Revisiting: Journal club: SURMOUNT-OSA and a hard endpoint in a soft field — setting out what I have, and where I think it stops being reliable.
Collecting what is known about SURMOUNT-OSA in one place, because it is currently spread across a category, two tag pages and a thread that is hard to find.
This is a summary rather than new work, and I have attributed each part to where I found it.
Taking post #4 at face value and following it one step further.
The baseline characteristics table is the single most useful page for the questions that get asked here, because it tells you who the result applies to.
Noting that the question and the thing people usually mean by it are different.
The claim about SURMOUNT-OSA upthread is stronger than its source supports. I have read the source. The source says "associated with" and the post says "causes".
I would be cautious about generalising from the SURMOUNT-OSA example above. It is a good example. It is one example.
Adding the measurement that post #8 says would settle it.
SURMOUNT-OSA is a good example of a question where the honest answer is boring and the interesting answers are unsupported. I would go with boring.
Agreed on all of that, and I have nothing to add to it.
Picking up post #12: that is the part I would want checked first.
Whatever the answer on SURMOUNT-OSA turns out to be, the method for getting there is the same: state the assumption, do the arithmetic in public, invite the correction.
Reading order that works for these sessions: registry entry, methods, baseline table, primary result, then abstract last. Reading the abstract first anchors everything that follows.
Not the whole picture, but the part of it I can speak to.
Reading this SURMOUNT-OSA thread as someone who came in with a fixed view: the third and seventh replies moved me and the confident ones did not.
The figures usually contain the finding and the text usually contains the interpretation. Separating them for the first twenty minutes is a discipline worth keeping.
I have separated what I observed from what I concluded, which does not always happen.
The practical version of SURMOUNT-OSA is three sentences long. The rigorous version is three pages and reaches the same conclusion with the conditions attached.
The useful distinction on SURMOUNT-OSA is between what was measured and what was inferred from it. Both end up in the same sentence and only one of them has error bars.
Sessions on negative or null results are consistently the most instructive and the hardest to get anyone to attend.
The disagreement above is smaller than it looks once the terms are fixed.
I read post #19 twice before replying, because I had assumed the opposite.
I changed my mind about SURMOUNT-OSA after someone here asked me for the source and I could not produce one. That is worth saying out loud because it is the ordinary way it happens.
Post #23 answers the question as asked. The question underneath it is different.
The baseline characteristics table is the single most useful page for the questions that get asked here, because it tells you who the result applies to.
Caveat: everything above assumes the paperwork is what it says it is.
I disagree with the framing of SURMOUNT-OSA above, and I think it is a substantive disagreement rather than a terminological one. Setting out why, so it can be checked.
The reasoning depends on an assumption that is doing a lot of work and is never stated. If the assumption holds, the conclusion follows. I do not think it holds generally.
That is the distinction I keep failing to hold on to. Written down now.
Everything in post #23 holds. The case it does not cover is the one I have.
The most common failure mode is spending fifty minutes on the effect size and ten on the population. Reversing that ratio would improve most sessions.
Nothing above should be read as advice about what anyone else should do.
I keep a log for SURMOUNT-OSA specifically because my memory of it turned out to be systematically wrong in one direction. Six weeks of notes cost nothing and settled it.
Building on post #27 rather than restating it.
Where the group cannot agree, record the disagreement rather than resolving it by seniority. The recorded disagreement is more honest and more useful later.
Genuinely open to being wrong about this one.