Back in 2009, the Journal of the American Podiatric Medical Association (JAPMA) published this impressively titled study: Extracorporeal Shockwave Therapy for Interdigital Neuroma: A Randomized, Placebo-Controlled, Double-Blind Trial that found:
The treatment group showed a significant difference before and after extracorporeal shockwave therapy (P < .0001). The sham group did not have a significant difference after 12 weeks (P = .1218).
and as such concluded
Conclusions: Extracorporeal shockwave therapy is a possible alternative to surgical excision for Morton’s neuroma
Given that impressively sounding title and the p values you could not help but be impressed, except that the way the results are stated in that quote above should have let off a huge alarm bell that the JAPMA editorial and peer review process had failed. The whole point of having a control group is that you do a statistical test comparing the outcomes between the intervention and control groups (ie a between groups analysis). What this study did was a within groups analysis which is not how you analyse a trial comparing two (or more) groups.
The study actually posted all the raw data in table, so I reanalyzed their data using the right analysis and published that as a letter to the editor of JAPMA. Others (here and here) also expressed the concerns about how the study was analysed. It turns out that the authors own data actually showed that shock wave therapy did not work for mortons neuroma!
JAPMA did not do the right ethical thing and retract the publication. They should have. They did not even link the abstract or paper on the website to the letters-to-the editor pointing out about the wrong analysis. They should have. If you read that paper and do not know about the issues of between groups vs within groups analysis, then without going quite a way out of your way to find the letters-to-the editor, you would believe that shock wave works for mortons neuroma. That is a problem.
Now fast forward to 2016 and this paper on the same topic, but different authors, appears in JAPMA: Extracorporeal Shockwave Therapy in Patients with Morton’s Neuroma: A Randomized, Placebo-Controlled Trial, again with a pretty impressive sounding title, that concluded:
These results suggest that ESWT may reduce pain in patients with Morton’s neuroma.
Guess what? JAPMA’s peer review and editorial processes let us down again
Guess what? The authors did a within groups analysis rather than the correct between groups analysis (the exact same wrong analysis as the 2009 paper above)
Guess what? They did not actually publish the mean and standard deviations of the outcome values in the paper (another failure of the editorial and peer review process to not get the authors to include that), but if you look at the outcomes in figure 3 (if you have access to the paper), notice that if you compare them between the two groups, they are about the same (you can only guess what the values are from the graph as they did not publish them).
Guess what? This study’s own data actually showed that shock wave therapy for mortons neuroma does not work, which is the opposite of what they conclude.
You think they would have learnt from the first one above, wouldn’t you?, but, no they repeated the same error again in the second one.
Am I going to write to the Journal another letter to the editor? No, can’t be bothered. The first example above shows that it does not work.
JAPMA should have retracted the first paper and now it should retract this second paper.
This is a serious ethical issue and JAPMA editorial staff would do no worse than appraise themselves on publication ethics. Anyone not familiar with the between groups vs within groups analysis will mistakenly believe the conclusions of these studies, when both studies show that shock wave therapy does not work for neuroma’s. What is the ethics of using shock wave therapy on neuromas based on these two papers that should not have made it through the editorial and peer review processes?
I have made no secret of my views on the editorial and peer review processes at JAPMA and have previously called them on another publication on my other blog on running. Guess what: one of the issues was the same as above, ie within group vs between group analysis. JAPMA’s Impact Factor has been languishing around 0.5 for a while now while all other foot and podiatry related journals have gone up. There is a reason for that.