|
John,
In a private email to you I wrote the response
(below) to a different email you sent directly to me so it may be a bit
difficult for others to track. In your previous email I thought
you were referring to the fact that the best use of ecologic studies is to form
hypotheses, not test them. Nonetheless, my post was a response
to Dr. Cohen regarding BEIR which showed that the relative risks based on
cumulative radon exposure are linear, but with a dependency on age, pack-year
rate of smoking, smoking duration, etc. Dr. Cohen is using an ecologic
study to test the validity of the LNT. My previous point was to test
the LNT or LNTassumption by using collective dose assumptions does not
add up to me. The reason it does not add up is the association between
radon exposure and risk will not be linear using his ecologic data
unless age, pack-year rate - smoking intensity, smoking duration, etc. are
considered, which is impossible using ecologic data. So if you
do not include these factors as he does not, he is not testing the LNT.
When he states the LNT fails, I consider it more a failure of his
study design to accommodate the non linear dependencies of pack year rate,
smoking duration, etc.
John, again, my goal is not to convince anyone,
but to state why I personally have questions with his ecologic
analyses. I do think this type of study design to test the LNT is
fatally flawed. I do think more work
is needed that will either verify the validity for certain types of
radiation or disprove the validity of the LNT, however I do not think the
examination should take the form of an ecologic study.
Bill
|