The final guidance adopts the same definition of "accessory" set forth in the 2015 draft: accessories are products that meet the definition of a medical device under section 201(h) of the Federal Food, Drug, and Cosmetic Act (FDC Act) and are intended specifically to support, supplement, and/or augment the performance of one or more parent devices. FDA maintains that whether a product meets this definition will generally be determined by its labeling and promotional materials, rather than those of the parent device. The final guidance also maintains the scope of eligible accessories by explicitly carving out accessories of a type previously classified under the FDC Act, approved in a premarket approval order, or cleared via 510(k) premarket notification, with the latter more explicitly delineated throughout the guidance.

Consistent with the draft guidance, the final version notes that accessories may actually pose different risks than their parent device and, therefore, should be eligible for separate classification. Historically, FDA has treated accessories, unless already separately classified, as automatically falling into the same classification as the parent device. The draft, and now final, guidance documents shift this policy to allow manufacturers to more readily seek reclassification for accessories if the risk they pose is different from that of the parent device(s). As explained in the final guidance, this means that the risks of a parent device should not be automatically imputed to the accessory; rather, FDA should evaluate the risks imposed by the accessory's impact on the parent device when used as intended and any unique risk of the accessory as an independent device. This policy change, first announced in the draft guidance, and then endorsed and required by the recently enacted 21st Century Cures Act, provides more flexibility to accessory manufacturers. It is also consistent with FDA's overall risk-based regulatory framework for medical devices generally.

Of interest, the final guidance clarifies the scope of the policy delineated therein. Specifically, the guidance in several places emphasizes the distinction between medical devices that are specifically intended for use with another device to supplement, support, or augment its performance and are therefore subject to FDA regulation as accessories, and other products that may be used in conjunction with a medical device but are not designed for that purpose. Providing some additional clarity, FDA specifically notes that products in the latter category include off-the-shelf replacement parts such as batteries which would not themselves meet the definition of a medical device, much less an accessory. The guidance also notes that device components cannot be accessories, because an accessory is a finished device and a component, by definition, is not.

While FDA's policy is not restricted to any particular type(s) of medical devices/accessories, it focuses in part on mobile health and software as an area where what satisfies the definition of accessory may be more ambiguous. For instance, the final guidance addresses Software as a Medical Device (SaMD), namely "software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device." While SaMD that meets the definition of a medical device is subject to FDA regulation, not all such products would be considered accessories for purposes of this guidance. For example, stand-alone software programs that analyze radiological images or heart rate data may not support, supplement, and/or augment the performance of the device that generated the data, in which case they would not constitute accessories. They would likely be actively regulated by FDA, but not as accessories. However, where SaMD does meet the definition of an accessory, FDA indicates it will be subject to the same risk-based classification framework as applies for other accessories.

Finally, the guidance addresses how to obtain classification of a new accessory type via the de novo process. This follows the same overall approach as set forth in the draft, with an additional requirement to include an explanation for why the subject accessory does not fit within any identified classification for the parent device(s). If the de novo request is granted, the accessory (and its type) will be placed in Class I or Class II and FDA will publish the required Special Controls for the device type. The accessory may be marketed immediately and serve as a predicate device for future 510(k) premarket notifications. If the de novo request is declined, however, the accessory remains in Class III and may be lawfully marketed only pursuant to an approved PMA. The guidance appendix provides more detailed instructions on the information to be provided to FDA in this type of de novo request. It remains to be seen whether the de novo process will be an efficient regulatory tool for clearing novel accessories.

Ultimately, the final guidance does not make significant changes to the accessories classification policy proposed in the draft released nearly two years ago and now made legally binding per the recently-enacted 21st Century Cures Act. Instead, the guidance further elucidates how FDA intends to apply this regulatory framework in practice, including discussion of some particular device types that could be affected. In the face of continuing technological innovation, particularly in the digital health space, the finalized policy should facilitate the marketing of certain accessories that, while used with high-risk devices, may themselves present a lower risk profile.