View Full Version : What criteria should we use for interpreting epidemiology?

02-19-2008, 08:35 AM
Dear all,

Lately in my professional reading I have noticed that the standards of
evidence varies greatly between different fields. For instance, the
mode of establishing causality used in science (which was devised by
Jakob Henle and Robert Koch for research in bacteriology) requires that
in every single instance the effect follows the cause. By contrast
epidemiology uses a system of causal inference that is based on the
philosophies of David Hume and John Stuart Mill (Morabia, 2005). The
question of identifying causes in epidemiology has always been a subject
of controversy. For instance, the controversy over the
interpretation of the statistical relationship between smoking and lung
cancer caused a landmark debate in that field in the second half of the
twentieth century. It led Bradford Hill to formulate the "pragmatics" of
risk factor epidemiology in 1965 (Berlivet, 2005). His model for
establishing causation ("The Hill causation model") is well known in the
public health field and widely used. For example Hill's nine proposed
"criteria" for determining causation were used was used to classify
Chrysotile asbestos as a cause of mesothelioma (Lemen, 2004); it met all
nine of them. However, "multiple causation" is the canon of contemporary
epidemiology and its "web of causation" is widely accepted though it is
a very poorly elaborated model (Krieger, 1994). Sadly public heath
debates these days are often based on evidence that use it in
questionable ways and this topic clearly deserves further debate. Has
anyone found a good review of criteria currently in use for interpreting


David McFarlane MAppSc (Ergonomics)
Ergonomist, WorkCover NSW


Any recommendation concerning the use or representation of a particular
brand of product in this document or any mention of them whatsoever
(whether this appears in the text, illustrations, photographs or in any
other form) is not to be taken to imply that WorkCover NSW approves or
endorses the product or the brand.

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

This message, including any attached files, is intended solely for the addressee named and may contain confidential
information. If you are not the intended recipient, please delete it and notify the sender. Any views expressed in this
message are those of the individual sender and are not necessarily the views of WorkCover NSW. Please consider the environment
before printing this E-mail.

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