One trial can support several publications without becoming several independent experiments. This post hoc SURPASS-CVOT report changes the outcome under examination, not the origin of its participants. Primary analysis
What changed
The analysis reused the 13,165-person trial population. Its six-component outcome combined all-cause death, myocardial infarction, stroke, coronary revascularization, heart-failure hospitalization, and a kidney composite. Tirzepatide versus dulaglutide produced a hazard ratio of 0.84, with a 95% confidence interval of 0.79–0.90. This was a post hoc analysis rather than the original trial's primary cardiovascular test. Primary report
Our appraisal: unpack the composite
A broader endpoint can capture important events missed by a narrower one. It can also change how often events occur and which components dominate the result. A favorable combined estimate does not show that each component improved by the same proportion, or that every component improved at all.
All-cause death and cardiovascular death are not interchangeable labels. Similarly, a procedure and a fatal event differ in clinical meaning. Readers need component-level results and definitions before treating the composite as one uniform kind of benefit.
Timing of endpoint selection matters. Calling an endpoint “primary” within a later analysis does not make it the prespecified primary endpoint of the original randomized trial. Post hoc findings may be informative, but their evidential status should remain visible.
Nested composites also overlap. Agreement among a six-component, five-component, and four-component result is not equivalent to replication in three new cohorts. Many of the same patients and events contribute repeatedly.
This report should therefore be linked to the original SURPASS-CVOT trial, not used to replace its conclusion about the original endpoint. Its value is in extending the questions asked about cardiovascular and kidney outcomes. Release review should examine component frequencies, multiplicity, event adjudication, and the analysis plan before making more specific claims.
Frequently asked questions
Was this a new randomized population?
No.
Was the expanded analysis post hoc?
Yes.
Did the composite include kidney outcomes?
Yes.
Is all-cause death identical to cardiovascular death?
No.
Does the overall result prove every component improved?
No.
Are nested composites independent replications?
No.
Does this replace the original primary endpoint?
No; both reports must retain their distinct questions.