Read on Twitter
https://journals.sagepub.com/doi/10.1177/0269881120922950#articleShareContainer Thread. Ppl critique efficacy of antidepressants, which is fair. Here @f_hieronymus and I examine this. Most MAs of acute tx hve effect size=0.3 for ADs in deprn. Some say this is not clinically sig, some say it shows they “work”. It is more complex imo
What is clinically sig change? 2 authors tk data frm paper on pharma trials of mirtazapine and tk a pt from graph of general clinician rated change and deprn rated by Hamilton Scale (HAM-D.) They said 7 pt change HAMD corresponded 2 minimally sig clinician rated change.
This led sme 2 state effect size 0.875 wld be min clinical sig. Which is higher than most tx in medicine. Odd. Taking 0.875 as min. change, almost all pts wld nd to b in remission. How does that make sense? Examine paper this arbitrary 7 pt change was (cherry) picked from
- orig authors talked of % change, as diff ppl have diff baseline scores. Also, mirtazapine has fx on HAMD prob nt related to depression-increased change in HAMD prob nt related to mood and diff to SSRIs (sedation, less GI side fx).
Wt ppl forget is HAMD is v old scale,measuring nt just depression but lots of somatic sx eg GI side fx, insomnia which can b caused by ADs acutely. When looking at rating of mood, SSRIs separate well frm placebo https://www.nature.com/articles/mp201553
So-how do we go forward? Nd pt rated outcomes, measures function and btr clinical scales. Definitely. But lessons from this-Journal Editors, peer reviewers, policy makers+researchers- interpret scales as just that-and b careful wt info ppl give, esp when it sounds dodgy Please.
