Clinical Evidence Requirements for De Novo Classification: What Device Makers Need to Know

De Novo Classification and the Clinical Evidence Question

If your medical device is novel — meaning there is no legally marketed predicate to support a 510(k) — the De Novo classification process under 21 CFR Part 860 Subpart D is often your most direct path to market. But one of the most consequential questions we hear from startup founders and regulatory teams alike is this: How much clinical evidence does FDA actually require for a De Novo request?

The honest answer is: it depends — and that ambiguity is exactly where companies lose time, money, and momentum. This post will break down FDA's clinical evidence expectations for De Novo submissions, grounded in the applicable regulations and guidance, so you can build a defensible strategy from day one.

The Regulatory Framework: What FDA Is Actually Asking For

The De Novo process is codified in Section 513(f)(2) of the Federal Food, Drug, and Cosmetic Act (FD&C Act) and implemented through 21 CFR 860.257. Under this pathway, FDA evaluates whether a novel, low-to-moderate risk device (Class I or II) has demonstrated reasonable assurance of safety and effectiveness — the same evidentiary standard that governs all device classification decisions.

FDA's primary guidance on submissions is the 2021 De Novo Classification Process Guidance. It specifies that a De Novo request must include a description of the device, proposed special controls, and performance data sufficient to support the classification. That last element — performance data — is where clinical evidence enters the picture.

Importantly, FDA does not mandate clinical data for every De Novo. The agency evaluates the totality of evidence, which can include bench testing, biocompatibility data, software verification and validation, usability studies, and clinical data. What drives the clinical evidence threshold is the device's intended use, user population, and risk profile.

When Clinical Data Is Expected

While bench and non-clinical testing may suffice for lower-risk devices with well-characterized performance parameters, FDA will generally expect clinical evidence when:

• The device makes diagnostic or therapeutic claims that cannot be validated through bench testing alone
• The intended use involves vulnerable populations (pediatric, elderly, or immunocompromised patients)
• There is no reliable in vitro or animal model that predicts human performance
• The device interfaces directly with the human body in ways that introduce novel biological or physiological risks
• The proposed special controls include a clinical performance requirement that must be prospectively validated

For software-based devices, particularly AI/ML-driven SaMD seeking De Novo classification, FDA increasingly expects clinical validation data that demonstrates the algorithm performs as intended across the target population — not just in controlled development datasets. Refer to FDA's 2021 AI/ML-Based SaMD Action Plan and the Predetermined Change Control Plan guidance for context on how clinical evidence intersects with software lifecycle requirements.

Study Design Considerations: Fit-for-Purpose Evidence

A common mistake we see is companies defaulting to full randomized controlled trial (RCT) designs when a leaner, well-designed study would satisfy FDA's evidentiary standard. FDA's Use of Real-World Evidence to Support Regulatory Decision-Making (2017) and the Considerations for the Design, Development, and Analytical Procedures for Clinical Investigations of Medical Devices (2020) guidance both reinforce a fit-for-purpose approach.

For De Novo, this often means:

• Prospective, single-arm studies with pre-specified performance endpoints and statistically justified sample sizes
• Objective Performance Criteria (OPCs) or Performance Goals derived from published literature or prior FDA decisions — avoiding the need for a concurrent control arm
• Pivotal feasibility data combined with robust post-market surveillance commitments as a condition of the De Novo order
• Human factors and usability validation studies that meet the evidentiary bar outlined in FDA's 2016 Human Factors guidance (FDA-HF-G-7)

The key is that your clinical study protocol — and the endpoints you select — must directly map to the safety and effectiveness questions FDA will ask during review. If there is misalignment between your study design and FDA's review criteria, you will receive an Additional Information (AI) request that can add six to twelve months to your timeline.

Pre-Submission Meetings: Your Most Underutilized Tool

Before you finalize your clinical study protocol, request a Pre-Submission (Q-Sub) meeting with FDA under the framework described in the 2023 Q-Submission Program Guidance. This mechanism allows you to present your proposed clinical study design, endpoints, and statistical analysis plan to the reviewing division for direct feedback — before you enroll a single subject.

Pre-Subs are not a guarantee of approval, but they dramatically reduce the risk of designing a study that misses FDA's evidentiary expectations. For novel devices with no regulatory precedent, this step is not optional — it is essential due diligence.

Special Controls and Post-Market Clinical Commitments

One often-overlooked aspect of De Novo is that FDA may grant classification while requiring post-market clinical data as a special control. This means your De Novo order may include a condition that you conduct a post-market study to confirm long-term safety or real-world effectiveness. Understanding this possibility upfront allows you to build your clinical roadmap — and budget — accordingly.

Build Your Clinical Evidence Strategy Before You Build Your Study

The De Novo pathway offers a legitimate, efficient route to market for innovative devices — but only when the clinical evidence strategy is designed with FDA's expectations in mind. Misaligned study designs, underpowered endpoints, and missing pre-submission engagement are the most common reasons De Novo requests stall or receive refuse-to-accept decisions.

At ADB Consulting & CRO Inc., we help medical device companies design clinical evidence strategies that are scientifically sound, regulatory-defensible, and commercially realistic. Whether you are evaluating your pathway options or preparing your De Novo submission package, we bring the regulatory depth and hands-on FDA experience to get it right the first time.

Ready to assess your clinical evidence requirements for De Novo? Book a free discovery call with Andre Butler and the ADB Consulting team at adbccro.com today.