- Design
- Retrospective statewide ambulance registry cohort
- Population
- 7,205 paramedic-suspected STEMI patients with a discharge diagnosis
- Primary outcome
- Final diagnosis and mortality
- Effect
- STEMI confirmed 60.1%; non-cardiac diagnosis HR 1.75 (1.47-2.09) for death vs STEMI
A retrospective statewide registry study in Victoria, Australia, followed 7,205 patients whose paramedics activated a STEMI pathway between 2020 and 2025 and who had a recorded hospital discharge diagnosis.
STEMI was confirmed in 60.1%. A further 11.4% had another acute coronary syndrome, 10.5% another cardiac diagnosis, 8.5% a non-cardiac diagnosis and 9.4% a symptom-only diagnosis. In-hospital mortality was 8.6%, and 17.5% died over a median 2.8 years. After adjustment, those with a non-cardiac final diagnosis had higher long-term mortality than confirmed STEMI (HR 1.75, 95% CI 1.47 to 2.09); other groups did not differ materially.
Pre-hospital activation is meant to be sensitive, so a 60% confirmation rate is not a failure. The message for the emergency department is that the minority who turn out not to have STEMI are not low risk: aortic dissection, pulmonary embolism, sepsis and intracranial events can present this way, and they need the same urgency of diagnosis, without losing rapid reperfusion for those who do need it.
- Keep rapid reperfusion pathways for paramedic-activated STEMI.
- Re-examine the ECG and history on arrival, and look actively for STEMI mimics.
- Consider aortic dissection, pulmonary embolism, sepsis and intracranial causes when the picture does not fit.
- Treat a non-cardiac final diagnosis as high risk, not reassurance.
Why it matters
The patients cancelled off the cath-lab pathway were the ones most likely to die later.
Don't overread it
Registry data; the higher mortality reflects the underlying diagnoses, not harm from the pathway.
The statistics, in plain English
A hazard ratio of 1.75 means about 75% higher risk of death over follow-up compared with confirmed STEMI, after adjusting for measured differences. Its confidence interval (1.47 to 2.09) stays well above 1. The other groups' intervals cross 1, so no difference was shown.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free