Predicate Device Search: How to Find and Select Predicates for 510(k)

Predicate Device Search: How to Find and Select Predicates for 510(k)

Key Takeaways

What Is a Predicate Device

A predicate device is a legally marketed device to which a new device is compared in a 510(k) submission. The comparison is used to demonstrate that the new device is substantially equivalent (SE) to the predicate.

Legal Definition

Under Section 513(i) of the FD&C Act and 21 CFR 807.92(a)(3), a predicate device is a device that was legally marketed before May 28, 1976 (preamendment device), or a device that has been reclassified from Class III to Class II or I, or a device that has been found substantially equivalent through the 510(k) process, or a device that was classified through the De Novo process.

Types of Legally Marketed Devices That Can Serve as Predicates

Devices that can NOT serve as predicates:

• Devices marketed only outside the United States
• Devices cleared under enforcement discretion
• Devices that have been recalled and are no longer marketed
• PMA-approved devices (PMA devices are Class III and are not predicates for 510(k))
• Devices with expired or revoked 510(k) clearances

FDA Databases for Predicate Search

1. FDA 510(k) Premarket Notification Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm

This database contains all 510(k) clearance decisions. Search by:

• Applicant name: Search for known competitors or established manufacturers
• Device name: Search for devices with similar names or descriptions
• Product code: Search by the three-letter product code for the device type
• Decision date: Filter by date range
• 510(k) number: Search specific clearance numbers

What you get: 510(k) number, applicant, device name, product code, clearance date, decision, and a link to the 510(k) summary or statement (when available).

Limitation: The 510(k) summary provides only a brief description of the device and the substantial equivalence determination. Detailed technical information is not publicly available unless the submitter chose to include it.

2. FDA Product Classification Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpcd/classification.cfm

This database contains the classification regulation for each device type, including:

• Product code: Three-letter code identifying the device type
• Device class: Class I, II, or III
• Review panel: The FDA review panel responsible for the device type
• Regulation number: The 21 CFR section governing the device type
• Submission type: Required premarket submission (510(k), De Novo, PMA, exempt)
• Special controls: Reference to applicable special controls guidance

Use this to: Identify the correct product code for your device, understand the classification and regulatory requirements, and identify the applicable Special Controls that your device must meet.

3. AccessGUDID (Global Unique Device Identification Database)

URL: https://accessgudid.nlm.nih.gov/

AccessGUDID contains device identification information submitted under the Unique Device Identification (UDI) system required by 21 CFR Part 830. Search by:

• Device name or description
• Company name
• Product code
• GUDID number

What you get: Device description, company name, product code, device classification, and sometimes more detailed device characteristics than the 510(k) database provides.

4. FDA De Novo Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/denovo.cfm

Contains all De Novo classification decisions. De Novo-classified devices can serve as predicates for 510(k) submissions. The De Novo decision summary typically includes more detailed information than a 510(k) summary, including the Special Controls that were established.

5. FDA Total Product Life Cycle (TPLC) Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfTPLC/tplc.cfm

The TPLC database links premarket submissions to post-market data, including recalls, adverse events, and compliance actions. Use this to evaluate the regulatory history of a potential predicate, including whether the predicate has been subject to recalls or safety issues.

6. FDA MAUDE (Manufacturer and User Facility Device Experience) Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm

Contains adverse event reports for medical devices. While not a primary predicate search tool, MAUDE can provide insight into the safety profile of a potential predicate, which is relevant for your risk analysis and SE argument.

Predicate Search Strategy

Phase 1: Identify the Product Code

Start by determining the correct product code for your device:

1. Search the Product Classification Database using keywords from your device description
2. Review the classification regulation and panel assignment for candidate product codes
3. Confirm the product code matches your device's intended use and technological characteristics
4. If no product code matches, the device may require De Novo classification

Phase 2: Search for Cleared Devices

Using the identified product code(s):

1. Search the 510(k) database by product code to retrieve all cleared devices in that category
2. Filter by date (recent clearances may be more relevant as predicates)
3. Review 510(k) summaries for cleared devices to understand intended use, technological characteristics, and testing performed
4. Search the De Novo database for the same product code to identify De Novo-classified devices

Phase 3: Evaluate Candidate Predicates

For each candidate predicate, evaluate:

Phase 4: Select the Primary Predicate

The strongest predicate is the one with the fewest differences from your device. Prioritize:

1. Same intended use (mandatory)
2. Same technological characteristics (strongly preferred)
3. Recent clearance date (indicates current regulatory expectations)
4. Detailed 510(k) summary available (easier to make the SE argument)
5. Clean regulatory history (no recalls, warning letters, or significant adverse events)

Substantial Equivalence Comparison

The SE Determination Framework

FDA's guidance "The 510(k) Program: Evaluating Substantial Equivalence in Premarket Notifications" (July 2014) describes the algorithm FDA uses:

Step 1: Does the new device have the same intended use as the predicate?

• Yes: Proceed to Step 2
• No: Device is Not Substantially Equivalent (NSE)

Step 2: Does the new device have the same technological characteristics as the predicate?

• Yes: Device is SE. Submit performance data confirming equivalent performance.
• No: Proceed to Step 3

Step 3: Do the different technological characteristics raise different questions of safety and effectiveness?

• No, and performance data demonstrates the device is at least as safe and effective: Device is SE
• Yes, or performance data is insufficient: Device is NSE

Intended Use Comparison

The intended use comparison is the first and most critical element. FDA defines intended use broadly to include:

• The disease, condition, or purpose for which the device is used
• The patient population
• The body part or body system
• The clinical setting

The new device's intended use must be the same as the predicate's intended use. "Similar" or "overlapping" intended use is insufficient. If the new device has a broader intended use than the predicate (e.g., the predicate is cleared for adult use and the new device claims adult and pediatric use), the expanded claim is not covered by the SE determination and may require additional clinical data or a separate submission.

Technological Characteristics Comparison

The comparison table is the core of the 510(k) submission. A well-constructed comparison table should include:

For each "Different" entry, the submission must explain why the difference does not raise new questions of safety and effectiveness, supported by performance data.

Split Predicates

What Is a Split Predicate

A split predicate approach uses two or more predicate devices in a single 510(k) submission: one predicate to support the intended use and a different predicate to support the technological characteristics. This approach is used when no single predicate has both the same intended use and similar technology.

FDA's Position on Split Predicates

FDA's guidance on substantial equivalence does not prohibit split predicates, but the agency has expressed caution. The key concern is that combining the intended use of one device with the technology of another may create a novel combination that was never evaluated as a complete device.

FDA evaluates split predicate submissions on a case-by-case basis, considering:

1. Whether the combination of intended use and technology raises new safety or effectiveness questions
2. Whether performance data adequately addresses the novel combination
3. Whether the split predicate approach is being used to circumvent classification requirements

When Split Predicates Are Acceptable

Best Practices for Split Predicates

1. Use a Pre-Submission meeting to confirm FDA's acceptance of the split predicate approach before filing
2. Provide a comprehensive comparison showing that the combination does not raise new questions
3. Include robust performance data demonstrating safety and effectiveness of the combined intended use and technology
4. Address the gap: explicitly discuss why no single predicate covers both intended use and technology

Multiple Predicates

Using More Than One Predicate (Not Split)

A 510(k) may reference multiple predicate devices to support different aspects of the SE argument, even when one primary predicate covers the main intended use and technology. Additional predicates (sometimes called "reference devices") can support:

• Specific technological features
• Specific performance claims
• Biocompatibility of specific materials
• Safety of a particular design element

This is distinct from a split predicate approach because the primary predicate covers the intended use, and additional predicates provide supplementary support.

Predicate Chains

When the primary predicate is itself a 510(k)-cleared device, it has its own predicate. This creates a predicate chain. FDA accepts predicate chains, but there are limitations:

• The predicate must be currently legally marketed (not recalled without replacement, not with revoked clearance)
• Long predicate chains may raise concerns about cumulative technological drift from the original classification basis
• FDA has expressed concern about "predicate creep," where successive 510(k) clearances gradually expand the intended use or technology beyond the original classification

Common Pitfalls in Predicate Selection

Pitfall 1: Choosing a Predicate with a Different Intended Use

The most common cause of 510(k) failure. If the predicate's intended use does not match the new device, the 510(k) will receive an NSE determination. Always verify the predicate's intended use from the actual 510(k) clearance letter or summary, not from the manufacturer's marketing materials.

Pitfall 2: Relying on a Recalled Predicate

A predicate that has been recalled and is no longer legally marketed cannot serve as a predicate. Check the FDA recall database and TPLC database before selecting a predicate.

Pitfall 3: Using a PMA-Approved Device as a Predicate

PMA-approved devices (Class III) cannot serve as predicates for 510(k) submissions. The 510(k) pathway is available only for Class I and II devices.

Pitfall 4: Insufficient Predicate Research

Selecting a predicate based solely on device name or product description without reviewing the 510(k) summary can lead to mismatches. Always read the full 510(k) summary or statement to understand the predicate's intended use, technological characteristics, and the testing that was performed.

Pitfall 5: Ignoring the Predicate's Special Controls

If the predicate was classified under specific Special Controls (e.g., through a De Novo order), the new device must also comply with those Special Controls. Failure to address Special Controls in the 510(k) can result in a Refuse to Accept determination.

Pitfall 6: Predicate Creep

Using a predicate that itself was cleared with a stretched SE argument. Each link in the predicate chain introduces risk. If FDA becomes concerned that the predicate chain has drifted too far from the original classification basis, they may issue an NSE determination even if the immediate predicate comparison is reasonable.

Pitfall 7: Selecting a Predicate from a Different Product Code

Using a predicate with a different product code is not automatically disqualifying, but it raises a flag. Different product codes may indicate different regulatory requirements, Special Controls, or device classifications. If you use a cross-product-code predicate, you must clearly justify why the comparison is appropriate.

Predicate Search Documentation

What to Document

Maintain a predicate search report that documents:

Why Documentation Matters

Predicate search documentation serves multiple purposes:

1. Regulatory submission: The 510(k) must describe the predicate selection rationale
2. FDA review: Reviewers may ask why a particular predicate was chosen over alternatives
3. Design controls: Predicate search is part of the design input process under 21 CFR 820.30
4. Audit readiness: Quality system audits may examine predicate selection as part of design control verification

Key Regulatory References

References