A "substantial proportion" of malignant lesions are found incidentally among patients referred to dermatologists for a different lesion flagged by the referring physician or health care professional, according to a report in the May issue of Archives of Dermatology.
In fact, almost half of all the 149 skin cancers identified in a cohort of 400 patients were not the suspect lesions for which the patients had been referred. They were instead found incidentally on other parts of the body by the dermatologist and had been missed by the referring clinician.
This finding calls into question the widespread use of teledermatology for triaging patients with suspected skin cancer lesions. "If the reviewing dermatologist has access only to a digital image of a specific lesion rather than interaction with the patient, other malignant lesions and/or lesions of concern that are outside the field of digital transmission may be overlooked," wrote Dr. Kate V. Viola of Yale University, New Haven, Conn., and her associates.
The investigators were interested in determining the proportion of suspicious lesions that actually turned out to be malignant, so they examined the records of 400 Veterans Affairs patients with a single suspect lesion who were referred to dermatologists at two major academic medical centers and six community-based clinics between 2006 and 2009.
As VA patients, 98% of the cohort comprised older white men. The mean age was 78 years, and 19% of the study subjects had a history of skin cancer.
The referring clinicians included attending physicians and residents in internal medicine, nurse practitioners, and physician assistants.
Most (56%) of the 400 suspicious lesions flagged by these referring clinicians were immediately ruled out as nonmalignant by the dermatologist on clinical examination and did not require biopsy.
A total of 176 index lesions were biopsied, and 88 of them – 22% of the original 400 – proved to be malignant.
Most of the malignancies were basal cell carcinomas (61 lesions), approximately one-fourth were squamous cell carcinomas (21 lesions), and 5 were melanomas.
The dermatologists, however, also identified and biopsied an additional 111 incidental lesions that usually occurred on different parts of the body from the index lesions. Just over half of these incidental lesions (61) proved to be malignant.
"Of great concern, [10%] of the incidental lesions discovered by the dermatologist were melanomas," Dr. Viola and her colleagues noted.
For 12 of these 61 incidental skin cancers (20%), the index lesion had been immediately ruled out as nonmalignant. Thus, if the dermatologists had looked no further than the flagged index lesion, 12 malignancies would never have been found, the investigators said (Arch. Dermatol. 2011;147:556-60).
Most of the melanomas were missed by the referring clinicians, they added.
The study findings clearly demonstrate that assessment of a single specific lesion of concern – which is typical in teledermatology – may lead to "underdiagnosis of clinically significant lesions that are not appreciated by the referring physician. Therefore, teledermatology must not be used as a substitute for a total body skin examination when skin cancer is suspected," the investigators wrote.
Dr. Viola and her associates did not report having any conflicts of interest.