top of page
Search

FDA's Final Human Factors Guidance Adds a Fourth Decision Point. Here's What Changes.

  • Writer: Sharmila Bhatt
    Sharmila Bhatt
  • Jul 14
  • 3 min read

For years, one of the more frustrating gray areas in device submissions has been human factors: the FDA has long requested HF information in premarket submissions, but for 510(k)s especially, it hasn't always been clear which device types actually needed HF validation testing included. On May 29, 2026, FDA finalized guidance intended to close that gap — Content of Human Factors Information in Medical Device Marketing Submissions — replacing the draft version it issued back in December 2022.


What stayed the same. The Final Guidance keeps the same basic structure as the draft: a risk-based flowchart that sorts submissions into three HF Submission Categories, each requiring a different depth of documentation.

  • Category 1 (modified devices only): a conclusion and high-level summary of the HF evaluation.

  • Category 2: a rationale for why there are no critical tasks (new devices) or no new/impacted critical tasks (modified devices).

  • Category 3: a full HF engineering/usability engineering report, including HF validation testing data.


What actually changed. The flowchart itself expanded. In the draft, a "yes" answer at Decision Point C — whether a use-related risk analysis identifies critical tasks — routed a submission straight into Category 3, the most burdensome tier. Industry had flagged this as sweeping too many submissions into full validation testing. The Final Guidance responds by adding a new fourth decision point, Decision Point D, which asks submitters to weigh three additional factors before concluding that HF validation data needs to go into the submission itself:

  • The user interface's history of use for the intended use, users, and environment(s);

  • The complexity of the user interface (FDA specifically calls out interfaces involving programming, monitoring, or maintenance, or with many interdependent steps);

  • Whether existing risk control measures are adequate.

In other words, a device with a well-established interface and solid existing risk controls may not need to submit full validation data, even if critical tasks exist — a meaningful narrowing from the draft.


The catch. FDA pairs Decision Point D with a notable footnote: even when validation data doesn't need to be submitted, the agency expects HF information to be maintained by the manufacturer as part of Quality Management System Regulation (QMSR) design verification and validation obligations — and reviewers may still ask for underlying study results during the course of review, or during a QMSR inspection. Decision Point D may reduce submission paperwork; it doesn't reduce the underlying documentation burden. Teams that read this as "we can skip the HF study" are reading it wrong — the more accurate read is "we may not have to attach it, but we'd better be able to produce it on request."


What to do next. Three concrete steps:

  1. If your team is unsure which HF Submission Category applies, FDA points submitters toward the pre-submission program rather than guessing — a Q-Sub is cheaper than a deficiency letter.

  2. FDA has already updated the eSTAR templates to require an HF Submission Category selection on every applicable submission, so this isn't an optional consideration anymore — it's a mandatory field.

  3. FDA is holding a Town Hall on the Final Guidance on July 22, 2026. For any team with a submission in the pipeline that touches human factors, that's worth attending live rather than reading the transcript later.

FDA has indicated it generally won't expect submissions received before August 1, 2026 to reflect the newly recommended content — but will review it if teams choose to include it. That's a narrow window for anything already in progress.


A note on tooling. This is exactly the kind of change that's easy to miss if guidance tracking is a manual, periodic process rather than a continuous one — the Final Guidance published seven weeks ago, and the Town Hall is one week out. Lexim Sphere's regulatory intelligence engines are built to surface changes like this as they publish and flag which submission strategies they touch, rather than relying on someone catching it during the next review cycle.


Sources

Comments

bottom of page