Page 14 - Noninvasive Diagnostic Techniques for the Detection of Skin Cancers

Methods

Systematic Searches of the Published Literature

®
We conducted an electronic literature search in MEDLINE for articles published between
1990 and March 2011.Our search strategy utilized both National Library of Medicine Medical
®
Subject Headings (MeSH ) and search terms. The identification of search terms for noninvasive
modalities was guided by information from the earlier Topic Refinement phase of this project.
To develop search terms for skin cancer, we used a previously published skin cancer systematic
18
review as a reference. Search terms for noninvasive modalities were crossed with search terms
for skin cancer. We restricted our search to literature published within the last 20 years because
all noninvasive modalities of interest are fairly new. We also restricted our search to English-
language human subject studies. Because the specific focus of the search was not screening and
detection for specific types of skin cancer, (e.g., melanoma or nonmelanoma), we did not seek to
identify or formally evaluate existing guidelines. Selected examples from the National Guideline
Clearinghouse (www.guideline.gov) are included for background information only. The details
of our search strategies are listed in Appendix A. In particular, we noted that while ‘dermoscopy’
®
has been used in MEDLINE since 2005 as an inclusive term for dermatoscopy,
epiluminescence microscopy, or surface microscopy, studies prior to 2005 needed to be
identified by the alternative terminologies. The first 200 abstracts were screened jointly by all
project investigators to ensure that screening criteria were well understood and applied
uniformly. Thereafter, investigators screened non-overlapping sets of the remaining citations. In
general, abstracts served as the primary source of information, which were supplemented by full-
text review of systematic reviews (n=5) and all randomized controlled trials (RCTs) (n=11). We
®
did not do a pre-MEDLINE search.
Because our report primarily focused on the use of noninvasive modalities for the early
diagnosis of skin cancer, we excluded studies of metastases detection, therapeutic interventions,
18
or modalities used exclusively for pre-surgical planning. Specifically, we did not consider F-
FDG positron emission tomography (PET), computed tomography (CT), single photon emission
computed tomography (SPECT), and lymphoscintigraphy in this technology brief because these
techniques are not generally used in initial screening of suspicious skin lesions or establishing
the diagnosis of skin cancer, but rather, in delineating loco-regional or distance spread of disease.
In addition, given our focus on the evaluation of primary cutaneous lesions, we also excluded
evaluation of subungal, ocular, and visceral sites of disease. Appendix B, Table 1 summarizes
our inclusion and exclusion criteria.

Searches of the Grey Literature

We performed an Internet search for key words to identify more recent noninvasive
techniques/modalities not in current widespread use, such as fluorescence, high-resolution
ultrasound, laser Doppler, bio-electrical impedance, polarized light photography, 3-D histograms
of color mapping, and thermography. For these searches, unless otherwise advised, we used the
Google search engine, and, for each search string entered, we perused the first 10 pages to
identify relevant links.
In addition to our Internet search, we reviewed major vendors or manufacturers’ Web sites
for information pertaining to the different noninvasive modalities. To identify major vendors or

6