PreHEART score for emergency medical services triage of chest pain: a randomised trial

Summary

PreHEART 3 was a pragmatic randomized, open-label, blinded-endpoint noninferiority trial of 5,170 adults attended by EMS for undifferentiated chest pain in the Netherlands. PreHEART-guided conveyance was noninferior to usual care for a hierarchical composite of unexpected serious adverse events at 3 days (win ratio 1.01, 95% CI 0.89-1.16; prespecified noninferiority lower bound >0.80) and increased non-conveyance from 24.5% to 34.6%. No deaths occurred within 3 days among the 1,526 patients who were not conveyed, and 30-day emergency presentations were reduced (RR 0.88).

Why it matters

A structured prehospital chest-pain score can reduce ED transports in a mature EMS system without worsening the prespecified short-term safety endpoint. Adoption would require local validation, reliable follow-up, and an EMS pathway capable of supporting non-conveyance rather than simply applying the score at the bedside.

Main findings

• PreHEART-guided care was noninferior to usual care (win ratio 1.01, 95% CI 0.89-1.16; P<0.001 for noninferiority).
• Non-conveyance increased to 34.6% versus 24.5%.
• No deaths occurred within 3 days among 1,526 non-conveyed patients.
• Thirty-day emergency presentations were reduced (RR 0.88).

Limitations

The trial was open-label and embedded in the Dutch EMS system, which may limit transport and follow-up generalizability. The primary endpoint was a hierarchical noninferiority construct, and exploratory sex-subgroup signals did not remain significant after multiplicity correction.