Predicate Device Selection: The 510(k) Decision That Makes or Breaks Your Submission

Of all the strategic decisions in a 510(k) submission, predicate selection is the one that most directly determines your outcome. Choose the right predicate and you define a clear, defensible path to clearance. Choose the wrong one and you may spend months responding to deficiency letters — or worse, withdraw and start over. In 20 years of FDA submissions, this is the single decision I see most often made poorly by first-time submitters.

What "Substantial Equivalence" Actually Requires

A 510(k) clears a device by establishing substantial equivalence to a predicate device — a legally marketed device cleared through 510(k) or on the market before May 28, 1976 (a "preamendments" device). Under 21 CFR 807.100, substantial equivalence requires that your device has the same intended use as the predicate AND either has the same technological characteristics, or has different technological characteristics that do not raise new questions of safety and effectiveness.

That second prong — different technological characteristics that don't raise new questions — is where most strategy lives. The goal is not necessarily to find a predicate that looks exactly like your device. The goal is to find a predicate that makes your differences look manageable.

The Predicate Selection Framework

Systematic predicate selection starts with FDA's 510(k) database (accessible at accessdata.fda.gov). Run searches by product code, device name, and intended use. You are looking for cleared devices that share your device's clinical purpose and patient population. Build a list of candidates — typically 10 to 20 — before narrowing down.

For each candidate predicate, evaluate:

• Intended use alignment: Does the predicate's cleared intended use statement match what your device will do? A single word difference in intended use — "diagnoses" versus "aids in diagnosing" — can turn substantial equivalence into a non-equivalence determination.
• Technological characteristics delta: What is different about your device? Materials, software, energy type, delivery mechanism? The narrower the delta, the cleaner the submission.
• Clearance history: Has FDA cleared devices with your specific technological differences using this predicate before? If there is a clearance chain showing progressive acceptance of your technology, that is powerful evidence of substantial equivalence.
• Special controls and performance standards: Does the predicate have associated performance standards or special controls that your device must also meet? Failing to test to those standards is a near-automatic deficiency.

Split Predicates: When One Is Not Enough

FDA permits the use of split predicates — using one predicate for intended use and a different predicate for technological characteristics. This is a legitimate strategy when no single predicate covers both dimensions. However, FDA has become increasingly cautious about split predicate submissions. The two predicates must both be within the same device type, and you must demonstrate that the combination does not create a novel device that raises new safety questions.

If your device requires a split predicate, a Pre-Submission meeting with FDA before filing is strongly advisable. This lets you test whether FDA agrees that your predicate combination is defensible before you invest in full testing.

Why Recent Predicates Are Usually Better

All other things being equal, a more recently cleared predicate is typically a stronger choice than an older one. Recent predicates reflect current FDA expectations for performance data, labeling, and special controls. If your predicate was cleared in 2004 and FDA's expectations for that device type have evolved significantly, you may find yourself arguing that your 2024 device meets the standards of a 2004 review — a position that FDA reviewers are unlikely to accept.

The exception: if an older predicate has a broader intended use statement that covers your device more cleanly than any recent clearance, it can be worth accepting the higher performance data burden in exchange for the cleaner intended use match.

The 510(k) Predicate Mistake That Triggers Most Deficiencies

The most common predicate mistake is selecting a predicate based on the device name rather than the cleared intended use. Two devices can have identical names and completely different cleared intended uses. Always read the original 510(k) summary and, when possible, the decision memo. FDA's clearance decision is based on the exact language in the submission — not on the product name in the database.

When There Is No Good Predicate: De Novo as the Answer

If no predicate reasonably supports a substantial equivalence argument, the De Novo pathway may be your best option. De Novo creates a new device type classification and establishes the first predicate for future 510(k) submitters. It requires more robust clinical evidence than a standard 510(k) and takes longer — but it is significantly faster and less expensive than a PMA, and it positions your company as the regulatory benchmark for a new device category.