RE: rem vs. rad

["Fulmer, Philip" <philip.fulmer@cplc.com>]

Hi, Keith.

I would think that we would run into a problem eventually at high doses
when overkill becomes a problem.  At some point, with the cells being
killed by the radiation, the probability of cancer induction/mortality
in these same cells is zero and the overall risk would be lessened,
resulting in a lower "RBE" with cancer as the endpoint.  Basically, we
have competing biological effects that are more likely at the higher
doses.  So while we can theoretically calculate dose equivalent at any
absorbed dose,  the conclusions that we draw from it at high absorbed
doses would be limited at some point, although I'm not sure what that
point would be.  For example, would the exceedance of any nonstochastic
threshold meet the test?  

Just to be on the safe side, I think I'll just try to avoid getting a
high absorbed dose.

Philip  

__________________________
Philip C. Fulmer, PhD, CHP 
Carolina Power & Light Company
Harris Energy and Environmental Center
3932 New Hill-Holleman Road
P. O. Box 327
New Hill, NC 27562-0327
philip.fulmer@cplc.com
(919) 362-3363      

>
>I think we're asking the wrong question.  Instead of basing the cutoff on
>dose, base it on effect.  Use rem for ANY dose when estimating stochastic
>risks.  Even if the dose is high, if I want a stochastic risk estimate, I
>have to use rem (would you try to estimate cancer risk using rads?).  As
>Bob Flood pointed out, the whole stochastic risk model - complete with
>radiation weighting factors (ie.Q) was "founded" on high doses, and if I'm
>estimating risk for a stochastic effect, by definition I have to use rem.
>For non-stochastic effects, one should use rad to determine, for instance
>if the dose is above the threshold for some particular effect.  
>Example:  I rapidly get 400 rem uniform whole body dose (just to make it
>interesting, it's from neutrons with Q=2, so absorbed dose is 200 rad).  To
>estimate my cancer risk, I use rem.  To evaluate the potential for and
>severity of a non-stochastic effect, I use rad.
>
>
>Keith Welch
>Thomas Jefferson National Accelerator Facility
>Newport News VA
>welch@cebaf.gov
>Ph: (757)269-7212
>FAX:(757)269-5048
>