The answer: a favorable composite result
In SUMMIT, cardiovascular death or a worsening heart-failure event occurred in 9.9% of participants assigned tirzepatide and 15.3% assigned placebo. The trial included 731 people with obesity and heart failure with preserved ejection fraction, or HFpEF. The hazard ratio was 0.62, with a 95% confidence interval of 0.41–0.95.[1]
The two percentages differ by 5.4 percentage points over the observed follow-up. The hazard ratio describes relative event rates over time; it is not another way of writing that absolute difference. Reporting both perspectives helps prevent a large relative figure from obscuring how many people experienced the outcome.
What preserved ejection fraction means for this study
The trial required an ejection fraction of at least 50% and a body-mass index of at least 30. Ejection fraction describes the proportion of blood pumped out of the left ventricle with each beat. “Preserved” should not be read as “no heart failure”: the study enrolled people who already had the disease. Its entry criteria define a specific clinical group, not everyone who has breathlessness or wants to prevent future heart problems.
| Part of the question | What was studied | Do not substitute |
|---|---|---|
| Disease | Established HFpEF | Any unexplained shortness of breath |
| Body size | BMI at least 30 | Every body-size category |
| Main event outcome | Cardiovascular death or worsening heart failure | Cardiovascular death alone |
| Comparison | Tirzepatide versus placebo on background care | One telehealth provider versus another |
Why a composite needs to be unpacked
A composite groups several events into one endpoint. This is often useful because it captures more clinically important events than any one component alone. But its interpretation depends on the components: preventing a worsening episode and preventing death are both valuable, yet they are not interchangeable results.
SUMMIT’s favorable combined result should therefore be described using its full name. A headline that removes “or worsening heart failure” changes the claim. Similarly, a reduction in the chance of having at least one event does not by itself describe every repeat event or the severity of every episode. Those require additional analyses.
- EventsDid death or worsening heart failure occur?
- Health statusHow did symptoms and daily limitations change?
- WeightHow did body weight change?
The study also found improved health status using a heart-failure questionnaire. This matters because daily limitations can be important even when a person has no hospital event. The questionnaire result and the event result reinforce different parts of the clinical picture; neither should be replaced with the weight-loss percentage.
How this differs from SELECT and an approved indication
Our SELECT article concerns a different medicine and a different cardiovascular trial population. It cannot be used to fill gaps in SUMMIT, and SUMMIT cannot establish that tirzepatide has the same effects across every type of cardiovascular disease. Clinical trial names are useful precisely because they keep these questions separate.
The current Zepbound label should be consulted for its approved indications; a publication about HFpEF is not, by itself, a heart-failure approval.[2] Someone with established heart failure needs the treatment decision integrated with their existing disease management, rather than having an online weight-loss subscription replace that care.
A practical care route
For coordinated care, compare the appointment and records arrangements discussed in our PlushCare review and Teladoc Health review. The study-reader tool can keep the composite endpoint visible. The useful service question is whether the treating teams can share the relevant history and decisions—not whether a provider repeats the most impressive trial headline.