Re: Environmental Epidemiology and Toxicology "Improvements in lo

[Bob Flood <bflood@SLAC.Stanford.EDU>]

The remark about the plant affect San Francisco is remarkably inapproriate.
By the reason given any such effects should also be seen in every other city
that receives food from that part of California. However, there's no
intention of actually examining whether the effect is really there - it's a
blind accusation with no supporting data, but it uses the forum of a
scientific journal to suggest credibility. However, it is nothing more than
speculation, and the journal should never have agreed to publish the article
with it included.

The author acknowledges that 3/4 of the Sacremento population live upwind of
the plant site, which makes them irrelevent to the study, but they included
them in the study anyway. I wonder what their inclusion did to the study's
result. Given that they are the overwhelming majority of people included in
the study, any effectsm good or bad, in the rest of the poopulation studied
would be masked. It is unlikely that this is an oversight or an accident.

Perhaps someone can educate me on basic math here: if 424 deaths represent a
rate of 9.5 deaths per 1000 births, then there were 4,028 births in the area
and time period of interest, right? But if there were 271 fetal deaths,
which is given as a rate of 6.1 deaths per 1000 births, there must have been
1,653 births. How can this be?

The study makes no effort to exclude or adjust for non-radiogenic causes of
death other than to exclide accidents, homicide, and suicide (probably not a
huge correction to infant death rates). It therefore attributes all deaths
or at least all changes in the incidence of infant death to radiation. Novel
approach to medicine.

I especially liked the I-131 data. no details in the data are given (must
have been an oversight), but my experience tells me that 2.67 and 1.82 pCi
(presumably per liter - he didn't actually say) are probably with one or one
and a half standard deviations of each other, meaning the difference is
smaller than the uncertainty in the measurements. But the 1.82 pCi is for
the calendar year beginning 6 months after the plant shut down! Perhaps the
8-day half-life for I-131 doesn't apply to epidemiology.

It would appear from the inclusion of irrelevent populations and unsupported
claims about effects 70 miles from the area of interest that this study
started with the answer and derived the question.


************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html