Nine new papers today — here are the top four. Paper one — only sixty per cent of paramedic-suspected STEMIs were confirmed at discharge. Paper two — bystander CPR keeps rising in older people, but outcomes in the very old haven't moved. Paper three — CPR-related injuries are common, and it's the minutes of compressions that drive them. Paper four — ice-water immersion is the fastest way to cool drug-induced hyperthermia.

Paper one — Annals of Emergency Medicine. A retrospective Victorian statewide registry cohort, just over seven thousand suspected STEMI visits with a recorded discharge diagnosis, from twenty-twenty to mid twenty-twenty-five. Paramedics activate STEMI pathways before any biomarkers or imaging, so the question is how often the suspicion holds up. The answer — STEMI was confirmed in sixty per cent. Another eleven per cent had non-ST-elevation coronary syndromes, ten per cent other cardiac diagnoses, and about nine per cent something non-cardiac. Two thirds got an angiogram, forty-four per cent got a stent, and thirty-day mortality was six and a half per cent. For crews, this is the cost of a sensitive pathway — over-triage is what stops true STEMIs being missed. It's registry data with coded discharge diagnoses, and roughly one in ten visits had no recorded diagnosis at all, so treat the exact splits with some caution.

Paper two — Resuscitation Plus. A Japanese national registry, over three hundred and twenty thousand people aged sixty-five and older with bystander-witnessed cardiac arrest, tracked from twenty-oh-five to twenty-twenty-two. Bystander CPR increased significantly in every age group, but it stalled around fifty per cent in family-witnessed arrests. Good neurological outcome at one month improved only in the sixty-five to seventy-four group, and only when the bystander wasn't family. In the ninety-plus group, bystander CPR reached about eighty per cent, yet good outcomes stayed flat at around one per cent. The uncomfortable read — community CPR training is working, but it isn't translating into better outcomes in the oldest patients, and whether that's patient factors or something fixable, the registry can't say. It's observational trend data from one country, so no causal claims.

Paper three — also Resuscitation Plus. A single-centre retrospective cohort of one hundred and thirty-six out-of-hospital arrests, with an average resuscitation length of twenty-four minutes. Forty-three per cent of patients had injuries from CPR, and seventeen per cent were clinically significant. On regression, what predicted injury was compressions running past twenty minutes — not mechanical CPR, and not ECPR. Mortality was higher in the injured group, seventy versus forty-three per cent, though that likely tracks arrest severity as much as the injuries themselves. The lesson for any prolonged arrest, anywhere — injuries come with the minutes, and survivors of long arrests may warrant imaging. Small single-centre sample, with selection toward ECMO-eligible patients, so hypothesis-generating only.

Paper four — Clinical Toxicology. A systematic review with meta-analysis, seventy-one studies covering just under two thousand patients and twelve cooling methods for drug-induced hyperthermia, with exercise-induced studies included to fill the evidence gaps. Ice-water immersion was the fastest method, at about zero point one seven degrees per minute, cold-water immersion right behind it, and ice-water-soaked sheets noticeably slower. The meta-analysis confirmed immersion clearly beats passive cooling. For stimulant toxicity on a hot day, the practical message is simple — get the patient into cold water, and don't rely on passive measures. The caveat — much of the evidence comes from exercise-induced rather than drug-induced hyperthermia, and a tub of ice water isn't always achievable on the road.

Quick recap. One — three in five paramedic-suspected STEMIs are confirmed at discharge. Two — bystander CPR keeps rising, but the very old aren't seeing better outcomes. Three — CPR injuries track compressions past twenty minutes, not the devices. Four — immersion wins for cooling, passive cooling loses.

That's the roundup for Friday the ninth of October. Five more on the dashboard today, including video laryngoscopy in kids and sepsis intubation complications. Back tomorrow.