The Peptide CommonsEst. May 2024
Independent. We sell nothing and are affiliated with no manufacturer or pharmacy. Every moderation action is logged in public
Evidence · Trials · continued

[2026 update] Composite endpoints and the component doing the work posts 61–88

This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1.

ZO
z.okonkwoTL28 Apr 2026#61

That is a fair summary of where the discussion has got to.

21 likes 4mo
SS
system_suitabilityTL3Analytical chemist9 Apr 2026#62
s.beaulieu, post #29: The thing about Composite endpoints that took me longest to accept is that a plausible mechanism is not evidence of an effect. It is a reason to look, not a result. Go to post

Post #58 describes the usual case. This is about the unusual one.

I would rather this thread reach "we do not know" about Composite endpoints than reach a confident answer that nobody can support when asked.

0 likes in reply to #29 4mo
NI
n.ibarraTL29 Apr 2026#63

Risk of bias: structured appraisal of internal validity. Key things to assess: randomisation method (was it truly random or could someone predict the next assignment), concealment (could randomisation be subverted), blinding (who was blinded and why or why not), completeness of outcome reporting.

0 likes 4mo
KO
k.otieno_statsTL3Statistician10 Apr 2026 · edited#64

Composite endpoints has been discussed here with more heat than it deserves, mostly because two definitions have been in play the whole time.

5 likes 4mo
SB
s.balogunTL211 Apr 2026#65

What would change my mind on Composite endpoints is a second dataset collected by someone with no stake in the first. Until then I hold it loosely and I would rather say so than pretend to more.

28 likes 4mo
TP
tracked_parcelTL2Regular11 Apr 2026#66

On post #62 — agreed on the reasoning, with one qualification.

A treatment-policy estimand asks what happens to people assigned to a strategy, including those who abandon it. A hypothetical estimand asks what would have happened had everyone continued. Both are legitimate and they give different numbers.

0 likes 4mo
RV
r.vukovicTL212 Apr 2026#67
n.abernathy, post #39: Composite endpoints is a question about a distribution, not about a value, and treating it as a value is what produces the confident wrong answers. Go to post

Composite endpoints should be read component by component. A composite driven entirely by its softest component is a different finding from one where the components move together.

Someone should write this up properly, and it should probably not be me.

2 likes in reply to #39 4mo
UC
unit_conversionTL3Regular13 Apr 2026#68

Two claims get bundled together under Composite endpoints and they need separating. The descriptive one — this is what was observed — is usually well supported. The causal one — this is why — usually is not.

Almost every disagreement in threads like this one dissolves once you say which of the two you are making.

9 likes 3mo
HK
h.karlsenTL214 Apr 2026#69

The strongest argument against my own position on Composite endpoints, stated as well as I can state it, since nobody else has yet.

0 likes 3mo
CA
c.adebayoTL214 Apr 2026#70

Surrogate endpoints: an endpoint that is not the outcome that matters but is measured as a stand-in. HbA1c is a surrogate for long-term glucose control and the short-term complications it prevents. Weight loss is a surrogate for metabolic health and long-term outcomes. Surrogates are useful but not identical to the endpoint that matters.

0 likes 3mo
K
KStephanopoulosTL3Regular15 Apr 2026#71

Reading this Composite endpoints thread as someone who came in with a fixed view: the third and seventh replies moved me and the confident ones did not.

13 likes 3mo
HC
h.castellanosTL216 Apr 2026#72

The number people quote for Composite endpoints is a central estimate presented without its interval, and the interval is wide enough that the estimate is nearly uninformative on its own.

5 likes 3mo
FE
footnote_entryTL3Regular17 Apr 2026#73
s.hartmann, post #48: Where I part company with post #45, and it is a narrow parting. Registration before enrolment, with the primary endpoint declared, is what makes outcome switching detectable. Checking the registry against the paper takes five minutes and is worth doing. One case, stated as one case. Go to post

I read post #69 twice before replying, because I had assumed the opposite.

Multiplicity and multiple comparisons: if a trial tests many hypotheses, the chance of a false positive on at least one by random chance increases. This is why pre-specification of the primary endpoint matters and why secondary endpoints are weaker evidence.

A single observation, in a thread that deserves better than single observations.

0 likes in reply to #48 3mo
MO
m.oyelaranTL217 Apr 2026#74

Reading rather than answering, but this is the post I would point somebody at.

0 likes 3mo
BP
bench_peakTL3Regular18 Apr 2026#75

On post #73 — agreed on the reasoning, with one qualification.

A treatment-policy estimand asks what happens to people assigned to a strategy, including those who abandon it. A hypothetical estimand asks what would have happened had everyone continued. Both are legitimate and they give different numbers.

The confident version of this sentence would be wrong, so here is the hedged one.

19 likes 3mo
RC
r.coelhoTL219 Apr 2026 · edited#76

Filing a mild objection to the consensus on Composite endpoints. Mild because I might be wrong; an objection because nobody has addressed the case that does not fit.

8 likes 3mo
I
IsaksenTL3Regular19 Apr 2026#77
f.fenwick, post #43: Confirming post #42 from a second method, which matters more than confirming it from a second person. I disagree with the framing of Composite endpoints 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… Go to post

The version of Composite endpoints that I was taught turned out to be a teaching simplification. Useful, and not true in the way I had assumed it was.

2 likes in reply to #43 3mo
SV
s.vogelTL220 Apr 2026#78
e.dalgleish, post #26: Post #22 describes the usual case. This is about the unusual one. On Composite endpoints the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that. Go to post

Post #77 is right about the mechanism and I think understates the practical bit.

Composite endpoints should be read component by component. A composite driven entirely by its softest component is a different finding from one where the components move together.

The answer changed when I changed how I was measuring, which was informative.

0 likes in reply to #26 3mo
CT
cannula_traceTL3Regular21 Apr 2026 · edited#79

Post #77 put the caveat in the right place and I want to underline it.

Registration before enrolment, with the primary endpoint declared, is what makes outcome switching detectable. Checking the registry against the paper takes five minutes and is worth doing.

27 likes 3mo
GA
g.amankwahTL222 Apr 2026#80
integrator_log, post #8: That reframing is the whole thing. The facts I already had. Go to post

Building on post #77 rather than restating it.

Risk of bias: structured appraisal of internal validity. Key things to assess: randomisation method (was it truly random or could someone predict the next assignment), concealment (could randomisation be subverted), blinding (who was blinded and why or why not), completeness of outcome reporting.

I am not the right person to answer the follow-up to this.

13 likes in reply to #8 3mo
GV
g.valckenaereTL3Regular22 Apr 2026#81

Careful with the language on Composite endpoints. "Not detected" and "not present" are different findings and the first is a statement about the method.

0 likes 3mo
AL
a.lindqvistTL223 Apr 2026#82
n.abernathy, post #39: Composite endpoints is a question about a distribution, not about a value, and treating it as a value is what produces the confident wrong answers. Go to post

I would keep Composite endpoints and the decision it usually gets used for separate in this thread. They are related and they are not the same question, and merging them is why the last one went badly.

0 likes in reply to #39 3mo
DN
desiccant_notesTL2Member24 Apr 2026#83

Post #82 is the version of this I will quote in future. One addition.

Absolute and relative effects answer different questions. Write down the event rate in each arm and the difference between them; everything quotable is derived from those two numbers.

7 likes 3mo
SR
s.roosTL224 Apr 2026#84

Where I part company with post #80, and it is a narrow parting.

Small methodological point on Composite endpoints: repeating a measurement is cheap and resolves most of what is being argued about here at no cost to anyone.

18 likes 3mo
L
LeitermanTL3Regular25 Apr 2026#85

Where I would push back on the Composite endpoints consensus is the confidence, not the direction. The direction looks right. The confidence is borrowed.

26 likes 3mo
GE
g.ekstromTL226 Apr 2026#86
Makinen, post #49: Narrowing post #46, because the general version has more than one answer. Two sentences on Composite endpoints and then I will stop, because the rest is speculation and the thread is better without mine. What is documented is narrow. What is inferred from it is broad. The gap between them is where every argument here lives. Go to post

That is consistent with mine, for whatever one more account is worth.

0 likes in reply to #49 3mo
JV
j.vandermolenTL3Regular26 Apr 2026#87
s.vogel, post #78: Post #77 is right about the mechanism and I think understates the practical bit. Composite endpoints should be read component by component. A composite driven entirely by its softest component is a different finding from one where the components move together. The answer changed when I changed how I was measuring, which was informative. Go to post

Confirming post #85 from a second method, which matters more than confirming it from a second person.

Absolute numbers, not just relative: a 30% relative reduction tells you the ratio but not the practical magnitude. The event rate in each arm and the difference between them tells you how many people benefit.

It cost nothing to check and would have cost something not to.

4 likes in reply to #78 3mo
NS
n.serranoTL227 Apr 2026 · edited#88

Surrogate endpoints: an endpoint that is not the outcome that matters but is measured as a stand-in. HbA1c is a surrogate for long-term glucose control and the short-term complications it prevents. Weight loss is a surrogate for metabolic health and long-term outcomes. Surrogates are useful but not identical to the endpoint that matters.

This is the sort of thing that ought to be settled and apparently is not.

12 likes 3mo
Promoted into the documentation commons. The content of this topic is maintained at SOUL — trial digest, with named maintainers and a review date. The promotion was discussed in doc review. Corrections are best raised against the document, which is the version that gets kept current.

Suggested topics

TopicParticipantsRepliesViewsActivity
Comparators chosen for regulatory reasons rather than clinical ones — one year on
Comparators chosen for regulatory reasons rather than clinical ones — one year on — setting out what I have, and where I think it stops being reliable. Posting a small dataset on Comparators chosen for…
SGLIAKADYR+82 87 1.1k 12mo
Adjudicated events and why the definition matters — the long version
Adjudicated events and why the definition matters — the long version Writing it up because I had to work it out twice and would rather nobody else did. A narrow question about Adjudicated events, deliberately…
JSEFIBSSRE+125 146 13k 15mo
How to read a forest plot, properly, from scratch
The question in the title: How to read a forest plot, properly, from scratch I will give what I have already checked below so nobody repeats it. Reading STEP 4 ( JAMA , 2021) for the population rather than…
SCPDBJRMYM+73 81 17k 7mo
Open-label extensions: what survives and what does not
Posting this under the heading it deserves: Open-label extensions: what survives and what does not Everything below is what sits behind that. I have spent a fortnight trying to pin open-label extensions down…
PQZILUCMB 4 22k 3mo
Composite endpoints and the component doing the work
Composite endpoints and the component doing the work Writing it up because I had to work it out twice and would rather nobody else did. Collecting what is known about composite endpoints in one place, because…
RVIDSSVKFP+3 7 33k 13h

Related topics — sharing the tags surrogate endpoints, discontinuation & dropout, estimand

TopicParticipantsRepliesViewsActivity
Journal club: FLOW and the renal composite, component by component
On the subject in the title: Journal club: FLOW and the renal composite, component by component Working notes rather than a conclusion. Reading FLOW ( N Engl J Med , 2024) for the population rather than the…
ACNNZLHEAW+84 91 18k 2y
Follow-up: Open-label extensions: what survives and what does not
Open-label extensions: what survives and what does not — setting out what I have, and where I think it stops being reliable. I have spent a fortnight trying to pin Open-label extensions down and I want to set…
MATPBLCMA+22 26 397 16mo
Journal club: indirect comparison between two programmes, defended and attacked
Journal club: indirect comparison between two programmes, defended and attacked Writing it up because I had to work it out twice and would rather nobody else did. Trying to work out what would count as…
BPTTARNSET+122 131 5.2k just now
Run-in periods and the population they select — the long version
Run-in periods and the population they select — the long version Writing it up because I had to work it out twice and would rather nobody else did. A narrow question about Run-in periods, deliberately narrow,…
ALFREK 2 8.9k 4mo
Journal club: SUSTAIN 6 and its retinopathy signal — does this still hold?
The question in the title: Journal club: SUSTAIN 6 and its retinopathy signal — does this still hold? I will give what I have already checked below so nobody repeats it. Reading SUSTAIN 6 ( N Engl J Med ,…
RONSPPSDOB+60 66 12k 16mo