Wednesday, May 18, 2005

Retired Doc's suggestion for medical curriculum,part 10, Beware of misleading bias in drug trials

Dr. Richard Smith, a editor of BMJ for 25 years,discusses some of the ways that drug companies get the results they want published in main line journals. Kevin,MD also has a recent post referencing Dr. Smith's insights.
Writing in PLoS Medicine (Smith, R (2005) PLoS Med 2 (5) : e 138)"Medical Journals are an extension of the marketing arm of pharmaceutical companies.", he gives handy hints for methods that drug companies can use to get the results desired from clinical trials.For example, trial your drugs against too low a dose of the competitor drug or conduct a trial against a treatment known to be inferior, or use multiple end points in the trial and select for publication those that give favorable result. Note he is not talking about making up data or that the trials are not conducted in a technically well done manner, but rather that bias is introduced in subtle ways. He also describes the practice of publishing the positive results more than once and to combine the results from various centers in multiple combinations. There are a number of ways that physicians can be- and apparently have been-mislead about the virtues of various medications.
As medical students learn the catechism of evidence based medicine they need to learn to be critical of trials even in the best of journals.There is good reason to believe that not only have readers of journals been taken in at times but so have the editors.Smith says editors are beginning to catch on to at least some of the manipulative techniques that have been used and may take appropriate defensive action. He admits it took him over 20 years to realize what was happening.
Medical students are appropriately taught that RCTs are the highest standard of proof in the on going search for medical truth. However,at least some the RCTs on which some standards of care are built may have been spun by the drug companies to give favorable press to this or that particular medication and present some thing far less that the whole truth. He quotes data indicating that at least 2/3 of trials published in JAMA, the Annals of Internal Medicine,NEJM and Lancet were funded by the drug industry. So it could be that these practices could be rather wide spread. Smith's article should be mandatory reading in the EBM course. It might serve as a partial antidote to the over exuberant enthusiasm for EBM precepts that some impressionable students might develop. Look at the nature of the evidence.
So what can the overwhelmed medical student do? Right now, the best I can suggest is to use the table of techniques in Smith's article but realize this is not likely to be exhaustive. The EMB gurus from McMaster might add a section in the " How to find current best evidence" chapter in their book, " Evidence Based Medicine " by Sackett et al on" How to detect bias in a drug trial publications".
Smith suggests that journals not publish trials and the protocols and results be available on the web . This, he says, is radical and not likely to happen.But with the issue more on the table now, editors will be more savvy and skeptical as, hopefully, will be the readers.
Between Ghost writing, the manipulative arsenal of big pharma described by Smith, the alleged marketing and other corporate techniques related to the cox-2 drugs and the recent " it's safe, wait, no it's not" antics of the FDA ( which in part is allegedly related to the aforementioned manipulations) it should be no wonder that patients may have less faith in what the doctors are prescribing and so should the docs.

2 comments:

marianne said...

Dear Dr. Gault,
My name is Marianne Szegedy-Maszak and I am working on a piece for the LA Times on doctor's blogs and would be very grateful if I could interview you. Could you please send me an email to szegedym@aol.com so we could set up a time for a telephone interview?

thanks so much,
Marianne

Anonymous said...

p.s. if there is anyone else out there who would like to talk with me about doctor's blogs, please get in touch with me at szegedym@aol.com.

thanks
marianne