Another horrific failure of reviewers. 6 subjects!? Did anyone actually check the study they registered? https://clinicaltrials.gov/ct2/show/NCT02790541 Their actual clinical trial had 62 subjects and it had a control arm. The paper has only 6 subjects and no controls. If I run an experiment with 60 people, I can find evidence for anything (precognition, telekinesis, etc.) if I'm allowed to cherry pick 6 data points…
it sounds like you are assuming they collected more data than they reported, is there evidence of that? They preregistered to collect 60, but they couldve simply been unable to collect that many (eg COVID). If it was preregistered perhaps the journal were obligated to publish the results no matter the outcome. Some journals do this ostensibly to combat publication bias. I didn’t read the whole thing: are the authors…
This is why we analyse on an intention-to-treat basis rather than who actually received the treatment. The publication makes no such attempt to explain this potential source of bias.