[e-drug] Wording of Papers, Use of Statistics

e-drug
1

E-DRUG: Wording of Papers, Use of Statistics
--------------------------------------------

In order to perform evidence-based decisions-as we call them-Professionals often rely on papers published in international Journals.

But sometimes the wording of the papers, the statistical arguments and, interpretation are often biased. And it can take academicians skilled at critical analysis to unearth bias. Often, bias is detected after performing a series of errors following use of biased evidence.

In this vein, I wish to raise two well-meaning ideas. And these also have their limitations, but it could be worth trying them.

One, evidence alone, without professional, technical logic is more dangerous than the lack of it.

Two, could there be some kind of well-monitored continuing education in critical, statistical analyses for editors of major Journals as well as peer reviewers? And no insult intended.

regards

George Kibumba, MPS (Uganda)
Msc. Student (Sept 2004-Sept 2005),
Pharmaceutical Services and Medicines Control,
University of Bradford,
Wardley House, Flat A3, Room 1,
Bradford, BD5 0AE, United Kingdom.
Personal e-mail: kibumba@yahoo.com
Student e-mail: G.Kibumba@bradford.ac.uk

2

E-DRUG: Wording of Papers, Use of Statistics (2)
------------------------------------------------

Dear E-druggers, dear George,

It has often been said - and repeated - that there are some things peer-reviewed scientific journals cannot do:

The bulk of the registration data will NOT be published because either the material is not scientifically interesting enough (and one of the main jobs of a chief editor is to make his/her journal readable, otherwise it will not be sold) or because the manufacturer has his reasons to reveal the data only to the registration authorities. Hopefully such data will in the future be stored in appropriate online databases, accessible to the public.

There still are three different categories of scientific publication media:

(1) the top-notch peer-reviewed journals where papers are sent out to at least two known experts in the field, who are supposed to have the knowledge George found wanting. The assessments as well as the original paper are again reviewed by the chief editor, who will return the paper to the author with all the comments and the request to review or rewrite the paper. A second full assessment may follow.

Then (2) there is a large group of "second rate" journals without proper reviewing system; many of those are directly or indirectly sponsored, otherwise they might not survive. Often papers are written by ghostwriters - who might be quite good, but are paid by industry. They are not really willing to be "taught", but the majority will not be quoted by Current Contents or Medline.

And finally (3) there is the well-known problem of the "scientific supplements" to peer-reviewed journals, often containing reports of sponsored meetings and congresses. These then are quoted by Medline.

So the readers should in the first place be taught how to select and read journals, not the other way round.

I can wholeheartedly recommend reading Marcia Angell's recent book, which should be available from your library.

Best wishes,

Leo Offerhaus
Former Managing Editor of a well known peer-reviewed scientific journal (clinical pharmacology).

3

E-DRUG: Wording of Papers, Use of Statistics (3)
------------------------------------------------

Dear George,

A comment on your view on <evidence alone, without
professional, technical logic is more dangerous than
the lack of it>

I dont think that evidence alone or the lack of it is
anyway dangerous rather what we do or not do with it,
or do in the absence of it.

History has shown us that having the necessary skill,
technology and knowledge and NOT using the evidence
can indeed be harmful.

Some example to illustrate this point:

1. Consider the number of patient with heart attacks
in the last two decades who died prematurely because
they had never been offered antiplatelet therapy with
low dose aspirin, though strong evidence existed for
its prophylactic use by 1980.

2. or the number of women with breast cancer, being
maimed and suffered greatly due to drastic breast
removal surgery that did not prolong their life
(though evidence showed many years ago that this type
of surgery is not saving extra lives, only add more
suffering).

Also think of the many thousands of patients with
HIV/TB in Africa who has not been receiving
cotrimoxazole prophylaxis for the last 5 years,
although we had the evidence from the Cote d'Ivoire
study (Anglaret, 1999), that it can significantly
reduce death and illness, even in a developing
country settings (not always optimal professional
skills, technical capacity or logistics)

For many of these patient NOT using the evidence even
when only imperfect skills and logistics are available
has been far more dangerous than evidence alone,
without professional, technical logic .

Klara

Dr Klara Tisocki
B. Pharm., M.Sc., Ph.D.
Phone: + 965 7975493
Fax: +1 775 4169388
Kuwait City
ktisocki@yahoo.ie