Endpoint hierarchy matters when a trial's results point in different directions. CEDIM 2 should not be summarized solely by its favorable early observation. Primary study
What the trial found
This randomized, double-blind, placebo-controlled cardiac trial enrolled 2,330 patients, fewer than the planned 4,000 because recruitment was slow. The six-month composite of death and heart failure was not significantly different between groups. Five-day mortality, a secondary endpoint, favored L-carnitine. Primary report
Our appraisal: neither outcome cancels the other
The primary endpoint describes the main prespecified clinical question. Its neutral result should remain prominent. The early mortality finding is also worth reporting, but its secondary status and statistical uncertainty should remain attached.
Stopping below the planned enrollment affects precision and power. It does not prove that the primary endpoint would have become significant with more participants, nor does it make the observed result irrelevant. The correct response is to examine the confidence interval and acknowledge the unresolved range of effects.
The two endpoints also have different time horizons and definitions. An early mortality comparison cannot be substituted for a later composite that includes heart failure. A benefit confined to one interval, if real, would still require explanation and replication.
Historical background care limits application to present-day treatment decisions. This trial is evidence about its tested clinical context, not a current recommendation to replace established myocardial-infarction care.
The useful conclusion is a neutral six-month primary result with a favorable early secondary signal in a trial that enrolled fewer patients than planned. Full review should inspect the multiplicity approach, follow-up completeness, event adjudication, and treatment context before stronger claims about mortality reduction are made.
Frequently asked questions
What was the primary endpoint?
Death or heart failure at six months.
Was it significantly improved?
No.
Was early mortality a secondary endpoint?
Yes.
Did that early comparison favor L-carnitine?
Yes, as reported.
Why did enrollment stop below plan?
Recruitment was slower than expected.
Does lower enrollment prove a hidden benefit?
No.
What must a fair summary retain?
Both endpoint definitions, their hierarchy, and their different results.