Skip to main content

What Clinical Data is Required

Clinical data in 510(k) submissions​

Clinical data is not always required for a 510(k). Whether clinical data is needed depends on:

  • The device's risk level and device type
  • Whether non-clinical data (bench testing, biocompatibility, software validation) is sufficient to demonstrate substantial equivalence
  • Whether the technological differences from the predicate raise clinical safety/effectiveness questions that only clinical data can address

When clinical data is typically required for 510(k):

  • Novel technological features that cannot be evaluated by bench testing alone
  • Devices where performance in the clinical setting cannot be inferred from lab data
  • Certain high-risk Class II devices (e.g., some IVDs, active implantables)

Clinical data in De Novo requests​

Clinical data requirements for De Novo vary by device type. Because De Novo devices are novel, FDA may require clinical data when:

  • Non-clinical data cannot adequately characterise the device's safety and effectiveness profile
  • The proposed special controls include clinical performance requirements

Clinical data in PMA applications​

PMA applications almost always require clinical data from well-controlled clinical investigations. The standard is valid scientific evidence — defined in 21 CFR § 860.7(c)(2) as evidence from well-controlled investigations, partially controlled investigations, studies and objective trials without matched controls, well-documented case histories conducted by qualified experts, and reports of significant human experience.

The pivotal study for a PMA must typically be conducted under an approved IDE.

What qualifies as valid scientific evidence​

Per FDA guidance, valid scientific evidence must:

  • Come from studies with appropriate controls
  • Use endpoints that are clinically meaningful
  • Be statistically adequate for the intended conclusions
  • Be collected under conditions approximating actual use

Official resources​

When non-clinical data alone may be sufficient: For devices with proven predicate equivalence, bench testing, biocompatibility testing, software validation, and sterilization validation may adequately support substantial equivalence without clinical data. This is most common for Class I devices and low-risk modifications to Class II devices.