171 Ashley Ave.
Charleston, SC 29425
843-792-1414
800-424-MUSC
|
 |
March 2008
Evidence-Based Tip
 |
Laura Cousineau, MLS
MUSC Library
Dept. of Pediatrics EBM Faculty |
Making sure you have ALL of the evidence
In addition to rounding with pediatrics, I also round at the Institute of Psychiatry once a week. If ever there was a place where "practicing medicine" refers to medicine as in pharmaceuticals, it is in psychiatry. It is not unusual for our patients to be on six or more medications.
With appropriate drug selection being so important, I read with interest the January 17 article in NEJM on the selective publication of antidepressant trials*. The first line of this article reads "Evidence-based medicine is valuable to the extent that the evidence base in complete and unbiased."
In teaching evidence-based medicine, we emphasize the difference between a review article and a systematic review. A review of the literature is subject to selection bias, and often chooses only articles selected from a single search of the literature. A systematic review must be excruciatingly thorough, using complex search strategies over several databases, and hand searches of appropriate journals. A systematic review should also look for grey literature; that is, conference reports, unpublished studies, theses and other such materials that do not turn up in literature searches alone. However, some reviews that are called systematic reviews have not been conducted with the thoroughness required.
As this NEJM article points out, that can make a difference in properly estimating the benefit or harm of a given therapy. A full 31 percent of the studies done on the 12 antidepressant agents tracked had not been published. There are many reasons why a study might not be published: a rejected manuscript, failure to submit, blocked by the sponsor, etc. What was startling for the investigation of these 12 drugs was that the non-published studies were overwhelmingly negative. This falsely increased the published benefit for the drugs by an average of 32 percent, and as high as 69 percent in one.
It is an important lesson for both those of us that teach and those of us who seek to practice evidence-based medicine. It helps us remember that our evidence pyramid, that hierarchy of studies with systematic reviews at the top, is only as sturdy as the quality of the studies in its levels. It reminds us of the need to understand the mechanics of studies, and to critically read the methodology sections of articles. For those of us who publish systematic reviews, it is also a reminder to pick up that phone, send those e-mails, search the websites of drug manufacturers and the FDA to find those unpublished studies. And you might want to enlist the help of your local medical librarian as well.
* Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. New England Journal of Medicine (Jan 17 2008) v358 n3 p252-60.
|
|
|
Print Version