of the most expensive aspects of product development, “so the requirements of this guidance have a major impact on both the cost and the time frames for bringing products in these categories to market,” he says.

Also See: FDA eyes plan to lessen review burden for software updates

Nonetheless, Thompson contends that although the draft guidance “sheds light on the issue,” it is difficult to discern exactly what validation is required for a particular piece of software. In addition, he says while the document acknowledges the value of collecting real-world evidence, it doesn’t clearly articulate the circumstances when real-world evidence can replace pre-market clinical study.

“Overall, I am very intrigued by the approach to risk stratification that IMDRF has been pursuing. It is possibly a workable framework, but I think it does require some tweaking,” adds Thompson. “Among other things, when software merely informs clinical decision-making, in the U.S. that means the doctor remains firmly in control of the decision-making. By law, FDA is not supposed to regulate the practice of medicine. So some of the validation required for software that merely informs clinical decision-making seems inappropriate. Further, although the document acknowledges the important role of transparency in the software, I’m not sure the document makes full use of that concept.”

The FDA will accept public comments on the draft SaMD guidance until December 13. However, Thompson asserts that there is a problem with the regulatory agency’s approach, namely that the IMDRF—not the FDA—will finalize the document.

“The International Medical Device Regulators Forum frankly has no authority under U.S. law. There’s a problem with saying that this document—when finalized by an international body—will become U.S. law,” he contends. “Honestly I think if FDA wants to adopt the final IMDRF guidance as U.S. policy, the agency will need to publish in the future the final IMDRF guidance and solicit comment on it. I understand why the FDA wanted to solicit comments now, because this is an opportunity to influence the final IMDRF document, (but this may) require two different comment periods.”

For reprint and licensing requests for this article, click here.