When Clinical Data is Needed to Support Substantial Equivalence for 510k Submissions (DRAFT)

This is a draft guidance. Not for implementation.

What You Need to Know? 👇

What are the four main scenarios when clinical data may be necessary for 510(k) submissions?

The four scenarios are: (1) differences in indications for use between new and predicate devices, (2) differences in technological characteristics, (3) when substantial equivalence cannot be determined by non-clinical testing alone, and (4) newly identified or increased risks for the predicate device.

When can real-world data be used as clinical evidence in 510(k) submissions?

Real-world data from registries, adverse event databases, medical records, and electronic health records can constitute valid scientific evidence if it demonstrates relevance and reliability. The data must meet the standards defined in 21 CFR 860.7(c)(2) for valid scientific evidence.

Does expanding indications for use to a higher-risk patient population always require clinical data?

Not always. Clinical data may be needed if the expansion results in increased risk or different benefit-risk profile compared to the predicate device. However, if the new indication represents similar or lower risk, non-clinical data may suffice.

Can manufacturers avoid clinical data requirements by choosing different predicate devices?

Yes, FDA recommends manufacturers avoid using predicate devices with newly identified or increased risks when alternative predicates exist. However, clinical data may still be required if no suitable alternative predicate is available without the identified risks.

What constitutes “different technological characteristics” that might require clinical data?

Significant changes in materials, device design, energy source, or other device features may require clinical data. Examples include switching from non-resorbable to resorbable materials in implants or changing critical reagents in diagnostic devices.

Are there device types that typically require clinical data due to limited non-clinical testing options?

Yes, devices for aesthetic purposes, schizophrenia treatment, hemostasis support, and blood screening often require clinical data because adequate non-clinical models are unavailable or cannot predict clinical outcomes reliably.

---

What You Need to Do 👇

Recommended Actions

1. Review the four scenarios requiring clinical data:
   • Differences in indications for use
   • Differences in technological characteristics
   • Inadequate non-clinical testing
   • Newly identified/increased risks for predicate device
2. Evaluate if your device falls under any scenario requiring clinical data
3. Consider submitting a pre-submission to discuss specific questions with FDA review division
4. Ensure clinical data constitutes valid scientific evidence per regulations
5. Provide adequate justification if using data from comparable devices
6. Review public FDA communications about new/increased risks for predicate devices
7. Consider alternative predicate devices without new/increased risks when possible
8. Consult device-specific guidance documents when available

Key Considerations

Clinical testing

• Clinical data may be needed to determine if new/modified indications fall within same intended use as predicate device
• Clinical data should constitute valid scientific evidence as defined in 21 CFR 860.7(c)(2)
• Data can include pre/post-market clinical investigations, literature, clinical experience reports, registries, adverse event databases, medical records
• Adequate justification needed when using data from comparable devices

Non-clinical testing

• When non-clinical testing is inadequate to establish SE due to:
  • No available model (analytical, bench, animal)
  • Model limitations preventing adequate assessment
  • Model not predictive of clinical outcomes
  • Anatomical/pathophysiological species-specific questions requiring clinical evidence

Labelling

• Proposed labeling used to determine device’s intended use
• Indications for use statement reviewed to determine if new device has same intended use as predicate

Safety

• New/increased risks identified for predicate device may require clinical data for new device
• Safety concerns identified through recalls and postmarket surveillance may trigger clinical data requirements

Other considerations

• Differences in technological characteristics requiring clinical data assessment:
  • Significant changes in materials
  • Significant changes in device design
  • Significant changes in energy source
  • Significant changes in other device features

---

Relevant Guidances 🔗

• Content and Decision-Making Process for 510k Submissions: Determining Substantial Equivalence
• General vs. Specific Indications for Use in Medical Device Substantial Equivalence Determinations
• Recommended Format and Content for Non-Clinical Bench Performance Testing in Medical Device Premarket Submissions
• Q-Submission Program: Strategic Framework for FDA Interactions in Medical Device Development
• Refuse to Accept Policy for 510k Submissions: Minimum Threshold Requirements for Administrative and Technical Review

Related references and norms 📂

• 21 CFR 860.7(c)(2): Definition of valid scientific evidence
• 21 CFR 807.92(a)(5): General/Specific Intended Use

---

Original guidance

• When Clinical Data is Needed to Support Substantial Equivalence for 510k Submissions
• HTML / PDF
• Issue date: 2023-09-07
• Last changed date: 2023-09-06
• Status: DRAFT
• Official FDA topics: Medical Devices, 510(k), Premarket, Biologics
• ReguVirta ID: f07ea4057fd20cf01bf0a7d4ae055d50