Re: Is radon as dangerous as the EPA sa

[JMUCKERHEIDE@delphi.com]

John Moulder writes:
in response to John Cameron:
 
> > The main point is that radon at ordinary levels is not 
> > a health hazard.
> 
> What does "not a health hazard" mean?
> 
> 1 -- To many, it means no statistically significant effect (at the levels to 
> which people are actually exposed) in reasonably designed studies (lab or 
> epi).  
> 
> 2 -- But this is not everybody's definition.  At the other extreme, the EPA is 
> willing to declare something a health hazard based on extrapolated risks of as 
> low as 1 in 10^5 (or is it 1 in 10^6?).
> 
> 3 -- There are also many people in this world who are willing to label 
> something as a health hazard on basis of the absence of data proving absolute 
> safety (look at the power line cancer controversy).

Right, there are many ignorant people in the world!?  :-)   And others who
take advantage of them. I suppose John should have said "no matter how you
look at it technically" or ... but this is a technical forum so we should
forgive his presumption that . OTOH: 

> > Bernie's work gives strong evidence that supports that conclusion.
> 
> It does is you use the first definition of a "health hazard", but not if you 
> use the other two.

This isn't quite accurate in the context of the statement. Cohen's data shows
unambiguous results that are roughly 22 standard deviations below EPA's
predicted "health effects". The EPA predicted "health effects" are simply
unjustified mathematical projections over orders of magnitude from miner data
(which itself ignores confounding factors in the miner data). 

This "data" also ignores the lack of effects in miners and other
highly-exposed populations to roughly 1000 times mean residential exposure
levels. In addition, the histology of lung cancer in non-smokers and in miners 
are different, further undercutting the biological plausibility of the
association between radon at residential levels and radon-caused effects.  

> > It strongly contradicts the LNT model no matter how you look at it.
> 
> I'm not sure about that.  The EPA radon model predicts that no increase in 
> lung cancer should be detectable for the radon levels to which people are 
> exposed in the vast majority of residences.  How can you "contradict" a model 
> that predicts "no detectable effect"?   Arguably, the only way to contradict 
> the model would be to show a threshold.  But if you really believe that there 
> is no lung cancer risk associated with residential radon, you cannot 
> contradict the EPA model.

It seems that a statistically significant negative slope in the range of
interest is at least as significant as "to show a threshold"; and the actual
data 22 SD below the EPA "prediction" is significant. It is disingenuous to
argue that "you can't see an effect at EPA's predicted levels so you can't
know they are false", ignoring the data.  (It is appropriate to be mindful
that the sole basis and result of EPA regs is to capture a sufficiently large
fraction of the housing stock to cause action, which would cause the public
expenditure of $20 billion for testing and remediation services. Then, of
course, if the public were stupid enough to do that, it would later be
ratcheted down to engage more of the residences and exponentially increase
costs and enhance bureaucratic programs.) 

> > I have long sought a published debate on the LNT model. It 
> > is urgently needed. The short anti-LNT articles in the June '95 
> > HPS Newsletter and Dan Strom's long pro-LNT article rebutting them was 
> > not a debate in any sense. 
> 
> Agreed, and in a more widely distributed forum than the HP Newsletter.  I'd 
> suggest HP itself.  Alternatively, if anyone want's to make a specific 
> proposal I will take it to the Rad Res Editorial Board.

I'll expect John to share the Rad Res response to his proposals of last year.
:-) 

There is a "Fishbowl Management/Planning" structured process of open, dynamic
participation, organized by alternative, topic and pro/con response. You print 
intermediate draft results (that encourages new participation, and new data,
without "starting over", and discourages saying the same thing over again
_LOUDER_  :-). 

Its very effective with a knowledgeable, impartial editor/facilitator. It's
not used since those with a weak case, who see winning by pandering and
appealing to public and political interests, won't participate. OTOH, when
Sierra Club and others refused, the facilitator recorded known arguments,
distributed the docs to the interested public, and they found that they
couldn't win by the standard effort to seek later court injunctions and to
mislead the body politic about the process and results, or that the agency had 
"not addressed alternative x, or topic y"; or they had not had an opportunity
to to be heard. 

It also fails because bureaucratic staff and "planners" and "managers" must
give up autocratic power over "studies" and results -- this is the real
failure mode for the process as a toold to be used by gov't agencies  :-) 

Thanks.

Regards, Jim Muckerheide
jmuckerheide@delphi.com