Page 42 - Noninvasive Diagnostic Techniques for the Detection of Skin Cancers

Next Steps

Only three trials examined the use of dermoscopy versus naked eye examination in the
evaluation of suspicious skin lesions. Only one reported diagnostic accuracy of dermoscopy
55
compared with naked eye examination. The focus of the other two studies did not concern
diagnostic accuracy (one examined the time needed to complete exam with and without
59
56
dermoscopy; the other studied the referral rate for biopsy with and without dermoscopy ) A
trial based on a larger number of dermatologists and primary care physicians in the U.S. and
using biopsy result as the reference standard and reporting on both intermediate (e.g., number of
lesions) and clinical outcomes (e.g., survival) would be informative.
Key informant input suggests that the main challenge to the use of dermoscopy in a primary
care setting lies in the training (or lack of) of nonexperts in its technique. As most of the studies
on training focused on dermatologists, it would be important to investigate the kinds of training
appropriate for primary care practitioners before this technique could be incorporated in a
primary care setting.
Because of the wide range of reported diagnostic sensitivities and specificities for the
standard dermoscopic algorithms in use, studies should be conducted to better understand the
determinants of this reported variability.
A limited number of studies reported on the use of computer-aided diagnoses of suspicious
skin lesion. It would be informative to further evaluate the performance of automated computer
diagnostic instruments vis-à-vis the ability of an experienced clinician to diagnose skin cancer. It
is also important to analyze further specific attributes of images captured by these noninvasive
devices (including dermoscopy, confocal laser scanning microscopy, etc.) for their use in
computer-aided diagnosis (such as a neural network classifier).
In contrast to the widespread use of many of photodynamic therapies for NMSC and benign
skin disorders, the clinical use of photodynamic diagnostic techniques is still investigational. At
present, its principal role may be in defining the borders of suspicious lesions, particularly for
NMSC, and in differentiating tumor tissue from normal tissue. Whether the use of alternative
light sources, like laser, with photodynamic compounds, is helpful in diagnosing melanoma
deserves further exploration.
Even though intermediate/process outcomes are important in the use of these noninvasive
technologies (e.g., decreasing the interval between referral to specialist and diagnosis and
initiation of treatment) to further appreciate their impact, long-term follow up studies with
patient-centered outcomes, such as survival and reduction of unnecessary biopsies, should also
be undertaken.
To improve understanding of the diffusion of these different technologies, it would be useful
to explore how specific population/patient/practice settings (e.g., rural versus urban, individuals
with or without previous history of cancer or of different ethnic/racial backgrounds, availability
of trained specialists) affect the use and adoption of these newer technologies.

33