c/trialwatch · created 4 Nov 2023
STEP, SURMOUNT, SURPASS, SELECT, FLOW, TRIUMPH — read the appendix.
FLOW (Perkovic et al., N Engl J Med , 2024) reported a hazard ratio of 0.76 for the composite kidney endpoint in people with type 2 diabetes and chronic kidney disease.…
The title is the argument: FLOW got published and nobody in here is talking about eGFR slope. Here is the rest of it. Intention-to-treat analyses everybody randomised…
SELECT: MACE HR 0.80, and the absolute numbers people skip — a position I have arrived at slowly and would like tested. Compared myself to a trial mean for about six…
relative risk reduction sells papers, absolute risk reduction makes decisions. It is the sort of thing everyone half-believes and nobody writes down. The…
Vendor-level purity averages over time, built from independent results rather than vendor claims. Transparent methodology, open dataset.
Browse resultsSURMOUNT-4 withdrawal arm, read the appendix not the abstract. Change my mind, genuinely — I have no stake in being right about this. Started keeping the trial…
Posting this as a discussion rather than a claim: STEP 1 vs SURMOUNT-1 side by side, with the caveats. Compared myself to a trial mean for about six months before…
new phase 3 numbers, no publication yet, calm down. It is the sort of thing everyone half-believes and nobody writes down. A confidence interval is the range of effects…
the placebo arm lost weight too. always ask by how much — a position I have arrived at slowly and would like tested. Why comparing across trials almost never works,…
intent-to-treat vs per-protocol explains half the arguments on this site. Change my mind, genuinely — I have no stake in being right about this. On means, which this…
Posting this as a discussion rather than a claim: REDEFINE 1 and what CagriSema actually adds over sema alone. Compared myself to a trial mean for about six months…
Something I keep coming back to: ATTAIN-1 — an oral pill with no food restrictions is a bigger deal than the % suggests. Quoted a figure here confidently, got asked…
number needed to treat, or i am not interested, and I am aware this is a minority view on this board. Read a press release and the publication three months apart. The…
Posting this as a discussion rather than a claim: watch where the curves separate, not just whether they do. Relative risk reduction without the baseline rate is…
proposal — a flair for FLOW posts. Nothing about this affects the ranking maths, before anyone asks. The registered protocol is public. Comparing the registered primary…
Something I keep coming back to: three years of SURMOUNT threads, summarised so you do not have to read them. Cardiovascular outcome trials are powered for events, not…
Reference standards, analytical reagents and system-suitability material. If you are running your own method, start with a known.
Reference standardsThinking out loud about this: SURPASS — 10 things I got wrong before I got it right. Started keeping the trial identifiers straight in a note file because I kept mixing…
Putting this to the board: the FLOW rule is doing its job and people should stop complaining. Why comparing across trials almost never works, with the specific failure…
STEP is not what most of this community thinks it is — with the reasoning, because a rule without a reason gets ignored. On means, which this board treats as targets…
Slightly embarrassed to be asking this, but: how do you actually verify SURPASS. Argued for a week about a result and then read the limitations section, which conceded…
how much of what we believe about endpoint actually comes from SURPASS threads. I am not trying to be the "source?" guy. I would just like a source. Read a press…
reading SURMOUNT threads from 2024 and half of it aged badly, and I am aware this is a minority view on this board. Open-label extensions lose their randomisation.…
Question in the title, detail here: can we stop arguing about TRIUMPH until somebody posts a number. Intention-to-treat analyses everybody randomised regardless of what…
unpopular opinion: most of what gets said here about hazard ratio is guesswork. Not a hot take, just something I have not seen said plainly here. A confidence interval…
The title is the whole question — does SELECT actually matter or is it forum lore at this point — but here is why I am asking. Went looking for the registered protocol…
TRIUMPH is doing more work than we give it credit for. It is the sort of thing everyone half-believes and nobody writes down. Spent an evening with the appendix tables…
Clinical trial reading group. New readouts, protocol amendments, endpoint definitions, dropout handling, and the difference between a press release and a publication. Absolute risk reduction and number-needed-to-treat are house dialect here.
Independent testing plus vendor-level trend data. Community-funded, unaffiliated.
peptidemeter.comSeveral compounds discussed on GLP Research Hub are sold for research use only and are not approved for human use anywhere. Nothing here is medical advice and none of it is written by your clinician. If a post reads like an instruction, treat it as a description of what one stranger did.
Reference