This section looks at
the difficulties of bringing the SSRI issues to light. Raising the
issues with the editors of the Journal of Psychopharmacology, the
British Medical Journal and the Lancet has produced an interesting
correspondence. The relevant articles, reviews and correspondence
are posted here. No criticism of Richard Smith or Richard Horton
should be inferred as they are among the ‘good guys’, although clearly
David Nutt sets the standard for all other editors to aspire to.
There is relevant background material in chapter 5 and chapter 8
in Let Them Eat Prozac.
“I think a version of
your paper could well be suitable for publication in the BMJ if
you can shorten it to not more than 2000 words… I think the Prozac
story is especially interesting, and it clearly would make sense
for something to be published in the BMJ when we have played such
a crucial part in the story. I remember clearly the meta-analysis
that we published, and I remember something about the debate around
the paper at the time. Some people said we shouldn’t publish the
paper because it would inevitably be biased, making the point, I
remember, that if the study had proved the link between Prozac and
suicide then they undoubtedly wouldn’t have sent the paper to us.
Others said that we couldn’t reject a paper simply because it came
from a pharmaceutical company and that we didn’t see many major
scientific problems with the study. In retrospect, there is clearly
a problem with the fact that the study suffers heavily from publication
bias[i], but I think
it’s true to say that we were all much less conscious of publication
bias in 1990 than we are in 1999”. – Richard Smith.
“I can understand that
you must be worrying that we keep rejecting your paper because we
are covering up a mistake. Perhaps unconsciously we are, but I
obviously don’t think that is the case. I think that we are rejecting
your papers because they are too long, too unfocussed and insufficiently
clear”. – Richard Smith.
“We were unforgivably
slow in dealing with your excellent paper… It was approved by our
reviewers.. no modifications were proposed. I am wondering whether
you would agree to our printing it as a guest editorial. I prefer
that papers which we are anxious to emphasize get this status.”
– Graham Dukes.
that none of the studies in the Beasley paper were designed to answer
the question, it is debatable whether publication bias has anything
to do with what happened to this article. The sheer embarrassment
of recognizing this may have played a part in Richard Smith’s inability
to accept any papers drawing attention to the issue.