The Rhythm | Issue 30
Issue 30 | Week of 26 October 2026
No major cardiology congress fell in this week's evidence window, so the findings below come from journal publication, ahead of Transcatheter Cardiovascular Therapeutics (TCT) 2026, which opens in San Diego on 31 October.
Cardiology Intelligence Briefing: the average is not the answer
A pooled result can be statistically sound and clinically incomplete. This week's evidence shows how much the interpretation can change beneath the headline, and why the reasons differ from paper to paper.
Point-of-care troponin testing shortened emergency department (ED) stays across six hospitals, but the benefit narrowed when the one non-randomised site was excluded, and it was smaller for Māori and Asian patients than for European patients.1 An interatrial shunt appeared neutral overall, yet its safety profile diverged by ejection fraction: signals of harm in one prespecified group contrasted with favourable but uncertain findings in another.2 Across sham-controlled trials, catheter ablation consistently reduced atrial fibrillation (AF) recurrence, while its effect on quality of life varied between studies and with the analytical approach.4
The other two papers expose a different problem: who was counted, and for how long. A common genetic variant enriched among people of African ancestry carried a greater estimated population burden of dilated cardiomyopathy than TTN truncating variants, usually described as the commonest genetic cause; studies drawn largely from European populations could not have found it.3 In adults aged 30 to 45, heavy coronary calcium was associated with a modelled 30-year cardiovascular mortality of 4.5%, against 1.5% with none, a figure that rests partly on extrapolation beyond the observed follow-up.5
None of this justifies treating every subgroup difference as meaningful. Some reflect genuine biological or clinical variation; others arise from trial design, analytical choices or chance. Distinguishing them takes more than a statistically significant headline.
Our view: averages are not misleading, but they are incomplete without context. Before applying any pooled result, ask who was studied, how the outcome was measured and whether the differences withstand scrutiny. The answers to those questions, not the headline, decide whether the evidence applies to the patient in front of you.
