Is Melanoma Overdiagnosed? What Rising Incidence and Flat Mortality Reveal

Much of the rise in melanoma is real disease that would never have hurt anyone. US diagnoses have grown roughly sixfold across four decades, yet the death rate barely shifted for most of that stretch, and that pairing is the textbook fingerprint of overdiagnosis. The lesions are genuine melanomas by every pathology rule, but a large fraction of them, especially the thinnest and the in situ cases, would never have caused symptoms or shortened a life. Saying so does not mean melanoma is harmless or that finding a dangerous tumor early is pointless. It means the numbers we count and the deaths we prevent have drifted apart.

Key points#

• Overdiagnosis means finding a true cancer that would never have caused harm, not making a wrong diagnosis.
• US melanoma incidence rose about 459 percent from 1975 to 2017 while mortality stayed nearly flat until roughly 2013.
• In situ (earliest-stage) diagnoses grew fastest of all, a pattern hard to explain except by indolent detection.
• More skin biopsies and lower diagnostic thresholds plausibly drive much of the increase.
• Population skin-cancer screening has never been shown to lower melanoma deaths, per the USPSTF.

What overdiagnosis actually means#

It helps to name the thing precisely, because the word is easy to misread. Overdiagnosis is not a mistake at the microscope. The pathologist is right, and the tissue truly meets the definition of melanoma. The trouble is that cancers vary enormously in tempo. Some race; some crawl; some sit still for the rest of a person's life. When a lesion belongs to that last group, finding and removing it cannot add a single day to your life, because it was never headed anywhere. What you get instead is the treatment and the label.

Epidemiologists cannot look at one mole and know its future, so they reason from populations instead. The logic is simple: if a wave of genuinely dangerous disease were sweeping through, more diagnoses today should mean more deaths tomorrow. When cases surge for decades and deaths do not follow, the extra cases are most likely the slow and the still, not the deadly ones arriving late.

The curve that started the argument#

An analysis by Kurtansky and colleagues in the Journal of Investigative Dermatology followed US data from 1975 to 2017 and found melanoma incidence up about 459 percent. The clearest signal sat in women aged 55 to 74, where new diagnoses rose year after year while deaths held essentially steady. The authors read that stubborn flat line beneath a rising one as the single sharpest sign of overdiagnosis in the whole dataset.

The pattern grows starker at the earliest stage. Melanoma in situ, confined to the outermost skin layer, climbed by roughly 4,675 percent over the same span, far outracing invasive disease and still rising when the data ended. A category can only expand that violently while almost no one in it dies if what is expanding is detection, not danger.

Where the extra cases come from#

Rising counts have an ordinary explanation that has nothing to do with more cancer: more looking, and more sensitive labeling of what we find. Welch and colleagues, writing in the New England Journal of Medicine, reported that the share of fee-for-service Medicare beneficiaries receiving a skin biopsy nearly doubled after 2004, and melanoma diagnoses in that group climbed right alongside. More biopsies surface more tiny lesions, and the threshold for calling a specimen malignant has drifted over time, so slides once read as benign now come back as melanoma. The authors judged intensified skin examination a more plausible engine for the rise than a hidden surge of lethal tumors.

Two honest complications#

A fair account has to hold two caveats in view rather than tidy them away.

The first is timing. Melanoma mortality did finally begin to fall around 2013, driven by genuinely effective new therapies for advanced disease. That is a treatment success, not a retreat of overdiagnosis, and if you read the fall carelessly it can paper over the decades-long gap between rising cases and steady deaths.

The second is a statistical trap called lead-time bias. Diagnose the same cancer earlier and the patient appears to survive longer counting from the diagnosis date, even if the day of death never moves. Because both overdiagnosis and lead time inflate survival without touching mortality, an improving five-year survival figure tells you little about whether early detection is saving lives. Deaths, not survival counted from diagnosis, are the honest scoreboard.

The comparison that put a number on it#

One of the more ingenious efforts to size the problem used a built-in control group. Adamson, Suarez, and Welch argued in JAMA Dermatology that melanoma biology does not differ enough across racial groups to explain wildly different diagnosis trends, so the experience of Black patients, examined far less intensively, can stand in for the true change in disease burden. From 1975 to 2014, melanoma incidence rose about sixfold in White men and roughly fourfold in White women, but by under 25 percent in Black patients. Treating the Black trend as the counterfactual, the authors estimated that around 60 percent of melanomas in White men and 59 percent in White women diagnosed in 2014 represented overdiagnosis. The confidence intervals are wide, but the direction lines up with the incidence-and-mortality picture.

Why screening the whole population has not helped#

If earlier detection reliably prevented death, screening everyone should bend the death curve. It has not. In 2023 the US Preventive Services Task Force again issued an I statement, concluding that the evidence is insufficient to weigh the benefits against the harms of visual skin examination in asymptomatic adults, and finding no reliable proof that such screening lowers melanoma mortality across a population. That verdict does not call skin cancer trivial. It records a real cost, the biopsies, scars, dollars, and worry, sitting next to a benefit that has never shown up in the deaths.

Reading the numbers without overcorrecting#

None of this softens the danger of a thick, ulcerated, or changing lesion, and none of it argues for ignoring a spot on your own skin that is growing or bleeding. The overdiagnosis case is deliberately narrow: most of the growth in melanoma statistics comes from small, early, often in situ lesions whose detection has not translated into fewer deaths. Hold that distinction and you can see why the claim that more skin checks straightforwardly equals more lives saved runs ahead of the evidence, and why pathologists and epidemiologists are working on ways to tell the lesions that need treating from the ones that never will. The honest stance keeps two things at once: humility about which single mole is which, and clarity that the population curves point to substantial overdiagnosis.