Task force maybe too stringent in not yet recommending melanoma screening
The U.S. Preventive Services Task Force announced today that it doesn't have enough evidence to recommend that clinicians perform visual screening for melanomas with patients with no known special risk for the skin cancer. In an invited commentary in JAMA—the journal of the American Medical Association—Drs. Martin Weinstock and Hensin Tsao agree that the evidence doesn't meet the task force's standards, but they also question whether those standards are appropriate.
"If you were to take a poll among practicing dermatologists, you'd find the vast majority believe that early detection reduces risk of death from melanoma," said Weinstock, a professor of dermatology in the Warren Alpert Medical School of Brown University and chief of dermatology at the Providence Veterans Affairs Medical Center. "Skin self-examination and clinician skin examination are a means of early detection. That's the predominant tool that we have. It makes eminent sense that primary care doctors should be trained to do examination for melanoma."
But Weinstock acknowledged that in the task force's typical hierarchy of evidence, the "gold standard" is not the opinions of experts, but systematic reviews of data from a series of well-controlled randomized clinical trials.
There isn't, and may never be, such evidence regarding melanoma screening, Weinstock said. Clinical trials to determine whether widespread melanoma screening would prevent deaths have to be very large to be definitive, principally because deaths from the condition are not common. No country has yet been willing to spend the many millions of dollars required to fund a large enough trial.
Instead the U.S. Preventive Services Task Force reviewed the mixed evidence from controlled observational and epidemiological studies. At least some of those suggest both that early screening could save lives and that presumed harms such as overdiagnosis, unnecessary specialist visits and unnecessary surgeries don't emerge when properly trained clinicians and even laypeople conduct visual screening, Weinstock said.
Given a likelihood of little harm but a potential life-saving benefit, Weinstock and Tsao of Massachusetts General Hospital argue, the task force could consider a different standard for judging the procedure.
"The evidentiary standard needs to be further refined to be appropriate to the modest magnitude of potential harms of a properly performed skin cancer screening," they wrote.
Skin cancer screening, Weinstock noted, is not an invasive procedure like a colonoscopy. For exams with potential serious medical consequences, the highest standards make clear sense. But for melanoma screening, a clinician simply looks at the skin, often while examining a patient during a routine office visit. Meanwhile, freely available training, such as the online course INFORMED that Weinstock helped to create, includes advice on when to appropriately reassure patients that skin lesions are not cancerous.
In their article, Weinstock and Tsao raised four other questions about the task force's approach to considering visual skin screenings.
- Overdiagnosis: While it is likely to a degree, Weinstock said, that is true of many screenings that are recommended, including for lung or breast cancer. If screening can save lives, it may still be worthwhile.
- Assessing risk: To determine whether someone has no known special risk, such as large numbers of moles or atypical moles, a clinician would have to do essentially the same visual inspection that would be done to screen for melanoma anyway.
- Advances in medications: New treatments for melanoma thankfully may save lives, but that could make it even harder to devise a definitive study that would show whether more widespread screening also saves lives.
- Self-examination: The task force is preparing a separate report on consumers conducting their own screening. Weinstock calls that an "artificial distinction" in that the dialogue between doctors and patients routinely means that self-screening and clinical screening are closely linked.
In the end, Weinstock and Tsao acknowledge that more evidence would be helpful. Weinstock is continuing to study whether presumed harms, such as excess anxiety, could emerge from expanded screening. But they also question how much evidence should be considered enough.
"Going forward, it is imperative to develop the requisite evidence and the appropriate evidentiary standards to advance this area of public health," they wrote.