Re: cancer clusters

["Tony Harrison" <Tony.Harrison@state.co.us>]

Epidemiology works best with exposures that have a high (> 2-3) relative

risk for a given outcome, like E. coli (currently big news here in

Colorado), HIV for AIDS and tobacco smoke for lung cancer.  Most of

those have been identified, or are so rare they aren't much of an issue

for the public.  Classic epi techniques are less useful for exposure

with low relative risks (between 1 and 2) or where exposure is very

rare.  Good epidemiologists know this and focus where they can do good. 

Bad epidemiologists know this and use it to milk the system (this

generally only works for a little while) or for political reasons.



Cancer (and other disease) clusters occur randomly, and epidemiologists

know this as well.  But until a cluster is thoroughly investigated, how

confident should we be that we are witnessing a collection of random

events, and not a true outbreak of preventable disease?



Tony Harrison, MSPH

Colorado Dept. of Public Health & Environment

Laboratory and Radiation Services Division

(303)692-3046

tony.harrison@state.co.us





>>> <RuthWeiner@AOL.COM> 07/18/02 08:10AM >>>

Three comments on clusters:



Disclaimer:  I am not an epidemiiologist.



1.  Several years ago one of my M. S. students investigated "leukemia 

clusters" near industrial ponds in northwest Washington.  She could not

show 

any association except geographical proximity: no consistent dose

response, 

not even consistent exposure to the water or consistency in the

duration of 

the geographic proximity.



2.  Two years ago, the New Yorker published an article on disease

clusters by 

Malcolm Gladwell.  As I recall, the author stated that geographical

clusters 

had not proved out in any case that he looked at.  Disease clusters

seemed to 

be much more the result of common behavior  than of geographical

location.



3.  I remember the study of spontaneous abortions in Alsea. Oregon,

where it 

was claimed that the unusually high incidence was related to effluents

from 

the local tree-spraying operations (spraying with pesticides).  It

turned out 

that there was not an unusually high incidence of spontaneous

abortion.



The epidemiology of AIDS, which did ultimately lead to elucidation of

the 

cause of the disease, strikes me as a classic example of good

epidemiology, 

particularly because it began with observation of a serious adverse

effect -- 

a complex of fatal diseases -- and not with any preconceived notion

about 

possible causative factors.











Ruth Weiner, Ph. D.

ruthweiner@aol.com

************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.

You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/