Intended for healthcare professionals

Research

Cochrane reviews compared with industry supported meta-analyses and other meta-analyses of the same drugs: systematic review

BMJ 2006; 333 doi: https://doi.org/10.1136/bmj.38973.444699.0B (Published 12 October 2006) Cite this as: BMJ 2006;333:782

Doctored meta-analyses are not the only way drug companies manipulate physician drug use.

Dear Editor,

The study by Jorgensen et al(1) is an eye opener to the data
manipulations used by the drug companies to sell their products. The value
of meta-analysis is severely hampered if data from all trials available
are not used, or if data not available to the public is used. This has
happened in most of the cases mentioned in the study concerned(1).
However, we hold that this is not the only way drug companies try to
manipulate physician-prescribing to their benefit.

Selective highlighting of the trials that show their product in good
light is a very common method known to doctors that have attended drug
company sponsored events. It requires a very astute clinician who is up-to
-date with his reading to fail to be impressed by the data presented.

Publication bias is another important pitfall. Data that is available
to the public and to the physicians is only those data the authors choose
to send to the journals, the drug company chooses to fund or the journals
choose to publish. Unfortunately, negative studies are not considered an
advance by many researchers and much of the negative data are never
published. This automatically skews all meta-analysis in favour of a new
therapy.

It is even more a cause for concern when authors choose to ignore or
manipulate data that shows a new drug in bad light as happened with the
Rofecoxib trial(2,3). The financial vested interests of the authors in the
concerned drug company has been reported specifically(4).

It is a great cause for concern that most of the new drugs reported
in Jorgensen's study(1) that have shown equivocal data in Cochrane reviews
are now quite commonly, if not exclusively, used for the disease in
question.

We suggest a few methods to avoid future bias in favour of new
treatments. A central website which is open access could be maintained
where authors are free to report any negative trials. Until a climate
develops where negative trials are encouraged and reported without any
publication bias, this seems to be the only option.

Physicians have an obligation to know in detail about the trial
before they change their prescribing habits. To this end, pharmaceutical
companies should be encouraged to provide full reprints of the trial
rather than just highlight the few graphs or tables that show their
product in good stead.

Reference:

1. JØrgensen AW, Hilden J, GØtzsche PC. Cochrane reviews compared
with industry supported meta-analyses and other meta-analyses of the same
drugs: systematic reviews. BMJ 2006:332:782-6.

2. Curfman GD, Morrissey S, Drazen JM. Expression of Concern:
Bombardier et al., "Comparison of Upper Gastrointestinal Toxicity of
Rofecoxib and Naproxen in Patients with Rheumatoid Arthritis," N Engl J
Med 2000;343:1520-8. N Engl J Med; 353: 2813-4.

3. Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis
B, Day R et al. Comparison of Upper Gastrointestinal Toxicity of Rofecoxib
and Naproxen in Patients with Rheumatoid Arthritis. N Engl J Med
2000;343(21):1520-8.

4. Supplement to: Bombardier C et al. Comparison of Upper
Gastrointestinal Toxicity of Rofecoxib and Naproxen in Patients with
Rheumatoid Arthritis. N Engl J Med 2000;343(21):1520-8. available online
http://content.nejm.org/cgi/content/full/343/21/1520/DC1

Competing interests:
None declared

Competing interests: No competing interests

16 October 2006
Rekha N Pillai
Foundation year 2 doctor
Manjith Narayanan, SpR, Paediatrics, Leicester.
Diana, Princess of Wales Hospital, Grimsby DN33 2BA