Using Multiple Predicates in a 510(k) Submission

FDA allows — and in many cases expects — 510(k) submissions to cite more than one predicate device. Multiple predicates are appropriate when no single cleared device encompasses all of the features, intended uses, or technological characteristics of your device. Understanding how to structure a multi-predicate argument is essential for devices that combine established features in new configurations.

When a single predicate isn't enough

Most devices that combine two established functions, add a new measurement modality, or serve a broader patient population than any single cleared predicate will benefit from a multi-predicate approach. Common scenarios:

• Your device measures two physiological parameters — one predicate covers each
• Your device has a therapeutic function and a monitoring function — split predicates address each
• Your device is intended for both adult and pediatric populations — one predicate covers each indication
• A feature of your device (e.g. wireless connectivity, specific material) is only present in predicates from a different product code

Split predicates: the most common multi-predicate approach

A split predicate argument uses one predicate to establish intended use equivalence and a separate predicate (or predicates) to justify specific technological characteristics. FDA's guidance on the 510(k) program explicitly acknowledges this approach.

Example: A wearable ECG monitor with a novel electrode material might cite:

• Predicate A (K221234) — for intended use: ambulatory ECG monitoring
• Predicate B (K190567) — for the specific electrode material, which was cleared for a different monitoring device but demonstrates the material's biocompatibility and electrical performance

The key requirement: you must demonstrate that combining features from multiple predicates does not create a device that raises new safety or effectiveness questions beyond what either predicate individually addressed.

How FDA evaluates multi-predicate submissions

FDA applies the same substantial equivalence standard to multi-predicate submissions. The reviewer will assess whether each predicate is appropriate for the feature or use it supports, whether the combination creates any new risks not addressed by either predicate, and whether the performance data covers any gaps created by differences from all predicates.

A multi-predicate submission with weak predicate selection for any individual feature is more likely to generate an AI request than a single-predicate submission, because FDA must evaluate each predicate relationship separately. Each link in the argument must be defensible.

Predicate combination red flags

• Mixing device classes — combining a Class I predicate and a Class II predicate in a way that implies your device should be Class I when FDA would classify it as Class II
• Cherry-picking characteristics — using predicate A only for the characteristics where it matches and ignoring the ones where it differs; FDA reviewers notice this
• Predicates from unrelated product codes — using a predicate from a completely different device category to justify a single feature, without a clear rationale for why that predicate is relevant
• Outdated predicates — using a predicate cleared under superseded standards when more recent clearances exist that reflect updated requirements

How to find the right combination of predicates

Start by identifying the distinct feature clusters of your device — intended use, each major technological module, any novel materials or connectivity. For each cluster, run separate predicate searches to find cleared devices that best represent that feature in isolation.

Once you have candidate predicates for each feature, compare them to each other: are they consistent in terms of device class, regulatory basis, and applicable standards? A set of predicates that are internally consistent makes for a much cleaner Section 4 comparison table.