Re: Cancer Assessment Press Release

[tmohaupt@wright.edu]

Hello Radsafe colleagues:

Steve Wing's findings of elevated cancer incidence of multiple 
myeloma at DOE sites was interesting.  He found the odds ratio for 
workers with doses >5 rem compared with workers with dose < 1rem to 
be 4.34.  Interestingly, the National Cancer Institute SEER Report 
(1973-1991) shows the ratio of the state with the highest mortality 
rate of multiple myeloma (District of Columbia, 4.5 deaths per 
100,000/year) and the state with the lowest mortality rate (Hawaii, 
1.7 deaths per 100,000/year) to be 2.6.  If such a difference is seen 
in very large control populations, I would not be surprised to see odds 
ratios of 4 when comparing smaller control groups.

Wouldn't it be nice to have a normal range of background cancers?  I 
don't think I've ever seen dose effect curves where the range of the effect at 
background is given.  Given the variability of the human corporeal 
health, this information would be handy to assess whether there is 
truely cause for concern.  For instance, I've seen the background 
cancer mortality as 16%., 17%, 20%, and 25%, but I've never seen it 
as X% +/- SD.  The normal range is used in many parts of medicine.  
When you get a CBC, your WBC count is given as an acceptible range to 
account for the huge width of acceptible values.  It seems to me that 
this would be a finer measurement of an area of concern than 
comparing two populations, especially when you can compare two control 
populations and yield elevated odds ratios.  [P.S.  I know that the CI is 
supposed to allude to variability of background,  but that's not very 
clear, is it?]

Tom Mohaupt
Wright State University
tmohaupt@wright.edu
Tom Mohaupt, MS, CHP
Wright State University
Voice:  (937) 775-2169
tmohaupt@wright.edu