29 July 2006

Why do people die?

Like crime statistics, statistics on the causes of death need to be considered carefully. No such careful consideration for The Scotsman, whose lead story today offers drama and moral opprobrium together:
"A RECORD number of women died from alcohol abuse in Scotland last year, new figures have revealed.
Experts said the increasing death toll was being fuelled by a generation of women whose attitude towards drinking was formed by the liberal values of the 1960s and 70s.
Statistics from the Registrar General for Scotland, published yesterday, showed 492 women died of alcohol-related diseases last year, compared with 441 in 2004. The increase was highest among women aged 30 to 60."

The full report is available here.

Gee, those liberal values of the 1960s and 70s have a lot to answer for (although I do not recall anything in those values that actually encouraged young women to get legless every saturday night). Nevertheless, it is perhaps worth noting that 492 female deaths in Scotland in 2005 amounted to less than 2 per cent of the total female deaths in Scotland in that year. Secondly, how much of the increase between 2004 and 2005 was due to the relative willingness of doctors to assign the cause of death to alcohol-related diseases? Doctors are human, after all, and perhaps it is becoming more acceptable to grieving relatives to be more honest on the death certificate. Thirdly, we all have to die of something: if the death of women in childbirth or of tuberculosis becomes less frequent, other causes inevitably become more prominent.

So can we trust the statistics? Sure, this is the GRO after all. Just don't be too quick to draw conclusions from them.

As an example, I offer you this from the same Scotsman article:
"By contrast, the number of alcohol- and smoking-related deaths among men dropped over the same period."

Nobody would suggest that this means we can go back to smoking in pubs. So let us not get too excited about the reverse trend for women. And, please, let us not assign blame to liberal values without good reason.

No comments: