Isn't there a difference between "treatment effect" as in how does the intervention affect participants, and "treatment effect" as in how much additional benefit does the intervention bring compared to placebo?
Comments
Log in with your Bluesky account to leave a comment
If I have a cold, take homeopathy and my cold is cured afterwards, my improvement is not a treatment effect. It likely had nothing to do with taking homeopathy.
So without controlling for other, more plausible explanations, I do not think it's fair to speak of a 'treatment effect.'
I think it's fair to say that observed benefits were 'clinically important', and group difference were 'statistically significant' and effect size was moderate.
Agree it's debatable what is clinically significant.
But given that the authors had pre-specified a MCID of 10 points in the protocol, I do not think it's fair to simply ignore this and claim the effect of 9 points was clinically significant without any discussion or justification.
They write "self-reported physical function improved statistically and clinically significantly in the intervention group after 2 to 8 outpatient encounters" and I think that's correct. In the other two instances they add "compared to controls" which I agree can be challenged.
Comments
So without controlling for other, more plausible explanations, I do not think it's fair to speak of a 'treatment effect.'
But given that the authors had pre-specified a MCID of 10 points in the protocol, I do not think it's fair to simply ignore this and claim the effect of 9 points was clinically significant without any discussion or justification.