Rule2026-05772
General and Plastic Surgery Devices; Reclassification of Optical Diagnostic Devices for Melanoma Detection and Electrical Impedance Spectrometers, To Be Renamed Software-Aided Adjunctive Diagnostic Devices for Use on Skin Lesions by Physicians Trained in the Diagnosis and Management of Skin Cancer
Primary source
Metadata and text below are from the Federal Register, a public-domain U.S. government work. Always verify the official published version before relying on it for any legal matter.
Published
March 25, 2026
Effective
April 24, 2026
Issuing agencies
Health and Human Services DepartmentFood and Drug Administration
Abstract
The Food and Drug Administration (FDA) is issuing a final order reclassifying optical diagnostic devices for melanoma detection (product code OYD) and electrical impedance spectrometers (product code ONV), both postamendments class III device types, into class II (special controls), subject to premarket notification. FDA is also renaming and codifying these devices under the new classification regulation named "software-aided adjunctive diagnostic devices for use on skin lesions by physicians trained in the diagnosis and management of skin cancer." FDA is also establishing the special controls necessary to provide a reasonable assurance of safety and effectiveness of these devices.
Full Text
<html>
<head>
<title>Federal Register, Volume 91 Issue 57 (Wednesday, March 25, 2026)</title>
</head>
<body><pre>
[Federal Register Volume 91, Number 57 (Wednesday, March 25, 2026)]
[Rules and Regulations]
[Pages 14445-14458]
From the Federal Register Online via the Government Publishing Office [<a href="http://www.gpo.gov">www.gpo.gov</a>]
[FR Doc No: 2026-05772]
========================================================================
Rules and Regulations
Federal Register
________________________________________________________________________
This section of the FEDERAL REGISTER contains regulatory documents
having general applicability and legal effect, most of which are keyed
to and codified in the Code of Federal Regulations, which is published
under 50 titles pursuant to 44 U.S.C. 1510.
The Code of Federal Regulations is sold by the Superintendent of Documents.
========================================================================
Federal Register / Vol. 91, No. 57 / Wednesday, March 25, 2026 /
Rules and Regulations
[[Page 14445]]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration
21 CFR Part 878
[Docket No. FDA-2022-N-0794]
General and Plastic Surgery Devices; Reclassification of Optical
Diagnostic Devices for Melanoma Detection and Electrical Impedance
Spectrometers, To Be Renamed Software-Aided Adjunctive Diagnostic
Devices for Use on Skin Lesions by Physicians Trained in the Diagnosis
and Management of Skin Cancer
AGENCY: Food and Drug Administration, HHS.
ACTION: Final amendment; final order.
-----------------------------------------------------------------------
SUMMARY: The Food and Drug Administration (FDA) is issuing a final
order reclassifying optical diagnostic devices for melanoma detection
(product code OYD) and electrical impedance spectrometers (product code
ONV), both postamendments class III device types, into class II
(special controls), subject to premarket notification. FDA is also
renaming and codifying these devices under the new classification
regulation named ``software-aided adjunctive diagnostic devices for use
on skin lesions by physicians trained in the diagnosis and management
of skin cancer.'' FDA is also establishing the special controls
necessary to provide a reasonable assurance of safety and effectiveness
of these devices.
DATES: This order is effective April 24, 2026.
FOR FURTHER INFORMATION CONTACT: Jessica Carr, Center for Devices and
Radiological Health, Food and Drug Administration, 10903 New Hampshire
Ave., Bldg. 66, Rm. 4522, Silver Spring, MD 20993, 301-796-5997,
<a href="/cdn-cgi/l/email-protection#f5bf9086869c9694dbb6948787b5939194db9d9d86db929a83"><span class="__cf_email__" data-cfemail="f9b39c8a8a909a98d7ba988b8bb99f9d98d791918ad79e968f">[email protected]</span></a>.
SUPPLEMENTARY INFORMATION:
I. Background--Regulatory Authorities
The Federal Food, Drug, and Cosmetic Act (FD&C Act), as amended,
establishes a comprehensive system for the regulation of medical
devices intended for human use. Section 513 of the FD&C Act (21 U.S.C.
360c) established three classes of devices, reflecting the regulatory
controls needed to provide reasonable assurance of their safety and
effectiveness. The three classes of devices are class I (general
controls), class II (special controls), and class III (premarket
approval).
Devices that were not introduced or delivered for introduction into
interstate commerce for commercial distribution prior to May 28, 1976
(generally referred to as postamendments devices) are automatically
classified by section 513(f)(1) of the FD&C Act into class III without
any FDA rulemaking process. Those devices remain in class III and
require premarket approval, unless and until (1) the Food and Drug
Administration (FDA, the Agency, or we) reclassifies the device into
class I or class II; or (2) FDA issues an order finding the device to
be substantially equivalent, in accordance with section 513(i) of the
FD&C Act, to a predicate device that does not require premarket
approval. FDA determines whether new devices are substantially
equivalent to predicate devices by means of the procedures in section
510(k) of the FD&C Act (21 U.S.C. 360(k)) and our implementing
regulations (part 807, subpart E (21 CFR part 807, subpart E)).
A postamendments device that has been initially classified into
class III under section 513(f)(1) of the FD&C Act may be reclassified
into class I or class II under section 513(f)(3) of the FD&C Act.
Section 513(f)(3) of the FD&C Act provides that FDA, acting by
administrative order, can reclassify the device into class I or class
II on its own initiative, or in response to a petition from the
manufacturer or importer of the device. To change the classification of
the device, the new class must have sufficient regulatory controls to
provide reasonable assurance of the safety and effectiveness of the
device for its intended use.
FDA relies upon ``valid scientific evidence,'' as defined in
section 513(a)(3) of the FD&C Act and 21 CFR 860.7(c)(2), in the
classification process to determine the level of regulation for
devices. To be considered in the reclassification process, the ``valid
scientific evidence'' upon which the Agency relies generally must be
publicly available. Publicly available information excludes trade
secret and/or confidential commercial information, e.g., the contents
of a pending premarket approval application (PMA) (see section 520(c)
of the FD&C Act (21 U.S.C. 360j(c))). Section 520(h)(4) of the FD&C Act
provides that FDA may use, for reclassification of a device, certain
information in a PMA 6 years after the application has been approved.
This includes information from clinical and preclinical tests or
studies that demonstrate the safety and effectiveness of the device,
but it does not include the descriptions of methods of manufacture and
product composition and other trade secrets.
Section 510(m) of the FD&C Act provides that FDA may exempt a class
II device from the requirements under section 510(k) of the FD&C Act if
FDA determines that a premarket notification (510(k)) is not necessary
to provide reasonable assurance of the safety and effectiveness of the
device type.
On June 30, 2022, FDA published a proposed order \1\ in the Federal
Register to reclassify optical diagnostic devices for melanoma
detection and electrical impedance spectrometers (product codes OYD and
ONV, respectively) from class III to class II subject to premarket
notification (87 FR 39025, the ``proposed order'').\2\ FDA has
considered the information available to the Agency, including the
deliberations of the General and Plastic Surgery Devices Advisory Panel
convened on
[[Page 14446]]
July 28-29, 2022 (the ``Panel'') to discuss software-aided \3\
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer and the proposed
reclassification (as discussed in section II of this document), as well
as comments from the public docket on the proposed order (as discussed
in section III of this document), to determine that there is sufficient
information to establish special controls to effectively mitigate the
risks to health (updated as discussed in section IV of this document).
FDA has also determined based on this information that the special
controls, together with general controls, provide a reasonable
assurance of safety and effectiveness when applied to these devices.
---------------------------------------------------------------------------
\1\ The ``ACTION'' caption for this proposed order was styled as
``Proposed amendment; proposed order; request for comments'' rather
than ``Proposed order.'' Beginning in December 2019, this editorial
change was made to indicate that the document ``amends'' the Code of
Federal Regulations. The change was made in accordance with the
Office of the Federal Register's (OFR) interpretations of the
Federal Register Act (44 U.S.C. chapter 15), its implementing
regulations (1 CFR 5.9 and parts 21 and 22), and the Document
Drafting Handbook.
\2\ In the proposed order, FDA proposed to reclassify optical
diagnostic devices for melanoma detection and electrical impedance
spectrometers under a new device classification regulation with the
name ``computer-aided devices which provide adjunctive diagnostic
information about lesions suspicious for melanoma.'' In this final
order, FDA is renaming these devices with the name ``software-aided
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer,'' to better
describe the devices that fit within this generic device type and
are subject to this reclassification order.
\3\ FDA regulates software that meets the definition of a
device, which is defined in section 201(h)(1) of the FD&C Act as
``an instrument, apparatus, implement, machine, contrivance,
implant, in vitro reagent, or other similar or related article,
including any component, part, or accessory, which is--recognized in
the official National Formulary, or the United States Pharmacopeia,
or any supplement to them, intended for use in the diagnosis of
disease or other conditions, or in the cure, mitigation, treatment,
or prevention of disease, in man or other animals, or intended to
affect the structure or any function of the body of man or other
animals, and which does not achieve its primary intended purposes
through chemical action within or on the body of man or other
animals and which is not dependent upon being metabolized for the
achievement of its primary intended purposes. The term `device' does
not include software functions excluded pursuant to section 520(o)''
of the FD&C Act.
---------------------------------------------------------------------------
Therefore, in accordance with section 513(f)(3) of the FD&C Act,
FDA, on its own initiative, is issuing this final order to reclassify
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer from
class III to class II (special controls).\4\ Absent the special
controls identified in this final order, general controls applicable to
the device type are insufficient to effectively mitigate the risks
identified for this device type, such as the risk of incorrect or
delayed diagnosis of skin cancer from false negative results, and
therefore insufficient to provide reasonable assurance of the safety
and effectiveness of these devices. FDA expects that the
reclassification of these devices will enable more manufacturers to
develop these types of devices such that patients will benefit from
increased access to adjunctive diagnostics for which there is a
reasonable assurance of safety and effectiveness.
---------------------------------------------------------------------------
\4\ FDA notes that the ``ACTION'' caption for this final order
is styled as ``Final amendment; final order,'' rather than ``Final
order.'' Beginning in December 2019, this editorial change was made
to indicate that the document ``amends'' the Code of Federal
Regulations. The change was made in accordance with the Office of
Federal Register's (OFR) interpretations of the Federal Register Act
(44 U.S.C. chapter 15), its implementing regulations (1 CFR 5.9 and
parts 21 and 22), and the Document Drafting Handbook.
---------------------------------------------------------------------------
For these class II devices, instead of a PMA, manufacturers may
submit a premarket notification and obtain FDA clearance of the devices
before marketing them. This action will decrease regulatory burden on
industry, as manufacturers will no longer have to submit a PMA for
these types of devices but can instead submit a 510(k) to the Agency
for review prior to marketing their device. A 510(k) typically results
in a shorter premarket review timeline compared to a PMA, which
ultimately provides patients with more timely access to these types of
devices.
II. Deliberations of the Panel Meeting
A. Panel Discussion
On July 28, 2022, the Panel met to discuss the general topic of
skin lesion analyzer technology and its application to detecting skin
cancers in various patient care settings (the proposed reclassification
of software-aided adjunctive diagnostic devices for use on skin lesions
by physicians trained in the diagnosis and management of skin cancer
was discussed on the second day of the Panel meeting, as described
later in this section). The Panel members were asked questions
regarding the diagnosing standard, or ground truth, used to confirm
lesion diagnosis in clinical testing of device accuracy; the acceptable
sensitivity and specificity thresholds for different diagnoses and
users; and patient characteristic considerations based on variable
incidence of skin lesions in the U.S. population (Ref. 1).
The Panel generally advised that the diagnostic standard, or ground
truth, for algorithm validation studies should be histological
diagnosis, while some Panel members believed alternative approaches
could be valuable depending on lesion type (such as consensus of a
group of experts in certain benign lesion cases).
The Panel advised that the applicable thresholds for performance
criteria, such as sensitivity and specificity, should be evaluated for
specific skin lesion types (e.g., melanoma, basal cell carcinoma, and
squamous cell carcinoma), intended users (e.g., dermatologist, primary
care physician, lay user), and device type, and should show that the
device improves the performance of the clinical user or improves
patient outcomes (i.e., earlier diagnosis). The Panel also discussed
that the sensitivity and specificity threshold should be higher for
standalone devices compared to devices intended for adjunctive use.
Some Panel members commented on the importance of prospective data or
post-market surveillance of real-world use to confirm device benefit.
The Panel agreed that the selection of performance thresholds should
take into account not only the risks of false negatives but also the
risks of false positives, including psychological impacts and other
effects.
Regarding patient characteristics that should be evaluated in
clinical testing, the Panel advised that all patient skin phototypes
should be studied with some flexibility in the data collection for low-
incidence populations, such as by balancing premarket data with post-
market study requirements and requiring transparency in the
demographics and prevalence data in the studied populations.
On July 29, 2022, the Panel met to discuss and make recommendations
regarding the proposed reclassification of software-aided adjunctive
diagnostic devices for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer from class III to class II
(Ref. 1). In particular, the Panel discussed the proposed
reclassification of MelaFind, a device that uses multispectral imaging
and that was approved by FDA in 2011 (PMA P090012) (Ref. 2). The Panel
also discussed the proposed reclassification of Nevisense, a device
that measures electrical impedance and that was approved by FDA in 2017
(PMA P150046) (Ref. 3). Both MelaFind and Nevisense are intended for
use on cutaneous lesions suspicious for skin cancer when a
dermatologist chooses to obtain additional information when considering
biopsy. At the Panel meeting, FDA presented the risks, mitigations, and
special controls identified in the proposed reclassification order for
a software-aided adjunctive diagnostic device for use on skin lesions
by physicians trained in the diagnosis and management of skin cancer.
The Panel agreed with inclusion of the risks identified by FDA, but
suggested additional risks be included, as detailed in section II.B.1.
The Panel agreed that general controls alone are not sufficient to
provide a reasonable assurance of safety and effectiveness for this
device type. The Panel had varying opinions on whether the devices are
life-supporting or life-sustaining, or for a use which is of
substantial importance in preventing impairment of human health, and
whether or not these devices present a ``potential unreasonable risk of
illness
[[Page 14447]]
or injury'' considering, among other factors, that the devices are
intended to be used adjunctively to standard of care but, in the
opinion of some Panel members, have a potentially high risk associated
with false negative output.
Some Panel members agreed that sufficient information exists to
establish special controls for this device type, though most Panel
members disagreed, suggesting more information is needed to understand
device performance and device benefits and risks in real-world use.
Several Panel members recommended additional clarity or detail in the
proposed special controls. Some believed that FDA's proposed special
controls were appropriate but not sufficient, recommending additional
controls such as requirements for post-market surveillance, metrics to
evaluate patient benefits and risks in addition to sensitivity and
specificity, evaluation of specific patient populations and subtypes,
prospective real-world use studies, and transparency regarding
algorithm development. Various Panel members believed that this device
type should not be reclassified from class III to class II.
B. FDA Responses to Panel Deliberations and Changes in the Final Order
FDA's responses to the recommendations of the Panel deliberations
are detailed in this section. As discussed in section III, FDA also
considered comments that were received from industry members,
professional societies, and other interested parties in developing this
final order. However, here in section II, we specifically address the
Panel recommendations and FDA's responses.
1. Risks to Health
The Panel suggested that additional risks to health presented by
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer be
included in FDA's overall assessment of the risks to health, such as
device bias in different populations, psychological impact of false
positives and false negatives, risks associated with specific locations
of device use, risks associated with inadequate user training, and
risks associated with algorithm performance drift in real-world use.
FDA recognizes the potential psychological impact of inaccurate
device output and considers this to be encompassed within the risks of
false negative results, false positive results, use error or improper
device use, and device failure or malfunction listed in table 1. FDA
also agrees that there are risks associated with inadequate user
training and considers these to be included in the risks of false
negative results, false positive results, and use error or improper
device use listed in table 1.
FDA agrees that devices may perform differently in different
patient sub-populations due to several factors, including bias in
training, tuning, and/or validation datasets. Likewise, FDA agrees that
device use on different anatomical locations may result in different
device performance. FDA considers these sources of performance
variability to be encompassed within the risks of false negative
results and false positive results listed in table 1. In response to
the Panel discussion, FDA has made several additions to the special
controls regarding clinical performance testing and labeling to
mitigate these specific sources of performance variability and their
contribution to device risk. These additions include requiring that
clinical testing evaluate patients across risk factors that represent
the intended patient population, including age, body site, skin
phototype, and other clinical factors as applicable; requiring that
labeling include a description of the patient population that was used
in development or training of the device algorithm; and requiring that
labeling include information related to the limitations of device
performance or subpopulations for which the device may not perform as
expected or for whom the device has not been validated.
Some Panel members also emphasized that interference with implants
could pose risks to patients, including device interference that
affects the implant performance, or interference of the implant that
affects the device performance. FDA agrees, and notes that the risk of
interference with other devices, including implants, was included among
the risks to health identified in the proposed order (87 FR 39025). FDA
also considers these risks to be encompassed within the risks of false
negative results, false positive results, and device failure or
malfunction listed in table 1.
Regarding the Panel's comments on algorithm performance drift in
real-world use, FDA notes that MelaFind and Nevisense have fixed
algorithms which do not change or adapt over time or with exposure to
lesions during clinical use. Devices with algorithms other than fixed
algorithms would represent a change in technology from the existing
devices, and would very likely raise different questions of safety and
effectiveness than the predicate device, which would preclude a finding
of substantial equivalence under 21 CFR 807.100(b). Such devices would
need to be evaluated according to 21 CFR 807.100(b)(2) to determine
whether such a device could be found substantially equivalent to a
predicate device within this classification regulation.
In addition, when a manufacturer of a software-aided adjunctive
diagnostic device for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer that has been cleared by FDA
intentionally changes the algorithm of its device, FDA would review the
change in a premarket submission if the change exceeds the regulatory
threshold of 21 CFR 807.81(a)(3) for submission and clearance of a new
510(k). FDA acknowledges that the device-aided user's accuracy may
change over time or may differ from that demonstrated in premarket
clinical studies with a retrospective design. FDA considers risks to
health related to the accuracy of the device-aided user to be
encompassed within the risk to health of false results due to use error
or improper device use listed in table 1. In response to the Panel
discussion, FDA has made several additions to the special controls
regarding clinical performance testing to mitigate these risks to
health. These additional special controls include requiring data
obtained from both premarket clinical performance validation testing
and post-market surveillance acquired under anticipated conditions of
use, unless FDA determines based on the totality of the premarket data
that data from post-market surveillance is not required.
2. Special Controls
FDA appreciates the perspective of various Panel members that more
information is necessary to understand the device performance and its
benefits and risks in real-world use. However, FDA believes that
sufficient information exists and is available to FDA through the
MelaFind PMA and associated panel considerations of that PMA,\5\
published peer-reviewed literature (as discussed in the proposed order
(87 FR 39025)),\6\ and information collected from FDA's publicly
available Medical Device
[[Page 14448]]
Report (MDR) database, Manufacturer and User Facility Device Experience
(MAUDE) database, and Medical Device Recall database since this device
type was first introduced to the market, to understand the benefits and
risks of this device type and establish special controls that
effectively mitigate those risks. FDA reviewed certain real-world data,
as discussed above, and does not agree that additional data from real-
world use is necessary prior to establishing special controls (we note
that data from real-world use is not a prerequisite to establishing
special controls), or that special controls cannot be established based
on current information, particularly as the intended use of the devices
is limited to adjunctive use (rather than standalone use, which would
present greater risks). For example, in response to feedback from the
Panel, the special controls have been revised to require a high
sensitivity (>90 percent for lesions with high metastatic potential, or
a clinically justified alternative), which can effectively mitigate the
risk of false negatives. See section IV for additional discussion of
the mitigation measures for the risks to health identified for this
device type. Some of the panel's concerns regarding real-world use
appeared to relate to uncertainty about adaptive algorithms, but as
noted above, devices with algorithms other than fixed algorithms are
very likely to raise different questions of safety and effectiveness
from the devices being reclassified here. In addition, to the extent
there is concern about FDA relying on data from only two authorized
devices, FDA notes that the statute contemplates that data from just
one device could provide sufficient information to establish special
controls for a device type under the de novo provisions (see 21 U.S.C.
360c(f)(2)).
---------------------------------------------------------------------------
\5\ The FDA General and Plastic Surgery Devices Panel reviewed
the MelaFind PMA at a meeting on November 18, 2010. Meeting
materials are available at available at <a href="https://wayback.archive-it.org/7993/20170403223449/https:/www.fda.gov/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/GeneralandPlasticSurgeryDevicesPanel/ucm205684.htm">https://wayback.archive-it.org/7993/20170403223449/https:/www.fda.gov/AdvisoryCommittees/CommitteesMeetingMaterials/MedicalDevices/MedicalDevicesAdvisoryCommittee/GeneralandPlasticSurgeryDevicesPanel/ucm205684.htm</a>.
\6\ Such literature included the pivotal study for Nevisense,
see Ref. 9 in the proposed order.
---------------------------------------------------------------------------
In addition, although FDA appreciates that this technology may
continue to evolve, the potential for such evolution does not suggest
that there is insufficient information to establish special controls
for the current device type. Future devices would only fall within this
device type, and thus be classified into class II subject to the
special controls established in this final order, if the device is
found to be ``substantially equivalent'' to a predicate software-aided
adjunctive diagnostic device for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer, which requires
that the device have the same intended use and the same technological
characteristics as the predicate device, or, if the device has
different technological characteristics, the device must be as safe and
effective as a legally marketed device and not raise different
questions of safety and effectiveness. A future device that raises
different questions of safety or effectiveness would not fall within
the device type being reclassified into class II in this final order.
As noted in section II.B.1, FDA has made several additions to the
clinical performance testing and labeling special controls in response
to the deliberations of the Panel. FDA has also revised and added other
special controls in response to recommendations from the Panel.
The Panel stated concerns regarding potential differences in device
performance as observed in retrospectively designed premarket studies
versus real-world device use. To partially address this concern, the
revised special controls require data obtained from both premarket
clinical performance validation testing and post-market surveillance
acquired under anticipated conditions of use, unless FDA determines
based on the totality of the premarket data that data from post-market
surveillance is not required. The post-market surveillance data may
provide additional information regarding device performance in the
overall patient population, as well as in low-incidence patient sub-
populations for whom relatively limited data was available at the time
of premarket review.
As identified in the proposed order, and consistent with the
recommendations of the Panel, the special controls allow for
flexibility in the endpoints used to demonstrate patient benefits and
risks in addition to sensitivity and specificity. The revised special
controls require demonstration of ``superior accuracy'' of device-aided
users' diagnostic characterization of the indicated lesions compared to
the accuracy of unaided users, but do not specify a particular endpoint
that must be used to evaluate accuracy. Additionally, to mitigate the
high risk of false negatives identified by the Panel, the revised
special controls require standalone device performance testing to
demonstrate at least 90 percent sensitivity of the device output for
lesions with high metastatic potential, or an alternative clinical
consideration must be provided to justify lower sensitivity. The
special controls also require that clinical justification be provided
for the reported specificity. In addition, consistent with Panel
feedback that applicable thresholds for performance criteria are
dependent, in part, on the intended user, the labeling special controls
require statements that the device is intended to be used by a
physician trained in the clinical diagnosis and management of skin
cancer (e.g., a dermatologist) and that the device is not intended for
use as a standalone diagnostic.
FDA agrees with the Panel recommendation that device evaluation
should consider specific patient populations and lesion subtypes. The
revised special controls require clinical testing and standalone
testing to evaluate patients across risk factors (including age, body
site, skin phototype, and other clinical factors as applicable) that
represent the intended patient population. Analysis of standalone
performance must include subgroup analysis by relevant risk factors.
Moreover, and in response to the Panel's feedback regarding
transparency, the revised special controls require device labeling to
include a description of the patient population that was used in
development or training of the device algorithm. The labeling must also
include a summary of the standalone and clinical performance testing
conducted with the device, including performance of the device for all
clinically relevant subgroups within the intended patient population,
and information related to the limitations of device performance or
subpopulations for which the device may not perform as expected or for
whom the device has not been validated.
FDA believes that these labeling special controls, along with other
special controls such as those requiring human factors testing, device
precision testing, the inclusion of information about device outputs in
device labeling, and the inclusion of information about user
qualifications in device labeling, will help to mitigate the risk of
false results due to use error and changes in performance of the device
in real-world use.
3. Reclassification From Class III to Class II
Under section 513(a)(1)(C) of the FD&C Act (21 U.S.C.
360c(a)(1)(C)), class III devices are those devices for which, among
other things, insufficient information exists to determine that general
controls and special controls would provide a reasonable assurance of
safety and effectiveness. In contrast, under section 513(a)(1)(B) of
the FD&C Act, class II devices are those which cannot be classified
into class I because general controls by themselves are insufficient to
provide reasonable assurance of the safety and effectiveness of the
device, but for which there is sufficient information to establish
[[Page 14449]]
special controls to provide such assurance. FDA agrees with the Panel
members that general controls are not sufficient to provide reasonable
assurance of the safety and effectiveness of software-aided adjunctive
diagnostic devices for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer. However, as noted above, FDA
believes that sufficient information exists and is available to FDA
through the MelaFind PMA and associated panel considerations of that
PMA, published peer-reviewed literature, and FDA's publicly available
MDR database, MAUDE database, and Medical Device Recall database to
establish special controls that effectively mitigate the risks to
health identified for this device type. Accordingly, FDA has determined
that software-aided adjunctive diagnostic devices for use on skin
lesions by physicians trained in the diagnosis and management of skin
cancer should be reclassified from class III to class II.
III. Comments on the Proposed Order and Panel
A. Introduction
FDA received comments from more than 50 commenters on the proposed
order (87 FR 39025) published in the Federal Register on June 30, 2022,
and for the subsequent Panel meeting held on July 28-29, 2022. The
comment period on the proposed order closed on August 29, 2022, and the
docket for the Panel meeting also closed on August 29, 2022. The
majority of the comments received by the close of the comment periods
were from individual medical professionals. Professional societies and
members of the medical device industry also provided comments. Some of
the comments contained one or more comments on one or more issues. We
received comments providing support for the proposed reclassification
as well as comments recommending against the proposed reclassification.
Many comments also included technical considerations for assessing the
safety and effectiveness of the devices subject to the proposed
reclassification.
We describe and respond to the comments in section III.B of this
document. The order of the comments and our response to them is purely
for organizational purposes and does not signify the comment's value or
importance nor the order in which comments were received. Certain
comments are grouped together under a single number because the subject
matter is similar. Please note that in some cases we separated
different issues discussed by the same commenter and designate them as
distinct comments for purposes of our responses.
B. Description of Comments and FDA Response
(Comment 1) Multiple comments supported the proposed
reclassification of software-aided adjunctive diagnostic devices for
use on skin lesions by physicians trained in the diagnosis and
management of skin cancer. Commenters stated that reclassification into
class II would be appropriate based on the device risks, that
reclassification would support patient access to dermatological care,
and that special controls could be established to provide a reasonable
assurance of safety and effectiveness, with some commenters emphasizing
the need to communicate risks due to false positives or false
negatives. One commenter stated that there has been widespread use of
physician adjunctive-use skin lesion analyzers in Europe for several
years and the commenter was unaware of any meaningful negative impacts
associated with such use. Many commenters also specified that
reclassification of these devices into class II would be appropriate as
these devices are intended to provide adjunctive information, not
intended for use as standalone diagnostic devices. Some of these
commenters also expressed that it would be appropriate to use these
devices for general screening or triage.
(Response 1) FDA agrees with the comments supporting
reclassification. Based on the available information, as discussed in
the proposed order and in section II of this document, and in
consideration of the comments received on the proposed order and the
Panel meeting, FDA has determined that reclassification of software-
aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer into
class II is appropriate. FDA made this determination because there is
sufficient information to establish special controls that together with
general controls will provide a reasonable assurance of safety and
effectiveness for these devices when used adjunctively by physicians
trained in the diagnosis and management of skin cancer (including
requirements that the device labeling provide information regarding
performance measures, including sensitivity, specificity and
statistical confidence intervals, and an identification of risks
associated with misinterpretation of the device outputs).
FDA also agrees with the comments that state that reclassification
is appropriate based on the intended use of these devices to provide
adjunctive information to aid in the evaluation of lesions suspicious
for skin cancer following identification of a suspicious skin lesion,
and not as a standalone diagnostic or for use to confirm a clinical
diagnosis. Based on the totality of information available, FDA believes
that general controls and special controls can provide a reasonable
assurance of safety and effectiveness for these devices when they are
intended to provide adjunctive information, not a diagnosis, and the
intended users are physicians trained in the diagnosis and management
of skin cancer. However, use of these devices as general screening
tools or for triage falls outside the approved intended uses for the
devices being reclassified, and hence is outside the scope of this
final order. FDA would need additional data and information to address
those uses.
The Agency believes that reclassification of software-aided
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer under this final
order will increase access to devices for which there is a reasonable
assurance of safety and effectiveness by reducing the regulatory burden
on manufacturers, while still providing reasonable assurance of safety
and effectiveness. Specifically, reclassifying this type of device from
class III into class II will reduce regulatory burdens on industry
because instead of submitting a PMA, manufacturers may submit a less
burdensome 510(k) to obtain FDA clearance of the device before
marketing it. The Agency also agrees with commenters that this
reclassification may have a positive impact on dermatological care.
(Comment 2) Multiple comments disagreed with the proposed
reclassification of software-aided adjunctive diagnostic devices for
use on skin lesions by physicians trained in the diagnosis and
management of skin cancer. Some commenters stated that it would be
inappropriate, difficult, or impossible to develop a uniform non-
clinical and clinical testing paradigm that can be applied to every
device of this type. Some commenters expressed concern that
reclassification would reduce FDA's ability to provide input on non-
clinical or clinical study design. Many commenters also indicated that
each specific device should be assessed to ensure appropriate device
performance and clinical validity catered to the device's specific
technological characteristics, intended users, and lesion types, and
that these devices should be maintained in class
[[Page 14450]]
III to maintain the strictest level of FDA premarket review of device
safety, performance, and labeling. Some commenters stated that there is
not sufficient information or experience with this device type to
support reclassification and the establishment of special controls to
provide a reasonable assurance of safety and effectiveness, or that the
consequences of an inaccurate diagnosis could result in unnecessary
procedures (such as biopsies and surgeries) or are too severe to
support reclassification.
(Response 2) FDA disagrees that reclassification would be
inappropriate based on the need for device-specific non-clinical and
clinical assessments or premarket review for the specific device. FDA
agrees that a uniform approach to non-clinical and clinical performance
testing is not appropriate for software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer, and there is no single clinical study
design that would be appropriate for all devices of this type. The
special controls for this device type outlined in this final order
include controls for non-clinical and clinical testing that could
support unique study designs that are appropriate for the specific
indications for use and technological characteristics for each device
that would be reviewed in a premarket notification. While every
clinical study developed to support a premarket submission for a
software-aided adjunctive diagnostic device for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer
should be designed to demonstrate that the device performs as intended
when used by the intended user in the intended patient population, the
final special controls for these devices allow for customized study
designs tailored specifically for each device considering the device
technology and indications for use. Reclassification of these devices
from class III to class II does not exclude device-specific non-
clinical and clinical testing. Rather, the special controls establish
requirements for non-clinical and clinical testing that is necessary to
support reasonable assurance of safety and effectiveness for the
device. The substantial equivalence framework in the 510(k) review
paradigm permits potential clearance of devices of the same device type
even when supported by data from different types of studies, consistent
with FDA's least burdensome provisions.\7\
---------------------------------------------------------------------------
\7\ See sections 513 and 515 of the FD&C Act. See also FDA's
guidance, ``The Least Burdensome Provisions: Concepts and
Principles; Guidance for Industry and Food and Drug Administration
Staff,'' Feb. 5, 2019. Available at <a href="https://www.fda.gov/media/73188/download">https://www.fda.gov/media/73188/download</a>.
---------------------------------------------------------------------------
FDA also agrees that these devices require appropriate regulatory
oversight through premarket review, but disagrees that classification
in class III and PMA approval is necessary to provide such oversight
for software-aided adjunctive diagnostic devices for use on skin
lesions by physicians trained in the diagnosis and management of skin
cancer. For example, PMA approval requires, among other things, the
submission of a complete description of the methods used in, and the
facilities and controls used for, the manufacture, processing, packing,
storage, and, where appropriate, installation of the device (21 CFR
814.20(b)(4)(v)), and FDA often conducts inspections prior to PMA
approval. PMA requirements also include annual reporting requirements,
and requirements regarding the submission and approval of PMA
supplements. As discussed in more detail in response to Comment 8, FDA
does not believe that compliance with such requirements is necessary to
provide a reasonable assurance of safety and effectiveness for
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer. This
final reclassification order establishes that instead of submission and
approval of a PMA, submission and clearance of a 510(k) will be
required prior to legally marketing these devices. FDA reviews the non-
clinical and clinical data and related valid scientific evidence
included in a 510(k) to assess substantial equivalence to a legally
marketed predicate device, including, as appropriate, conformance to
special controls. FDA will also evaluate device labeling to ensure that
labeling is consistent with the labeling requirements established in
the special controls. Clearance of a 510(k) reflects FDA's
determination that the device has the same intended use and
technological characteristics as the predicate, or, if the device has
different technological characteristics, that the device is as safe and
as effective as a legally marketed device and does not raise different
questions of safety and effectiveness than the predicate device (see
section 513(i) of the FD&C Act).
Additionally, some commenters expressed concern that
reclassification would reduce FDA's ability to provide input on non-
clinical or clinical study design. However, regardless of whether a
device is classified in class III or class II, FDA may provide input on
non-clinical or clinical study design. A manufacturer may seek FDA
input on non-clinical or clinical study design by utilizing our Q-
Submission program, through which FDA may provide input on device-
specific requirements and recommendations for non-clinical and clinical
studies intended to support device-specific indications for use.
Additional information regarding the Q-Submission program can be found
in FDA's final guidance document entitled ``Requests for Feedback and
Meetings for Medical Device Submissions: The Q-Submission Program''
(Ref. 4). In addition, reclassification does not change the regulatory
requirements related to clinical study oversight and investigational
device exemptions (IDEs). FDA may provide specific feedback and study
design considerations for clinical trials as a part of IDE review for
significant risk studies. The special controls also require that
specific study design elements consider the indicated lesions, intended
patient population, and intended users to provide a reasonable
assurance of safety and effectiveness.
Ultimately, as discussed in section II.B.2 of this document, FDA
believes there is sufficient information and experience available to
support the reclassification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer into class II, including sufficient
information to establish special controls for these devices such that
general controls and special controls together provide a reasonable
assurance of safety and effectiveness. Considering the intended use of
these devices as an adjunctive source of information for clinical
decision making (rather than to provide a diagnosis), FDA believes
these controls effectively mitigate the potential negative consequences
of an incorrect diagnosis based on the adjunctive information from the
device.
(Comment 3) Multiple comments disagreed with the proposed
reclassification based on the concern that software-aided adjunctive
diagnostic devices for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer may be used by the general
public (as an over-the-counter device) or by other unqualified users to
assess skin lesions (their own or others'), or based on the concern
that these devices may be used as standalone diagnostic devices in the
absence of a clinical assessment by a physician trained in the clinical
diagnosis and management of skin cancer (e.g., a dermatologist). Some
of these commenters indicated that use
[[Page 14451]]
by untrained users, including non-dermatologist medical providers and/
or lay users, would likely increase the risk of device misuse and
result in unacceptable rates of false positive and false negative
results, leading to patient harm (such as delay in diagnosis of skin
cancer and/or unnecessary procedures and patient anxiety) or increased
demand for dermatological services. Commenters stated that diagnosis by
a medical professional trained in the diagnosis and management of skin
cancer is the gold standard for skin cancer diagnosis and that a
software-aided adjunctive diagnostic device for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer is
not a substitute for a clinician's diagnosis and histopathology.
(Response 3) FDA agrees that software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer are not a substitute for diagnosis of
skin lesions by a trained clinician and would present different risks
if intended for use by untrained users. The devices being reclassified
in this final order are identified as being intended for prescription
use only and for use by a physician trained in the clinical diagnosis
and management of skin cancer (e.g., a dermatologist) as an adjunctive
device following identification of a suspicious skin lesion. These
devices are not intended to be used by the general public or by medical
personnel untrained in the clinical diagnosis and management of skin
cancer, and are not intended for use as a standalone diagnostic or to
confirm a clinical diagnosis.
FDA agrees that the expertise of a trained clinician is needed to
appropriately select lesions for assessment consistent with the
intended use and instructions for use for these devices, and consistent
with the clinical studies that validate device performance. As such,
the intended users of these devices are physicians trained in the
clinical diagnosis and management of skin cancer. The clinical
performance validation testing special controls require that lesions
used in clinical testing must be selected by representative users
(e.g., dermatologists) and a justification must be provided for the
quantity and range of mimic lesions per diagnosis, which is intended to
ensure that the device works safely and effectively on the lesions that
a user (e.g., a dermatologist) would choose to use the device on.
Additionally, the labeling special controls require that the labeling
include a statement that the device is not intended for use as a
standalone diagnostic and include the user qualifications needed for
safe use of the device, including a description of required user
training and a statement that the device is intended to be used by a
physician trained in the clinical diagnosis and management of skin
cancer.
(Comment 4) Some commenters stated that reclassified software-aided
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer should include
devices intended for use by non-specialists and by general practitioner
care providers.
(Response 4) FDA disagrees with these comments and believes that
physicians trained in the clinical diagnosis and management of skin
cancer are the appropriate intended users of these devices based on the
intended use, performance testing for the devices, and the expertise
and training needed to utilize these devices as adjunctive sources of
information for clinical decision making for suspicious skin lesions.
This is consistent with the indications for use of approved software-
aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer (see
Refs. 2 and 3). While different types of skin lesion analyzers may have
different intended users,\8\ different users present different risks,
and the Agency believes that the assessment of a clinician trained in
the diagnosis and management of skin cancer is important for the safe
and effective use of this device type.
---------------------------------------------------------------------------
\8\ FDA has separately classified ``software-aided adjunctive
diagnostic device[s] for use by physicians on lesions suspicious for
skin cancer'' into class II under 21 CFR 878.1830. Please see De
Novo classification order DEN230008, available at <a href="https://www.accessdata.fda.gov/cdrh_docs/pdf23/DEN230008.pdf">https://www.accessdata.fda.gov/cdrh_docs/pdf23/DEN230008.pdf</a>. The
publication of this classification in the Federal Register and
codification in the Code of Federal Regulations are currently
pending at the time of publication of this final order.
---------------------------------------------------------------------------
(Comment 5) Several commenters provided feedback on what may be
considered the ground truth as a comparator for device performance in
clinical studies for a software-aided adjunctive diagnostic device for
use on skin lesions by physicians trained in the diagnosis and
management of skin cancer. Other commenters provided feedback on what
type of data should be considered as an input for these devices. These
comments generally agreed that histologically confirmed diagnosis of
biopsied tissue assessed by a trained dermatopathologist or a panel of
trained dermatopathologists would be appropriate as a ground truth
comparator for clinical studies, though one comment noted that biopsy
of non-suspicious lesions may not be needed to establish ground truth
in a clinical trial because a large number of benign lesions are
already considered suspicious enough to biopsy and can be used to form
a ground truth for benign lesions. Some commenters stated that
dermoscopic images should be used as inputs for these devices, rather
than standard images or other types of input data.
(Response 5) FDA agrees that in many cases, histologically
confirmed diagnosis by a trained dermatopathologist or a panel of
trained dermatopathologists provides an appropriate ground truth
comparator for use in clinical studies of a software-aided adjunctive
diagnostic device for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer, though the exact source of
information used in clinical trials as a comparator may be specific to
each clinical trial design, including the range and number of
suspicious and non-suspicious lesions sampled for the trial. FDA also
believes that it is important to consider that, as technology and
clinical practices evolve, additional comparators may be possible with
appropriate justification and supportive valid scientific evidence. To
that end, the clinical performance special controls include a
requirement that justification must be provided for the determination
of ground truth.
FDA also acknowledges that dermoscopic images may be used as inputs
for software-aided adjunctive diagnostic devices for use on skin
lesions by physicians trained in the diagnosis and management of skin
cancer. However, we note that current devices utilize other types of
inputs, including multispectral images or electrical impedance
measurements, and as discussed elsewhere in this document, FDA has
determined that general controls and the special controls established
in this final order effectively mitigate the risks to health identified
for these devices.
(Comment 6) Multiple comments provided input regarding statistical
performance targets and study endpoints for software-aided adjunctive
diagnostic devices for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer, including the extent to which
such targets and endpoints should be specified in special controls.
Generally, these comments indicated that
[[Page 14452]]
sensitivity, specificity, accuracy, and other indicators of statistical
performance and validity should be assured to be appropriate for each
device based on the intended use and intended user, and that there is
not a single set of statistical performance targets that would be
appropriate for every device and every indication, though some
commentators suggested that the special controls should include
specific targets for statistical performance requirements. Some
commenters emphasized the importance of ensuring that performance goals
and study requirements are achievable for each type of study conducted
to support device performance, and that use of a software-aided
adjunctive diagnostic device for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer should improve
physicians' ability to detect skin cancer without causing an
unacceptable increase in systemic care burden, particularly if the care
burden is unnecessary biopsies.
Specific feedback regarding statistical performance included a
recommendation that suitable studies of software-aided adjunctive
diagnostic devices for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer should demonstrate sensitivity
with a lower bound of the confidence interval greater than 90 percent.
One commenter recommended that device sensitivity for the indicated
skin cancer(s) should be superior to unaided providers in the study or
in literature, and that the sum of device sensitivity and specificity
should be greater than one to demonstrate that the device performance
is statistically meaningful and is a beneficial adjunct to inform
dermatologist decision making. Another commenter recommended that,
ultimately, the primary effectiveness criterion should be that use of
the device improves dermatologists' correct assessment of whether to
biopsy a lesion suspicious for skin cancer.
(Response 6) FDA agrees that the device, when used by a provider
trained in the diagnosis and management of skin cancer as an adjunctive
device to aid in the evaluation of lesions suspicious for skin cancer
following identification of a suspicious skin lesion, should enable
superior accuracy in the provider's decision-making process for
assessing skin lesions compared to unaided providers. FDA also agrees
that the performance goals and study endpoints for clinical studies
used to test these devices should be reasonable and achievable. The
performance special controls established in this final order therefore
include clinical testing with clinically justified endpoints for device
sensitivity and specificity relative to ground truth for a
representative range of individuals with different risk factors and a
justified quantity and range of mimic lesions, which are benign skin
lesions that are visually very similar to malignant lesions.
Performance data must also demonstrate superior accuracy of device-
aided users' diagnostic characterization of lesions compared to that of
unaided users. For lesions with high metastatic potential, FDA believes
that a sensitivity of at least 90 percent, or a clinically justified
alternative, is necessary to mitigate the risks related to false
negatives, and as such has established requirements in the special
controls to that effect. Device specificity and other endpoints should
be clinically justified and appropriate for the specific indications
and labeling to mitigate the risks related to false positives, which
may also reduce any potential increase in care burden such as from
unnecessary biopsies.
(Comment 7) Multiple comments provided recommendations for study
design considerations related to patient demographics and lesion
selection. Specific recommendations included that there should be a
sufficient number of subjects across multiple clinical sites in studies
to allow for subgroup analysis. Demographic considerations recommended
in these comments included ethnicity, sex, age, low/intermediate/high-
risk populations, and Fitzpatrick skin type \9\ of the intended patient
population. Some comments noted that patients with darker Fitzpatrick
skin types (IV-VI) are often underserved in dermatology and
underrepresented in clinical studies related to assessing skin lesions;
these comments supported that clinical trials should utilize
investigational patient populations representative of the intended
patient population of the device. Some comments further stated that
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer ought
to be effective at assessing lesions in patients with darker
Fitzpatrick skin types, while other comments stated that a phased
approach to considering different skin types would be appropriate. One
comment asserted that special controls addressing labeling may be used
to minimize risk in underrepresented populations. Comments also
recommended that a broad range of lesion types, sizes, and anatomic
sites should be included in clinical studies. Additionally, commenters
stated that the classification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer should reflect the difference in risk
associated with devices indicated for use for melanoma as compared to
devices indicated for use for other types of lesions and skin cancers.
---------------------------------------------------------------------------
\9\ The Fitzpatrick skin type classification scale is a
recognized standard in dermatology used to estimate the response of
different types of skin to ultraviolet (UV) light. Fitzpatrick skin
types range from Type I to VI where individuals with Type I always
burn and never tan (pale white skin; blond or red hair; blue eyes;
freckles) and individuals with Type VI never burn (deeply pigmented
skin; dark hair and eyes). Fitzpatrick skin type is an independent
risk factor for skin cancer, notably melanoma, basal cell carcinoma,
and squamous cell carcinoma. See National Cancer Institute,
``Genetics of Skin Cancer (PDQ)--Health Professional Version,''
available at <a href="https://www.cancer.gov/types/skin/hp/skin-genetics-pdq">https://www.cancer.gov/types/skin/hp/skin-genetics-pdq</a>
(last updated on May 9, 2025); Fitzpatrick, T.B., ``The Validity and
Practicality of Sun-Reactive Skin Types I Through VI,'' Arch
Dermatol, 124(6):869-71, 1988, available at <a href="https://doi.org/10.1001/archderm.124.6.869">https://doi.org/10.1001/archderm.124.6.869</a>.
---------------------------------------------------------------------------
(Response 7) FDA believes that clinical studies supporting the
safety and effectiveness of a software-aided adjunctive diagnostic
device for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer should consider patient populations
representative of the intended patient population for the device and a
wide variety of lesions to support the specific indication(s) and
intended patient population for the device. The special controls state
that clinical testing must evaluate patients across risk factors
(including age, body site, skin phototype, and other clinical factors
as applicable) that represent the intended patient population and that
analysis of standalone performance must include subgroup analysis by
relevant risk factors to help demonstrate that the device performs as
intended in the intended patient population. The special controls
require both premarket clinical performance validation testing and
post-market surveillance, in part to provide a reasonable assurance of
safety and effectiveness in patients with relatively lower incidence of
skin cancer, unless FDA determines based on the totality of the
premarket data that data from post-market surveillance is not required.
The clinical performance special controls also state that a
justification must be provided for the quantity and range of mimic
lesions per diagnosis used in testing, that data must demonstrate
superior accuracy of device-aided users' diagnostic characterization of
the indicated lesions compared to the accuracy of unaided
[[Page 14453]]
users, and that testing must demonstrate at least 90 percent
sensitivity of the device output for lesions with high metastatic
potential (or an alternative clinical consideration must be provided to
justify lower sensitivity), thereby contributing to the risk mitigation
for higher risk types of skin cancers, and different types of skin
cancers and lesions. The special controls established in this final
order therefore mitigate the risk associated with use for different
lesion types, including lesions with high metastatic potential such as
melanoma.
Additionally, FDA agrees that labeling special controls, among the
other special controls established in this final order, are necessary
to provide a reasonable assurance of safety and effectiveness of
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer. The
labeling special controls state that device labeling must include,
among other things, information about device performance for all
clinically relevant subgroups within the intended patient
population(s), a description of the patient population that was used in
development or training of the device algorithm, and information about
subpopulations for which the device may not perform as expected or for
whom the device has not been validated.
(Comment 8) Multiple comments stated that FDA should conduct
premarket inspections and establish post-market requirements for
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer
including post-market surveillance studies, manufacturing facility
inspections, and annual reports. One comment stated that databases
should be maintained to help ensure the safety of patients on which
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer are
used. Some commenters raised concerns regarding how future changes to
device hardware or software, especially artificial intelligence and
machine learning (AI/ML) algorithms, might be managed under the 510(k)
paradigm and suggested that PMA annual reporting requirements or PMA
supplement requirements may be the only way to provide appropriate
oversight of such changes, and that such changes may need to be
supported by additional clinical data. Another commenter supported the
proposed reclassification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer and suggested that special controls
should require adequate documentation on AI/ML data management,
training, and validation.
(Response 8) FDA agrees with parts of these comments. The special
controls established in this final order include a requirement that
data obtained from post-market surveillance demonstrate that the device
performs as intended in the intended patient population and under
anticipated conditions of use, unless FDA determines based on the
totality of the premarket data that data from post-market surveillance
is not required. For any such required post-market surveillance, FDA
believes that to conduct the surveillance in a timely fashion, a
manufacturer generally should submit a complete study protocol for the
post-market surveillance study consistent with the special control
requirements within 30 days of receipt of a 510(k) decision letter
finding the device substantially equivalent to a predicate device. FDA
generally expects to work with the manufacturer to approve the study
protocol within 90 days of that letter. In addition, to ensure the
surveillance is conducted in a timely fashion, FDA believes that from
the date of protocol approval, the first study subject should generally
be enrolled within 6 months; 20 percent of subjects should generally be
enrolled within 12 months; 50 percent of subjects should generally be
enrolled within 24 months; and 100 percent of subjects should generally
be enrolled within 36 months. FDA also believes that manufacturers
should generally submit post-market study progress reports every year
until subject enrollment has been completed, and annually thereafter,
from the date of the 510(k) decision letter; that if any enrollment
milestones are not met, manufacturers should generally submit
enrollment status reports every 6 months in addition to the annual
post-market study progress reports, until FDA notifies the manufacturer
otherwise; and that manufacturers should generally submit a final post-
market study report 3 months from study completion (i.e., last
subject's last follow-up date). FDA anticipates that specific timelines
may be discussed with the manufacturer at the time of and/or following
receipt of the 510(k) decision letter.
FDA also maintains the MDR database, MAUDE database, and Medical
Device Recall database, which allows for additional post-market
surveillance of these devices and helps to ensure continued safety for
marketed devices.
FDA does not agree that premarket site inspections are necessary to
provide reasonable assurance of safety and effectiveness for these
devices. There is a low risk of batch variability in the manufacturing
of these devices, and the hardware of these devices can generally be
characterized with well-established methods and standards. The special
controls identified in this final order establish requirements for
validating both software and hardware components of the devices
premarket, including that testing must include a description of
compatible hardware and processes, pre-specified compatibility testing
protocols, and dataset(s). The special controls also establish
requirements relating to device precision, electromagnetic
compatibility, and electrical, mechanical, and thermal safety,
biocompatibility, and software verification, validation, and hazard
analysis. Additionally, routine or for-cause inspections, which may
consider compliance with quality management system requirements \10\
applicable to the manufacturing of the device, allow for post-market
oversight of these devices with respect to inspections.
---------------------------------------------------------------------------
\10\ On February 2, 2024, FDA issued a final rule amending the
device Quality System Regulation, 21 CFR part 820, to align more
closely with international consensus standards for devices (89 FR
7496). This final rule took effect on February 2, 2026. This rule
withdrew the majority of the previous requirements in part 820 and
instead incorporated by reference the 2016 edition of the
International Organization for Standardization (ISO) 13485, Medical
devices--Quality management systems--Requirements for regulatory
purposes, in part 820. As stated in the final rule, the requirements
in ISO 13485 are, when taken in totality, substantially similar to
the requirements of the previous part 820, providing a similar level
of assurance in a firm's quality management system and ability to
consistently manufacture devices that are safe and effective and
otherwise in compliance with the FD&C Act.
---------------------------------------------------------------------------
FDA also disagrees that the PMA annual reporting requirements
(distinct from annual reporting associated with a post-market
surveillance study) or PMA supplement requirements are necessary to
ensure that changes to a software-aided adjunctive diagnostic device
for use on skin lesions by physicians trained in the diagnosis and
management of skin cancer that has received marketing authorization
have appropriate oversight. Under the 510(k) paradigm, a new 510(k) is
required for any change or modification to a cleared device that could
significantly affect the safety or effectiveness of the device, or for
a major change or modification in
[[Page 14454]]
intended use.\11\ Information on more minor changes, such as might be
submitted in a PMA supplement or reported in a PMA periodic report, is
not needed for reasonable assurance of safety and effectiveness here
considering that the intended use of the devices being reclassified is
limited to adjunctive use after a provider has identified a suspicious
skin lesion and is not for use to confirm a clinical diagnosis, and
because, as discussed in section IV, the special controls established
in this final order, in addition to general controls, are sufficient to
mitigate the risks to health that may be associated with the use of a
software-aided adjunctive diagnostic device for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer.
---------------------------------------------------------------------------
\11\ In accordance with 21 CFR 807.81(a)(3), a 510(k) is
required for significant changes or modifications to a device in
design, components, method of manufacture, or intended use, which
include: (1) a change or modification that ``could significantly
affect the safety or effectiveness of the device, e.g., a
significant change or modification in design, material, chemical
composition, energy source, or manufacturing process,'' or (2) ``a
major change or modification in the intended use of the device.''
---------------------------------------------------------------------------
Additionally, manufacturers may wish to use predetermined change
control plans (PCCPs) as a way to implement future modifications to
their devices without needing to submit a new 510(k) for each
significant change or modification \12\ while continuing to provide a
reasonable assurance of device safety and effectiveness.\13\ FDA
reviews a PCCP as part of a marketing submission for a device to ensure
the continued safety and effectiveness of the device without
necessitating additional marketing submissions for implementing each
significant change or modification described in the PCCP. When used
appropriately, PCCPs authorized by FDA are expected to be least
burdensome for manufacturers and FDA.\14\
---------------------------------------------------------------------------
\12\ For the purpose of this final order, reference to a
``significant change or modification'' means a significant change or
modification that would generally require a new premarket
notification under 21 CFR 807.81(a)(3).
\13\ Section 3308 of the Food and Drug Omnibus Reform Act of
2022, Title III of Division FF of the Consolidated Appropriations
Act, 2023, Public Law 117-328 (``FDORA''), enacted on December 29,
2022, added section 515C ``Predetermined Change Control Plans for
Devices'' to the FD&C Act. Section 515C has provisions regarding
PCCPs for devices requiring premarket approval or premarket
notification. Under section 515C, supplemental applications (section
515C(a)) and new premarket notifications (section 515C(b)) are not
required for a change to a device that would otherwise require a
premarket approval supplement or new premarket notification if the
change is consistent with a PCCP approved or cleared by FDA.
\14\ Sections 513 and 515 of the FD&C Act. See also FDA's
guidance, ``The Least Burdensome Provisions: Concepts and
Principles; Guidance for Industry and Food and Drug Administration
Staff,'' Feb. 5, 2019. Available at <a href="https://www.fda.gov/media/73188/download">https://www.fda.gov/media/73188/download</a>.
---------------------------------------------------------------------------
Based on all available information, including feedback from the
July 2022 reclassification panel meeting, the Agency believes it likely
that changes to the AI/ML algorithm in a software-aided adjunctive
diagnostic device for use on skin lesions by physicians trained in the
diagnosis and management of skin cancer could significantly impact
device effectiveness or safety. As such, these changes would likely
require a new 510(k) (unless the changes are implemented consistent
with a cleared PCCP), as would changes to device hardware (including
signal capturing hardware), the device output, or other software
aspects that could significantly impact device safety or effectiveness.
Examples of such changes include expansion or modification of the AI/ML
algorithm training data, modification to cut-off values or thresholds
used to determine device output, and addition, removal, or modification
of device outputs. It is important to note that the devices subject to
this reclassification do not utilize adaptive algorithms (i.e.,
algorithms that evolve dynamically due to continuous learning while
they are used). FDA also notes that FDA's quality management system
requirements include documentation requirements related to, among other
things, device modifications and validation (see 21 CFR part 820).
FDA agrees that some changes to device software or hardware may
require new clinical data, which may require premarket review within
the context of a new premarket submission.
(Comment 9) Several comments expressed concerns related to
layperson use of mobile phone-based software applications intended to
provide information about or diagnose skin lesions. Commenters stated
that these devices have different risks and may have greater risks than
the software-aided adjunctive diagnostic devices for use on skin
lesions by physicians trained in the diagnosis and management of skin
cancer being reclassified in this order, and stated that such mobile
phone-based software applications intended for use by lay persons
should not be a part of this reclassification. Others expressed a
concern that reclassification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer might be interpreted as indicating an
Agency position on layperson, mobile phone-based devices with similar
intended uses. Some commenters provided input regarding requirements
that should be established for mobile phone-based software applications
intended for use by lay persons on skin lesions, such as that they
should use digital dermoscopy images, not smartphone images.
(Response 9) FDA agrees that mobile phone-based or other web-based
software applications intended for use by lay users to provide
information about or diagnose skin lesions have different risks than
the devices being reclassified in this final order. The devices subject
to this reclassification order are not mobile phone or web-based
applications available to lay users. At the time of publication of this
final order, FDA has not classified, cleared, approved, or granted
authorization for a layperson use mobile phone-based or web-based
application intended to provide diagnostic or adjunctive information
about skin lesions, and layperson, mobile phone-based or web-based
devices are not within the scope of this final order. Comments related
to the technological characteristics of mobile phone applications
intended for use by lay users are outside the scope of this
reclassification order.
(Comment 10) One comment stated that FDA's method of notifying the
dermatologic community about the opportunity for public comment on the
proposed reclassification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer was inadequate, noting that certain
interested persons were not informed that the topic was under
discussion.
(Response 10) FDA disagrees that its method of informing interested
persons of the opportunity for public comment on the proposed
reclassification of software-aided adjunctive diagnostic devices for
use on skin lesions by physicians trained in the diagnosis and
management of skin cancer was inadequate. FDA's procedures were
conducted consistent with the Agency's regulations under 21 CFR 860.134
and section 513(f)(3) of the FD&C Act, including publication of a
proposed order in the Federal Register (87 FR 39025), provision of a
60-day comment period for interested persons to submit comments to a
public docket, and the convening of a meeting of the appropriate
classification panel to discuss the proposed reclassification
(including the establishment of an additional docket for public comment
and opportunities for interested persons to attend and/or present data,
[[Page 14455]]
information, or views at the Panel meeting). These procedures provided
meaningful opportunity for public comment; FDA received comments from
more than 50 commenters on the proposed order and Panel meeting,
including comments from a variety of entities such as individual
medical professionals, professional societies, and members of the
medical device industry. FDA therefore believes that the Agency
provided sufficient opportunity for interested parties to comment on
the proposed reclassification of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer.
IV. Changes in the Final Order
As described in sections II and III of this document, FDA has made
revisions in this final order in response to feedback from the Panel
and comments regarding the proposed reclassification order that were
submitted to public dockets.
Based, in part, on the Panel feedback and comments regarding the
proposed reclassification order, FDA has revised the list of risks to
health associated with the use of software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer, the special controls that FDA has
determined will mitigate these risks, and Table 1, ``Risks to Health
and Mitigation Measures for Software-Aided Adjunctive Diagnostic
Devices for Use on Skin Lesions by Physicians Trained in the Diagnosis
and Management of Skin Cancer''.
FDA has identified the following risks to health associated with
the use of software-aided adjunctive diagnostic devices for use on skin
lesions by physicians trained in the diagnosis and management of skin
cancer:
<bullet> False negative results or false positive results: False
negative results could result in incorrect or delayed diagnoses and
delays in skin cancer treatment. These delays may allow an undetected
condition to worsen and potentially increase skin cancer associated
morbidity and mortality. False positive results may result in
complications such as incorrect management of the patient, including
unnecessary invasive biopsy procedures and more frequent screenings, as
well as the potential administration of inappropriate treatments and/or
the withholding of appropriate treatments, with possible adverse
effects.
<bullet> Use error or improper device use: Use error or improper
device use could lead to false results or failure to generate a result.
The device could be misused to analyze data from an unintended patient
population, an unintended anatomical site, or lesions having an
unintended attribute, or to analyze data acquired with incompatible
hardware or incompatible acquisition settings, potentially resulting in
the device not operating at its expected performance level. The device
could also be misused if the user does not follow the appropriate use
protocol for using the device to assess lesions of interest. Examples
of not following the appropriate use protocol include overreliance on
the device output or not using the device in an adjunctive manner,
i.e., using the device output alone to make a patient management
decision, which may lead to lower accuracy. Inaccurate results may
result in the same complications associated with false negative or
false positive results as previously discussed.
<bullet> Device failure or malfunction: Device failure or
malfunction could result in the absence or delay of device output, or
incorrect device output, which could lead to inaccurate patient
assessment. Inaccurate results may result in the same complications
associated with false negative or false positive results as previously
discussed.
<bullet> Electrical, thermal, mechanical, or light exposure-related
injury: While in operation, the device may discharge electricity that
could shock the user or patient. Electrical discharge or exposure to
device-generated heat may cause thermal injury or discomfort. Moving
parts may cause mechanical injury. For devices that utilize energy
(e.g., light) to provide adjunctive diagnostic information, accidental
eye exposure to the energy source could cause eye injury.
<bullet> Interference with other devices: Individuals with
electrically powered implants could experience an adverse interaction
with the device due to electromagnetic interference or radiofrequency
interference.
<bullet> Adverse tissue reaction: A patient could experience skin
irritation and/or allergic reaction associated with non-biocompatible
materials in patient-contacting components of the device.
<bullet> Infection and cross contamination: If components of the
device that must be sterile are not adequately sterilized or if
reusable components are not adequately reprocessed (i.e., are not
cleaned and sterilized or disinfected) between uses, the device may
introduce pathogenic organisms to patients which may result in an
infection.
FDA has determined that the following special controls will
mitigate these risks to health, and that these special controls, in
addition to general controls, will provide a reasonable assurance of
safety and effectiveness for software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer:
<bullet> The risk of false negative results or false positive
results can be mitigated through clinical performance testing--
including standalone testing that demonstrates at least 90 percent
sensitivity of the device output for lesions with high metastatic
potential (or an alternative clinical consideration must be provided to
justify lower sensitivity), and with a clinical justification provided
for the reported specificity--as well as non-clinical performance
testing and post-market surveillance (unless FDA determines based on
the totality of the premarket data that data from post-market
surveillance is not required). The clinical performance testing must
demonstrate superior accuracy of device-aided users' diagnostic
characterization of the indicated lesions compared to the accuracy of
unaided users. The non-clinical performance testing, among other
information, must demonstrate that the device performs as intended
under anticipated conditions of use, including a description of
compatible hardware and processes, pre-specified compatibility testing
protocols, and dataset(s). In addition, post-market surveillance data
may address potential differences in device performance as observed in
retrospectively designed premarket studies versus real-world device use
and provide additional information regarding device performance in the
overall patient population, including in low-incidence patient sub-
populations for whom relatively limited data was available at the time
of premarket review. The risk of false positive results and false
negative results can be further mitigated by special controls that
require information in labeling to provide detailed instructions for
use and inform the user of the expected device performance for all
clinically relevant subgroups within the intended patient population.
<bullet> The risk associated with use error or improper device use
can be mitigated by performance testing that demonstrates device
precision, including repeatability and reproducibility of device
performance, across operators and challenging use conditions. In
addition, this risk can be mitigated by requiring that the device
labeling include information regarding performance of the device for
all
[[Page 14456]]
clinically relevant subgroups within the intended patient population,
as well as a description of the patient population that was used in
development or training of the device algorithm. This risk can be
further mitigated by special controls that require the device labeling
to include information related to the limitations of device performance
or subpopulations for which the device may not perform as expected or
for whom the device has not been validated. The risk resulting from not
following the device instructions for use can be mitigated by special
controls that require a human factors assessment to demonstrate that
intended users can correctly use the device according to the intended
use. This risk can be further mitigated by requiring that the device
labeling include information needed to facilitate interpretation of all
device outputs and identification of the risks associated with
misinterpretation of the device outputs, and by special controls
requiring that the device labeling provide a description of user
training required prior to use and a statement that the device is not
intended for use as a standalone diagnostic.
<bullet> The risk of device failure or malfunction can be mitigated
by requiring non-clinical performance testing and software
verification, validation, and hazard analysis, and by requiring
performance testing that demonstrates device precision, including
repeatability and reproducibility of device performance, across
operators and challenging use conditions. This risk can be further
mitigated by requiring that instructions for device maintenance and
validated methods and instructions for reprocessing of any reusable
components be included in the labeling.
<bullet> The risk of electrical, thermal, mechanical or light-
related hazards leading to user injury or discomfort can be mitigated
by special controls that require testing that demonstrates electrical,
mechanical, and thermal safety; software verification, validation and
hazard analysis; and device labeling that includes instructions on
appropriate usage and maintenance of the device. The risk of eye injury
due to energy (e.g., light) exposure can be mitigated by special
controls that require labeling that warns users about exclusion of
lesions close to the eye and unsafe exposure to any energy-emitting
components of the device.
<bullet> The risk that the device may interfere with other devices
due to radiofrequency or electromagnetic interference can be mitigated
by requiring testing that demonstrates electromagnetic compatibility.
<bullet> The risk of adverse tissue reaction for patient-contacting
devices can be mitigated by special controls that require elements of
the device that may contact the patient to be demonstrated to be
biocompatible and labeling that includes, in addition to user
qualifications needed for safe use of the device, instructions for
device maintenance and validated methods and instructions for
reprocessing of any reusable components.
<bullet> The risks of infection and cross contamination for
patient-contacting components can be mitigated by special controls that
require sterilization validation, shelf-life testing, and labeling that
includes validated methods and instructions for reprocessing of any
reusable components (i.e., cleaning and sterilization or disinfection).
Table 1--Risks to Health and Mitigation Measures for Software-Aided
Adjunctive Diagnostic Devices for Use on Skin Lesions by Physicians
Trained in the Diagnosis and Management of Skin Cancer
------------------------------------------------------------------------
Identified risk to health Mitigation measures
------------------------------------------------------------------------
False negative results or Clinical performance testing. Post-market
false positive results. surveillance. Non-clinical performance
testing. Labeling.
Use error or improper device Precision testing. Human factors testing.
use. Labeling.
Device failure or malfunction Non-clinical performance testing.
Precision testing. Software
verification, validation, and hazard
analysis. Labeling.
Electrical, thermal, Electrical, mechanical, and thermal
mechanical, or light safety testing. Software verification,
exposure-related injury. validation, and hazard analysis.
Labeling.
Interference with other Electromagnetic compatibility testing.
devices.
Adverse tissue reaction...... Biocompatibility evaluation. Labeling.
Infection and cross Sterilization validation. Shelf-life
contamination. testing. Cleaning and disinfection
validation. Labeling.
------------------------------------------------------------------------
V. The Final Order
In this final order, FDA is adopting relevant findings from the
June 30, 2022, proposed order (87 FR 39025). FDA has made revisions in
this final order in response to the Panel deliberations (see section
II) and comments received (see section III). FDA is issuing this final
order to reclassify optical diagnostic devices for melanoma detection
and electrical impedance spectrometers from class III into class II
under a new device classification regulation with the name software-
aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer, and
to establish special controls by revising 21 CFR part 878 (adding 21
CFR 878.1820). The identification for Sec. 878.1820(a)(1) has been
revised to provide a more accurate description of the devices in this
classification regulation.
Further, in this final order, FDA has identified the special
controls under section 513(a)(1)(B) of the FD&C Act that, along with
general controls, provide a reasonable assurance of the safety and
effectiveness for software-aided adjunctive diagnostic devices for use
on skin lesions by physicians trained in the diagnosis and management
of skin cancer. In this final order, the Agency has made refinements to
the special controls as previously described in the proposed order to
further mitigate the risks to health associated with the use of
software-aided adjunctive diagnostic devices for use on skin lesions by
physicians trained in the diagnosis and management of skin cancer.
Specifically, and among other things, FDA added new special controls
requiring that data from post-market surveillance demonstrate that the
device performs as intended in the intended patient population and
under anticipated conditions of use (unless FDA determines based on the
totality of the premarket data that data from post-market surveillance
is not required); requiring that standalone testing demonstrate at
least 90 percent sensitivity of the device output for lesions with high
metastatic potential (or an alternative clinical consideration must be
provided to justify lower sensitivity), and that a clinical
[[Page 14457]]
justification be provided for the reported specificity; requiring that
clinical testing evaluate patients across risk factors that represent
the intended patient population, including age, body site, skin
phototype, and other clinical factors as applicable; requiring that
performance testing demonstrate device precision, including
repeatability and reproducibility of device performance, across
operators and challenging use conditions; and requiring that labeling
include a description of the patient population that was used in
development or training of the device algorithm.
Under the FD&C Act, 510(k) submissions are required to reasonably
assure the safety and effectiveness of class II devices unless FDA
determines that the device type should be exempt under section
510(m).\15\ FDA has not made this determination for software-aided
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer and, therefore,
this class II device type is not exempt from 510(k) requirements. Thus,
under sections 510(k) and 513(f) of the FD&C Act, persons who intend to
market this device type must submit a 510(k) containing information on
the software-aided adjunctive diagnostic device for use on skin lesions
by physicians trained in the diagnosis and management of skin cancer
that they intend to market and must obtain FDA clearance of the device
prior to marketing it.
---------------------------------------------------------------------------
\15\ In considering whether to exempt class II devices from
premarket notification, FDA considers whether premarket notification
for the type of device is necessary to provide reasonable assurance
of safety and effectiveness of the device. FDA generally considers
the factors initially identified in 63 FR 3142 (January 21, 1998)
and further explained in FDA's guidance ``Procedures for Class II
Device Exemptions from Premarket Notification, Guidance for Industry
and CDRH Staff'' to determine whether premarket notification is
necessary for class II devices. FDA also considers that even when
exempting devices from the 510(k) requirements, these devices would
still be subject to certain limitations on exemptions, for example,
the general limitations set forth in 21 CFR 878.9.
---------------------------------------------------------------------------
Under this final order, software-aided adjunctive diagnostic
devices for use on skin lesions by physicians trained in the diagnosis
and management of skin cancer are prescription use devices under Sec.
801.109 (21 CFR 801.109). Prescription devices are exempt from the
requirement for adequate directions for use for the layperson under
section 502(f)(1) of the FD&C Act (21 U.S.C. 352(f)(1)) and 21 CFR
801.5, as long as the conditions of Sec. 801.109 are met. The device
would continue to be subject to the submission and device clearance
requirements of sections 510(k) and 513 of the FD&C Act and of part
807, subpart E, of FDA's regulations (21 CFR part 807).
VI. Effective Date
This final order is effective 30 days after the date of its
publication in the Federal Register.
VII. Analysis of Environmental Impact
The Agency has determined under 21 CFR 25.34(b) that this action is
of a type that does not individually or cumulatively have a significant
effect on the human environment. Therefore, neither an environmental
assessment nor an environmental impact statement is required.
VIII. Paperwork Reduction Act of 1995
This final administrative order refers to previously approved
collections of information found in FDA regulations. The previously
approved collections of information are subject to review by the Office
of Management and Budget (OMB) under the Paperwork Reduction Act of
1995 (44 U.S.C. 3501-3521). The collections of information in 21 CFR
part 820 (Quality Management System Regulation) have been approved
under OMB control number 0910-0073; the collections of information in
21 CFR part 812 (Investigational Device Exemptions) have been approved
under OMB control number 0910-0078; the collections of information in
part 807, subpart E (Premarket Notification Procedures), have been
approved under OMB control number 0910-0120; the collections of
information in 21 CFR part 822 (Postmarket Surveillance) have been
approved under OMB control number 0910-0449; and the collections of
information under 21 CFR part 801 (Device Labeling) have been approved
under OMB control number 0910-0485.
IX. Codification of Orders
Under section 513(f)(3) of the FD&C Act, FDA may issue final orders
to reclassify devices. FDA will continue to codify classifications and
reclassifications in the Code of Federal Regulations (CFR). Changes
resulting from final orders will appear in the CFR as newly codified
orders. Therefore, under section 513(f)(3) of the FD&C Act, we are
codifying in this final order the classification of software-aided
adjunctive diagnostic devices for use on skin lesions by physicians
trained in the diagnosis and management of skin cancer in the new Sec.
878.1820, under which these devices are reclassified from class III
into class II.
X. References
The following references are on display at the Dockets Management
Staff (HFA-305), Food and Drug Administration, 5630 Fishers Lane, Rm.
1061, Rockville, MD 20852, 240-402-7500, and are available for viewing
by interested persons between 9 a.m. and 4 p.m., Monday through Friday;
they are also available electronically at <a href="https://www.regulations.gov">https://www.regulations.gov</a>.
Although FDA verified the website addresses in this document, please
note that websites are subject to change over time.
1. FDA, July 28-29, 2022, Meeting of the General and Plastic Surgery
Devices Panel Meeting Materials (available at <a href="https://www.fda.gov/advisory-committees/advisory-committee-calendar/july-28-29-2022-general-and-plastic-surgery-devices-panel-medical-devices-advisory-committee-meeting">https://www.fda.gov/advisory-committees/advisory-committee-calendar/july-28-29-2022-general-and-plastic-surgery-devices-panel-medical-devices-advisory-committee-meeting</a>).
2. P090012 Approval Order, available at <a href="https://www.accessdata.fda.gov/cdrh_docs/pdf9/P090012A.pdf">https://www.accessdata.fda.gov/cdrh_docs/pdf9/P090012A.pdf</a>.
3. P150046 Approval Order, available at <a href="https://www.accessdata.fda.gov/cdrh_docs/pdf15/P150046A.pdf">https://www.accessdata.fda.gov/cdrh_docs/pdf15/P150046A.pdf</a>.
4. FDA, ``Requests for Feedback and Meetings for Medical Device
Submissions: The Q-Submission Program; Guidance for Industry and
Food and Drug Administration Staff,'' May 29, 2025. Available at
<a href="https://www.fda.gov/regulatory-information/search-fda-guidance-documents/requests-feedback-and-meetings-medical-device-submissions-q-submission-program">https://www.fda.gov/regulatory-information/search-fda-guidance-documents/requests-feedback-and-meetings-medical-device-submissions-q-submission-program</a>.
List of Subjects in 21 CFR Part 878
Medical devices.
Therefore, under the Federal Food, Drug, and Cosmetic Act (21
U.S.C. 321 et seq., as amended) and under authority delegated to the
Commissioner of Food and Drugs, 21 CFR part 878 is amended as follows:
PART 878--GENERAL AND PLASTIC SURGERY DEVICES
0
1. The authority citation for part 878 continues to read as follows:
Authority: 21 U.S.C. 351, 360, 360c, 360e, 360j, 360l, 371.
0
2. Add Sec. 878.1820 to subpart B to read as follows:
Sec. 878.1820 Software-aided adjunctive diagnostic device for use on
skin lesions by physicians trained in the diagnosis and management of
skin cancer.
(a) Identification. A software-aided adjunctive diagnostic device
for use on skin lesions by physicians trained in the diagnosis and
management of skin cancer is a device that uses a software algorithm to
analyze optical or other physical properties of a skin lesion and
returns a classification of the skin
[[Page 14458]]
lesion. The device is intended for prescription use by a physician
trained in the clinical diagnosis and management of skin cancer (e.g.,
a dermatologist) as an adjunctive device to aid in the evaluation of
lesions suspicious for skin cancer following identification of a
suspicious skin lesion. The device is not intended for use as a
standalone diagnostic and is not for use to confirm a clinical
diagnosis.
(b) Classification. Class II (special controls). The special
controls for this device are:
(1) Data obtained from premarket clinical performance validation
testing, or a combination of premarket clinical performance validation
testing and post-market surveillance (in accordance with paragraph
(b)(2) of this section), must:
(i) Demonstrate superior accuracy of device-aided users' diagnostic
characterization of the indicated lesions compared to the accuracy of
unaided users in the intended patient population and under anticipated
conditions of use;
(ii) Include an evaluation of patients across risk factors
(including age, body site, skin phototype, and other clinical factors
as applicable) that represent the intended patient population under
anticipated conditions of use; and
(iii) Include standalone device performance testing that
demonstrates the accuracy of the device output relative to ground truth
in the intended patient population and under anticipated conditions of
use, including the following:
(A) Testing must demonstrate at least 90% sensitivity of the device
output for lesions with high metastatic potential, or an alternative
clinical consideration must be provided to justify lower sensitivity.
Clinical justification must be provided to support the reported
specificity.
(B) Lesions must be selected by representative users (e.g.,
dermatologists) and a justification must be provided for the quantity
and range of mimic lesions per diagnosis.
(C) Justification must be provided to support the determination of
ground truth.
(D) Testing must include a representative range of individuals with
different risk factors (including age, body site, skin phototype, and
other clinical factors as applicable), and analysis of standalone
performance must include subgroup analysis by relevant risk factors.
(2) Data obtained from post-market surveillance must demonstrate,
in consideration of the premarket data obtained in accordance with
paragraph (b)(1) of this section, that the device performs in
accordance with paragraph (b)(1) of this section, unless FDA
determines, based on the totality of the premarket data, that data from
post-market surveillance is not required to demonstrate that the device
performs as intended. Such post-market surveillance must be conducted
per a protocol determined appropriate by FDA to demonstrate that the
device performs as intended (in consideration of the premarket data
obtained in accordance with paragraph (b)(1) of this section), and must
include initiation, enrollment, and reporting requirements to ensure
timely periodic updates to FDA on post-market surveillance progress and
outcomes.
(3) Non-clinical performance testing must demonstrate that the
device performs as intended under anticipated conditions of use,
including compatibility testing of the device software with specific
signal or image acquisition hardware. Testing must include a
description of compatible hardware and processes, pre-specified
compatibility testing protocols, and dataset(s).
(4) Performance testing must demonstrate device precision,
including repeatability and reproducibility of device performance,
across operators and challenging use conditions.
(5) Performance testing must demonstrate electromagnetic
compatibility, and electrical, mechanical, and thermal safety of any
electrical components of the device.
(6) Performance testing must validate reprocessing instructions for
reusable components of the device.
(7) Sterilization validation must be conducted for components that
must be sterile. Performance testing must also demonstrate continued
sterility and package integrity of components that must be sterile, as
well as continued device functionality, over the identified shelf life
of the device.
(8) The patient-contacting components of the device must be
demonstrated to be biocompatible.
(9) Software verification, validation, and hazard analysis must be
performed.
(10) A human factors assessment must demonstrate that the device
can be safely used by intended users.
(11) Labeling must include:
(i) A summary of standalone and clinical performance testing
conducted with the device. The summary must describe performance
measures, including sensitivity and specificity, and statistical
confidence intervals, as well as performance of the device for all
clinically relevant subgroups within the intended patient population;
(ii) A description of the patient population that was used in
development or training of the device algorithm;
(iii) Information related to the limitations of device performance
or subpopulations for which the device may not perform as expected or
for whom the device has not been validated;
(iv) Information needed to facilitate interpretation of all device
outputs, and identification of the risks associated with
misinterpretation of the device outputs;
(v) A statement that the device is not intended for use as a
standalone diagnostic and is not for use to confirm a clinical
diagnosis;
(vi) User qualifications needed for safe use of the device,
including a description of user training required prior to use, and a
statement that the device is intended to be used by a physician trained
in the clinical diagnosis and management of skin cancer (e.g., a
dermatologist);
(vii) Warnings to avoid unsafe exposure to any energy-emitting
components of the device (e.g., excluding use of the device on lesions
close to the eye); and
(viii) Instructions for device maintenance and validated methods
and instructions for reprocessing of any reusable components.
Grace R. Graham,
Deputy Commissioner for Policy, Legislation, and International Affairs.
[FR Doc. 2026-05772 Filed 3-24-26; 8:45 am]
BILLING CODE 4164-01-P
</pre><script data-cfasync="false" src="/cdn-cgi/scripts/5c5dd728/cloudflare-static/email-decode.min.js"></script></body>
</html>Indexed from Federal Register on March 25, 2026.
This is legal information, not legal advice. Laws vary by jurisdiction and change frequently. Always verify current law with official sources and consult a licensed attorney in your jurisdiction for advice on your specific situation.