Page 15 - Noninvasive Diagnostic Techniques for the Detection of Skin Cancers

manufacturers and to obtain FDA clearance status of relevant devices (see Appendix C, Table
C1), we searched the FDA Center for Devices and Radiological Health (CDRH) database. We
compiled a list of noninvasive devices using information from published reviews as well as
information gleaned from our Technical Experts. For potential harms with the relevant devices,
we queried the FDA Manufacturer and User Facility Device Experience (MAUDE) database for
any reported harms with the use of the relevant devices.
Also, we searched the ClinicalTrials.gov registry for any ongoing or completed relevant trials
involving the various noninvasive modalities of interest. We used both device categories and
®
name of device (such as MoleMax and Vivascope ) and device category when searching the
registry (see Appendix C, Table C2).

Interviews With Key Informants
A representative panel of Key Informants was identified through the earlier Topic
Refinement phase of this project. These individuals included medical experts/practitioners in
dermatology, oncology, and family medicine, a patient with skin cancer and a representative of
patient advocacy group, as well as scientists and representatives of professional societies.
Technical experts within the Key Informant group were individually interviewed by phone or
email for their responses to Guiding Questions 1, 2, and 4. Interview questions were tailored to
the unique perspective and expertise of each Technical Expert. The final structure of the brief in
reference to classification of technologies by degree of clinical use was also reviewed. All
external participants completed disclosure of interest forms and any conflicts were considered
and balanced.

Data Collection

Data from abstracts of qualified studies were extracted into an Excel spreadsheet. We
extracted information on items of interest (population, intervention, comparator, outcome
[PICO]), using customized forms, including information specific to the particular noninvasive
modality. We also extracted data on the citation (first author name, journal, and year of
publication), condition being evaluated, study size and setting, particular noninvasive modality,
and details relevant to the technical specification of the particular noninvasive modality. We
categorized, wherever possible, qualified studies into the two major categories: reviews, which
included systematic review, general or narrative review, and technical report; or primary studies,
which included RCTs, comparative cohort, non-comparative cohort or case series, diagnostic
tests, or single case reports. Studies that did not fit into one of these two groups were considered
“other.” An abstract was classified as a systematic review if it described the study as a meta-
analysis, pooled analysis, or systematic review. An abstract was also classified as a systematic
review if it reported that there were inclusion or exclusion criteria in selecting the studies. We
documented the outcomes as test accuracy (e.g., sensitivity, specificity, false positive rate, false
negative rate), clinical outcomes (e.g., survival), characterization of the skin lesion (e.g., lesion
size, lesion color), and other. The “other” category pertained to intermediate outcomes (e.g.,
number of lesion) or process outcomes (e.g., patient satisfaction, healthcare utilization), training
for optimizing the use of device, or description of use.

7