Latent cancer fatalities and premature deaths

["Dukelow, James S Jr" <jim.dukelow@pnl.gov>]

Ruth Weiner wrote
------------------------------

Date: Tue, 28 Nov 2000 19:01:43 -0700
From: "ruth_weiner" <ruth_weiner@email.msn.com>
To: <radsafe@romulus.ehs.uiuc.edu>
Subject: Re: Deaths from fossil fuel burning air pollution

Jerry makes some very good points.  We continue to refer to the product of
population radiation dose (rem) and the conversion factor 5E-4 (or
person-Sv*5E-6) as "latent cancer fatalities" meaning, I suppose, that that
many cancers, which might prove to be fatal, would occur in the exposed
population.  I don't like LCF particularly, but it is certainly better than
"deaths."  Years of life (potentially) lost is a much better measure and is
moreover comparable to other risks that shorten life.  Everyone dies, after
all.  The probability of dying is=1.  "Premature" deaths is vague because
life expectancy is so variable.

	<snip>

A couple of comments:

The conversion factors for LCF are calculated so that the conversion is actually
to fatalities.  Thus, the given dose is assumed -- by whatever method is used to
derive the conversion factor -- to produce a larger number of cancer cases, some
fraction of which will turn out to be fatal.  For some cancers, such as thyroid
cancer, that fraction may be small -- on the order of  0.1 or 0.05.

For most epi studies looking at mortality, premature deaths is what is actually
being measured, in that the number of deaths in the exposed group over a defined
time period is compared with the number of deaths in the control group.  There
is a trap here, though, one into which the inimitable Steven Wing has fallen in
some of his work.

In epi studies of DOE worker populations, the exposed group typically shows a
healthy worker effect, with fewer overall deaths during the follow-up period
than the control group.  Inevitably, though, if the same population is
reanalyzed for longer and longer follow-up periods, as has been done with
several DOE worker populations, the healthy worker effect will begin to
disappear, since the human mortality rate is still one per person.  As the
overall mortality of the exposed group approaches that of the control group, the
statistical fluctuations in various subgroups will be more likely to produce
"significant" increases in particular diseases or other causes of death than
they were with a shorter follow-up period.

Adopting Ruth's suggestion re Years of Life Lost, would eliminate this
particular problem.

Best regards.

Jim Dukelow
Pacific Northwest National Laboratory
Richland, WA
jim.dukelow@pnl.gov

These comments are mine and have not been reviewed and/or approved by my
management or by the U.S. Department of Energy.
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html