On 10/03/2011 00:24, Andy Dingley wrote:
On Mar 9, 8:53 pm, Andy wrote:
You can't really tell whether some of
the cases were associated with something else,

You can't with lung diseases, but you can with mesothelioma. There's
really nothing much else that causes it.

Again, how do you know?

BTW it says here

http://en.wikipedia.org/wiki/Mesothelioma

Mesothelioma can be associated with glass dust.


Legally there's no requirement to prove that asbestos caused a
condition anyway. The legal test is that it _probably_ caused it,
which is accepted as being proven if the exposure caused at least a
doubling of risk (sounds low, but actually it makes statistical
sense). A doubling of risk is not an epidemiologically serious
increase in risk - in most studies of rare conditions with small
sample groups it's hard to demonstrate significance - but it can be
enough to show cause.

As you say, much easier to prove. The odd 1% error in the numbers is
then unimportant.

Andy