In the most recent issue of the Journal of the American Podiatric Medical Association is this paper: Robert Fridman, Jarrett D. Cain, and Lowell Weil, Jr: Extracorporeal Shockwave Therapy for Interdigital Neuroma: A Randomized, Placebo-Controlled, Double-Blind Trial. J Am Podiatr Med Assoc 2009 99: 191-193. This was a randomised controlled trial comparing shockwave therapy for mortons neuroma to a placebo.
The purpose of a RCT is to determine how much more effective an intervention is compared to a placebo. To do this the correct way to analyze a RCT is to compare the outcomes BETWEEN the intervention and placebo groups (I could cite some references here, but every book on clinical trials says that, so take your pick).
The authors in this study did a WITHIN groups analysis which is the wrong way! They found a statistically significant difference between baseline and outcome in the shockwave group and no difference in the placebo group. Doing a WITHIN groups comparison, you have no way of knowing how much of the change in the shockwave group was due to the placebo effect which is why you do a BETWEEN groups comparison. On the basis of the analysis that the authors did, they concluded that: “Extracorporeal shockwave therapy is a possible alternative to surgical excision for Morton’s neuroma” and made recommendations for its use.
In the paper the authors tabulated the individual results for each participant. I put those numbers into a stats program and did a BETWEEN groups comparison and got a p value for the difference of 0.27 – not even close to being statistically significantly different! The authors actually showed that shockwave therapy for neuromas was no better than a placebo! – the opposite of what they claimed!
Not to mention that this research was also very underpowered and there were dropouts in the placebo group that should have been included in the analysis (look up ‘intention to treat’).
I seriously question how this publication made it past the journal’s review process. Comments here: Shockwave therapy for Morton's neuroma.
Back to home page
