OT - Medical journals & Big Pharma
Posted: Thu Sep 08, 2005 12:53 am
Another slant on how articles such as the Lancet piece come to be written:
Richard Smith, editor of the British Medical Journal (BMJ) for 25 years, wrote a piece this May on how pharmaceutical companies use medical journals to promote their drugs (http://medicine.plosjournals.org, 18 May 2005). They have refined the art of designing clinical trials and studies that provide the results they want -- and the journals love to print positive stories.
And how do the drug companies create trials that increase their chances of a favorable result? According to Smith, they:
1) Conduct a trial of their drug against a treatment known to be inferior;
2) Conduct a trial in which the competitor drug is given at too low a dose to be effective (or too high a dose, which makes it seem more toxic than the company's drug);
3) Conduct trials too small to show differences from the competition;
4) Use multiple endpoints in trials (or do multicenter trials, or conduct subgroup analyses) -- and select only those that give favorable results for publication;
5) Structure the results so they're more likely to impress (e.g., a reduction in relative risk, not absolute risk).
So be aware of these deck-stackers when next you read a medical journal article. Then go out & see "The Constant Gardener."
Peace,
Cinnabar
[Non-text portions of this message have been removed]
Richard Smith, editor of the British Medical Journal (BMJ) for 25 years, wrote a piece this May on how pharmaceutical companies use medical journals to promote their drugs (http://medicine.plosjournals.org, 18 May 2005). They have refined the art of designing clinical trials and studies that provide the results they want -- and the journals love to print positive stories.
And how do the drug companies create trials that increase their chances of a favorable result? According to Smith, they:
1) Conduct a trial of their drug against a treatment known to be inferior;
2) Conduct a trial in which the competitor drug is given at too low a dose to be effective (or too high a dose, which makes it seem more toxic than the company's drug);
3) Conduct trials too small to show differences from the competition;
4) Use multiple endpoints in trials (or do multicenter trials, or conduct subgroup analyses) -- and select only those that give favorable results for publication;
5) Structure the results so they're more likely to impress (e.g., a reduction in relative risk, not absolute risk).
So be aware of these deck-stackers when next you read a medical journal article. Then go out & see "The Constant Gardener."
Peace,
Cinnabar
[Non-text portions of this message have been removed]