> You simply completely ignored my core statement.

Most of it was junk, or included stuff about TMS which has zero relevance to antidepressants. Here's the junk:

> The truth is we took someone from being passively suicidal to functioning normally, and we fail to look at the self-reported problems, we just report them. The self reported problems tend to change from "I don't care about anything" to "I'm unhappy at my job" or "I'm stressed at how much I have to do with work and my kids and home." These are actually major improvements, the patient has gone from being actually clinically depressed to significant improvement but continued unhappiness with life circumstances as opposed to unhappiness with life in general.

Sorry, but, no, going from passively suicidal to normal function (including not getting out bed) would move HAM-D and almost other other depression rating scales massively, so we would see this in studies, but we don't. This doesn't pass the sniff test.

Could antidepressants help with other symptoms that are milder than suicidality, but which are not measured? Plausibly, but why would I trust you, a faithful zealot (literally: you will "fight tooth and nail", and "antidepressants are amazing")? Where are the papers? You are too biased to take at word.

> We're engaged in a long term research study to help demonstrate the value of a tightly integrated therapy/psych team and more advanced treatments and when you get everyone in the room pulling in the same direction patient outcomes are amazing.

Great, when it is published, I look forward to reading it. Doctors have felt this way since the dawn of time though, and since you are a zealot, I don't care until I see the paper.

That being said, yes, there is broad evidence that combination treatments (e.g. antidepressants + psychotherapy) are generally better than either alone, and it is quite reasonable when there is proper, regular communication amongst those teams that it is a bit better still. There are studies looking at this, but again, you find average effect sizes that really fail to meet anything like a minimally important difference.

And again in these cases it remains unclear how much of the patients that do experience clinically meaningful change are changing because of the antidepressants (or how much of the meaningful difference can be attributed to the antidepressants). It's the same basic problem, the science here is really, really poor.

> a metastudy [which] is only useful for indicating future research avenues, not drawing clinical decisions

Nonsense, you'd better be thinking about and looking at these kinds of things when making clinical decisions.

> You linked 4 studies you don't even understand because you misrepresented the data they report. One of them actually PROVES my point

You've shown me nothing to indicate you have a better understanding, none of them prove anything you are saying. Best you can say is some interpretations of some studies looking at the distribution of the treatment effect might be consistent with more individuals receiving antidepressants having large positive effects, but, as I said, this is remains only weakly supported in general, i.e. it isn't a clear finding and just remains a "maybe".

The MID issue is huge, and that you work in this area on a research committee board and can't deal with this, and appeal to authority without seeming to understand the basic measurement issues at hand here, is deeply concerning. As a true believer, your beliefs about the efficacy of esketamine are also widely overstated, and, we can be sure, ignore the MID issue as usual.

Nice try though.

EDIT: Okay, since burnte is too lazy to do the work, I did some looking for proper MID-based responder analyses. There are a some, and one is with eskatimine [1], and finds:

> By Day 28, 86.5% of patients reached or exceeded the PHQ-9 [MID] in the esketamine/AD group compared to 70% in the placebo/AD group. The most appropriate [MID] for the MADRS was -10 points. By Day 28, 78.2% of patients reached or exceeded the MADRS [MID] in the esketamine/AD group compared to 65.0% in the placebo/AD group.

So this is real research! And maybe esketamine really is something new. But then, why is competent research nearly impossible to find and results always presented in a way that aren't clinically interpretable and/or prevent seeing the distributions in a way that could easily answer these questions? Hmmm, maybe because the placebo vs. treatment distributions look so similar, like this: https://www.nature.com/articles/s41398-022-01882-5/figures/1

[1] https://pubmed.ncbi.nlm.nih.gov/33261932/