I have heard it said that some epidemiological journals no longer accept
traditional covariate based adjustments for confounders, requiring authors
instead to carry out propensity score based stratification. If so this is,
in my opinion, an inappropriate policy based on a misunderstanding.
However, perhaps I am misinformed. Does anybody know of such a policy by
any journal or has anybody been required to amend a paper accordingly or
had it rejected on such grounds?
I would be interetsed to collect any experiences.
Thanks in advance
Stephen Senn
University of Glasgow
|