23 May 2010

On "facts"

I've just responded to a sponsorship appeal for a medical charity. That's not the issue. But I am a little concerned about the acknowledgement email, which says;
Thank you very much for your kind donation. 1 in 50 will suffer from MM [Malignant Melanoma] during their lifetime, 1 in 4  of those sufferers will die from it. These are FACTS ...
Are they? This suggests that 0.5% (1 in 200) of all deaths, from all causes, worldwide, are accounted for by this particular form of cancer. I've pursued this on the net, and I can find no evidence to support such broad-brush claims. (Show me and I'll take this post down and publish a retraction.)

"According to a WHO report about 48,000 melanoma related deaths occur worldwide per year." On that basis the total annual death toll of humanity would be about 9.6m p.a. But if there are about 6bn of us, and assuming optimistically that all of us live 100 years, even those figures would produce a death toll of 60m p.a. ...

I, like most respondents to the appeal, am not really bothered about this kind of statistic. Indeed, an appeal to the rarity of a condition might be equally potent in eliciting a donation. But it does make me uneasy when such assertions are made without evidence. It undermines my confidence in such appeals.


  1. Perhaps overstated, but not so wildly out: http://info.cancerresearchuk.org/cancerstats/types/skin/incidence/index.htm

  2. The writer has picked the two worst figures possible from the literature, which do not match each other. The 1 in 50 figure is that for the white population in Australia, and the 1 in 4 figure is an outdated UK one.

    The massive rise in diagnosis for MM in the UK has been matched by a greater drop in lethality than would have been be expected as a result simply of earlier diagnosis, such that some writers are suspecting much of the declared increase represents false positives brought about by more stringent monitoring.

    Histopathological diagnosis of MM is notoriously difficult, as this rather boldly titled paper discusses: http://www.medicine.ox.ac.uk/bandolier/band37/b37-2.html

  3. Great aunt alice9:18 pm

    Love the mangled stats. There are lies...


Comments welcome, but I am afraid I have had to turn moderation back on, because of inappropriate use. Even so, I shall process them as soon as I can.