We Are Sunshine

Mayo Study Built On 1 Case A Year

Thursday, April 5th, 2012

Mayo Dermatology’s baseless attack on indoor tanning this week is predicated entirely on one small Minnesota county that averaged just 1 melanoma a year in women 18-39 in the 1970s — as small a sample as is statistically possible for comparison.

Olmsted County, Minnesota — a county of 144,000 that is home to the Mayo Clinic — now diagnoses about 9 melanomas a year in women 18-39, an increase Mayo calls 8-fold after adjusting for the increased population. Even worse, in promoting its one-county study, Mayo suggested that increase — measured in a county with 15 times more dermatologists-per-capita than the national average — is somehow more accurate than the much-larger National Cancer Institute national database, which did not show that increase.

By saying 1 case increased to 9 over a 40-year span, they convinced the American news media to run with the story anyway this week. And they did – as if this study were a national database with large numbers. Headlines suggested that “Melanoma cases rising – young women at greatest risk.” (The Detroit Free Press)

That’s NOT what the study said. The study alleged an increase in women — but even Mayo admitted that melanoma is STILL much more common in men than in women, and that older men are most at risk. But that’s NOT what they led reporters write this week.

“They shoved the term ‘8-fold increase’ down reporters throats and didn’t explain that ‘8-fold’ meant increasing from 1 case to 9 cases total in an entire county, and that there wasn’t an increase in thick lesions in that county or that mortality rates actually declined in that county, which isn’t possible if incidence truly is increasing,” Smart Tan Executive Director Joseph Levy said. “That entire 40-year increase is predicated on that county generating less than one more melanoma case a month over a 40-year period. And there were 68 dermatologists doing it. Increased detection, and not increase in disease, is so obviously the default explanation for what happened in Olmsted and why it didn’t actually happen in the rest of the country.”

It gets worse. Mayo didn’t bother to report that in 1970 that county detected less than half than the national average for melanomas in women — about 5.4 per 100,000 verses about 11 per 100,000 nationwide. Olmsted’s rates were LOWER than the rest of the nation. When you start from a lower baseline and you begin DETECTING more aggressively, and you have 15 times more dermatologists per capita DOING that detection, that’s one way to suggest that there’s a bigger increase.

It gets STILL worse. Mayo didn’t bother to discuss that its data show no measurable increase in THICK melanomas in the 40 years in the study. There weren’t ANY in that category in the 1970s. There were just 2 in the last decade, which is actually a big reduction from the 1990s, when they found 5 in 10 years.

Mayo didn’t bother to tell you that the dermatology community worldwide has written that you can’t have an ACTUAL increase in disease if all you are detecting is an increase in THIN lesions that you didn’t used to even find. That’s the subject of peer-reviewed studies in the past two years in the Journal of the National Cancer Institute and the British Journal of Dermatology.

All of that on top of the errors we’ve documented in the past two days. That the National Cancer Institute’s data don’t suggest what Mayo alleges. That Dermatology itself attacked Mayo for a similar paper in 2005. And that the rest of the world’s melanoma data don’t match Mayo’s.

The story keeps getting worse. As does Mayo Dermatology’s academic integrity.