Paramedic Papers Daily — 3 November 2026

Five new papers today — here are the top four. Paper one — prehospital stroke triage missed one in five reperfusion candidates, and a second expert check on the phone cut that. Paper two — reported loss of consciousness didn't flag post-traumatic amnesia after mild head injury. Paper three — a systematic review of wearables that detect cardiac arrest automatically. Paper four — models that predict ground versus helicopter travel times, and where they struggle.

Paper one — Swiss Medical Weekly. A retrospective single-centre quality-control study of one thousand and seventy-one patients with suspected acute ischaemic stroke, testing how well the prehospital triage system spotted candidates for reperfusion. Forty-one per cent were eligible. The ambulance service under-triaged twenty-one per cent of those — no stroke pathway activated — and over-triaged sixty-three per cent. The interesting bit: crews phone a coordinated medical team that includes a neurologist, and that second expert step cut under-triage to eleven per cent and over-triage to fifty-seven per cent. Seizures, intracranial haemorrhage and functional disorders were the usual false positives. The catch — a physician-staffed call team isn't how Australian dispatch works. But the benchmark numbers are worth having against Australian stroke screens.

Paper two — Emergency Medicine Australasia. A single-centre retrospective cohort of six thousand two hundred and fifty-nine adults with head injuries, all tested in the ED with the Abbreviated Westmead Post-Traumatic Amnesia Scale. Ninety-one point eight per cent passed. Post-traumatic amnesia was more common with an initial Glasgow Coma Scale of thirteen or fourteen, in male patients, after low falls, and after hours. But reported loss of consciousness made no real difference — pass rates were ninety point six per cent with documented loss of consciousness, ninety-two per cent without. The takeaway for crews handing over head injuries is that the initial GCS carries the signal; a reported knock-out on its own doesn't. Caveats: ED data, retrospective records, and no link back to the prehospital notes.

Paper three — Resuscitation. A systematic review of wearable devices that detect cardiac arrest automatically and raise the alarm — smartwatches and the like, catching the collapse the wearer can't report themselves. The abstract isn't published yet, so the findings can't be assessed. Still worth flagging, because automated detection feeds bystander response and dispatch — the front end of every out-of-hospital arrest. One to watch.

Paper four — the Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine. A modelling study that developed and validated ways to predict travel times for ground crews versus helicopters, using an open-source routing tool adjusted with historical data. Ground times predicted well — R squared of point seven nine for drive time to the scene, point seven zero for the trip to hospital. Helicopters were much harder — point four six and point three eight. Mean absolute percentage error sat around eighteen to twenty per cent across every method. The point of all this is the helicopter decision: they're a costly, finite resource, and dispatchers need a decent estimate of the actual time benefit before sending one. The tools are open-source, so any service could adapt them — though accuracy will shift with local geography.

Quick recap. One — Swiss stroke triage missed twenty-one per cent of reperfusion candidates, and a second expert check on the phone cut that to eleven. Two — reported loss of consciousness didn't predict post-traumatic amnesia; the initial GCS did. Three — a systematic review of wearables that auto-detect cardiac arrest, findings still under wraps. Four — ground travel times predict well, helicopter times don't.

Also on the dashboard today — a protocol mapping anti-violence measures for ambulance personnel. That's the roundup for Tuesday the third of November. Back tomorrow.