September 26th, 2003
Boston Cure Project
"Effect of awareness of a randomized controlled trial on use of experimental therapy"
Clark WF, Garg AX, Blake PG, Rock GA, Heidenheim AP, Sackett DL.
JAMA. 2003 Sep 10;290(10):1351-5
In a study published in JAMA, a team of Canadian researchers investigated whether randomized clinical trials of experimental treatments affect the overall demand for those treatments. They found that when a trial for an available treatment is initiated for a new indication (disorder), overall use of that treatment increases among people with that disorder. In other words, they found that rather than waiting for the results of the trial, physicians and/or their patients become more motivated to use the treatment in advance of the results.
The researchers looked at the usage of apheresis (blood filtering) in MS and two other diseases at a group of Canadian apheresis centers. A randomized clinical trial of apheresis was performed for each disorder to follow up on anecdotal evidence and/or clarify the results of past flawed studies. In the case of MS, over twice as many people on average underwent apheresis during the years of the trial (1986-1988) compared with the years before the trial. The case was similar for the two other diseases.
The study discusses a number of potential factors that may have led to this phenomenon, but concludes that it is most likely due to physicians becoming more confident in the efficacy of a treatment just because a trial is being conducted. Unfortunately this confidence is not well-founded -- for instance, the MS trial found apheresis to be ineffective and the usage levels dropped again after the results were announced.
I think the moral of this story is that until a randomized clinical
trial is actually completed and the results are published, no one should
draw any conclusions about the effectiveness of a treatment. And anyone
who doesn't know the benefits of a treatment should be very certain of
the risks before giving it a try.
Copyright © 2003, Boston Cure Project