FDA Asked the Question It Already Answered — Then Deleted the Answer
On January 7, FDA withdrew the SaMD Clinical Evaluation guidance. On March 31, FDA asked the public what guidance is needed for validating digital health technologies.
On March 31, 2026, the FDA’s Center for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER) published a Request for Information asking stakeholders a pointed question: what guidance would support the use of digital health technologies (DHTs) in clinical investigations?
It’s a reasonable question. DHTs (e.g., wearables, sensors, AI-driven measurement tools, mobile apps, etc.) are increasingly used as a method of data collection in clinical trials. Getting their validation right is mandatory or trials have to trash the data. Endpoint integrity, participant safety, and regulatory defensibility all depend on that data integrity.
Here’s the problem:
On January 7, 2026, just a few months before publishing that RFI, FDA quietly withdrew its guidance document titled Software as a Medical Device (SaMD): Clinical Evaluation. No press release. No public comment period. No replacement document. Just gone.
That guidance, originally issued in 2017 and built on internationally harmonized IMDRF principles, provided exactly the kind of structured framework for validating software-driven clinical tools (i.e., DHT) that FDA is now asking the public to help reconstruct. It addressed valid clinical association, analytical validation, and clinical validation. These are the three pillars of demonstrating that a software output means what it claims to mean in a clinical context. Which is, not coincidentally, precisely what you need when a DHT is generating endpoints in a drug trial.
While the SaMD Clinical Evaluation guidance applied to device regulation (not drug trial oversight), through this March RFI, CDER and CBER are asking about DHTs used in drug and biologics development. These are not identical regulatory lanes but the underlying validation problem (i.e., how do you know this technology is measuring what it claims to measure, reliably, in a clinical population?) is THE SAME QUESTION. And now that answer was just discarded by the FDA…
What’s The Big Deal?
What makes this particularly difficult to defend is the timing. The January guidance changes were framed publicly by Commissioner Makary as designed to “cut unnecessary regulation and promote innovation.”
For those of you who actually read the updated CDS and General Wellness guidances, I’m sure you’d agree that they largely confirmed existing regulatory interpretations. They did NOT open new pathways. Rather, clarified the edges of what was already understood. That’s why you don’t see any webinars or changes in practice…
To be fair, there were some changes: the removal of QMS requirements for certain general wellness devices. That was the most tangible change, and it introduced ambiguity rather than clarity: validation is still expected, but the framework for what counts as “adequate validation” is now thinner. And THAT is what this latest RFI feels like a joke.
Withdrawing the SaMD Clinical Evaluation guidance is not the same as reducing regulatory burden. It removes shared vocabulary, reduces predictability, and leaves developers, including those building DHTs for clinical investigations, without a defensible reference point (which is exactly what the RFI is now asking for…)
The innovation that actually suffers is the kind that requires upfront investment in rigorous validation: the kind that produces trustworthy clinical endpoints.
The March 31 RFI asks four questions. One of them is: What areas of guidance would support the use of DHTs in clinical investigations? The honest answer, at this moment, includes: restore or replace the clinical evaluation framework you withdrew in January!
For the IRB/HRPP Community, This Matters.
When DHTs generate data used in clinical investigations involving human participants, the adequacy of that technology’s validation is a research ethics question, not just a regulatory one. If the evidentiary standard for a DHT endpoint is unclear (because there is no coherent framework for demonstrating valid clinical association and analytical reliability) then the risk-benefit assessment for participants enrolled in studies relying on those endpoints is also unclear.
The comment period for Docket No. FDA-2026-N-2476 closes June 1, 2026. If you work in digital health, clinical research, or research ethics governance, this is a consequential opportunity. The field has a chance to make visible the contradiction between what FDA removed in January and what it is asking for now, and to argue, on the record, for the kind of stable, principled guidance framework that serves participants, developers, and the integrity of clinical evidence alike.
**If you want to see the comments I submitted to the FDA, DM/email me. You know how to reach me. :)
FDA created the vacuum. The public comment process is one mechanism for demanding it be filled.