The practice-changing evidence, trials and guidelines from the last six months — rewritten in plain clinical language and expanded with the "why" for exams and the shop floor.
Anticoagulating a patient on extracorporeal membrane oxygenation (ECMO) is a constant tug-of-war: the circuit's artificial surfaces trigger clotting, but the same patients are critically ill and prone to catastrophic bleeding. The RATE trial tested whether a deliberately lower-dose unfractionated heparin (UFH) strategy is safer without paying for it in circuit or patient thrombosis.
330 patients (≈60% venoarterial, 40% venovenous). Low-dose UFH gave a non-significant reduction in severe bleeding versus standard dose (58% vs 65%) and did not increase severe thromboembolic complications (11% vs 10%). But interpretation is limited: it was open-label, anti-Xa monitoring was inconsistent, and the "standard" comparator may have been dosed higher than many modern units use.
Bleeding — not clotting — is now the leading cause of death on ECMO, so any strategy that trims haemorrhage without adding clots is attractive. This is a genuine signal, but not yet practice-changing on its own.
Gentler anticoagulation on ECMO looks promising and safe, but the evidence isn't strong enough to change routine practice — keep titrating to your unit's protocol and monitoring.
After return of spontaneous circulation (ROSC), the injured brain is vulnerable to both hypoxia and hyperoxia — excess oxygen drives free-radical reperfusion injury. This trial asked whether a conservative SpO₂ target is as good as a liberal one in ventilated arrest survivors.
1840 patients. Targeting SpO₂ 95% (allowing FiO₂ as low as 0.21) gave similar favourable neurological outcome, mortality and length of stay at 180 days compared with higher targets (allowable minimum FiO₂ 0.3), with no increase in adverse events.
It's tempting to leave a post-arrest patient on 100% oxygen "to be safe," but that reflexive hyperoxia may harm the brain you're trying to protect. Once you have a reliable saturation, wean the FiO₂.
Titrate oxygen down after ROSC — aim for the lower end of SpO₂ 95–98% rather than reflexively running 100% FiO₂.
In intermediate-high-risk pulmonary embolism — RV strain on imaging plus raised biomarkers, but not yet shocked — the dilemma is whether to escalate beyond anticoagulation. This trial compared ultrasound-facilitated catheter-directed thrombolysis (low-dose alteplase delivered into the clot) with anticoagulation alone.
544 patients. The catheter arm reduced the composite of PE-related death, cardiorespiratory decompensation or symptomatic PE recurrence (4% vs 10%), driven almost entirely by fewer decompensation events (3.7% vs 10.3%). Major bleeding was numerically higher (4% vs 2%) but not statistically significant, and RV size at 48 h didn't differ. Limitations: low event rates, no blinding, and few patients over 75.
Catheter-directed lysis uses a fraction of the systemic thrombolysis dose, aiming to relieve RV strain while limiting bleeding. The benefit here is preventing deterioration, not saving lives — so patient selection and a PE response team matter.
Supports offering catheter-directed therapy to selected intermediate-high-risk PE patients to prevent decompensation. It does not reduce mortality — weigh bleeding risk and involve your PE team.
The international sepsis guidelines were refreshed, with a consistent theme: resuscitate enough, then stop — several nudges away from over-treatment.
Years of "more is better" fluid and broad-spectrum antibiotics have their own harms — tissue oedema, organ dysfunction, resistance, Candida and C. difficile. The pendulum is swinging toward precision and restraint.
Screen early, resuscitate judiciously to MAP 60–70, then deliberately de-escalate fluids and antibiotics rather than leaving them running.
HFNC feels better and looks better — warmed, humidified, high-flow oxygen with a little PEEP — but its effect on hard outcomes in acute hypoxaemic respiratory failure has been unclear. This trial compared it with plain low-flow oxygen.
1110 patients. HFNC did not reduce 28-day mortality. It produced a non-significant fall in intubation (42.4% vs 48.4%) and rapidly improved respiratory rate, dyspnoea and CO₂ — real comfort and physiology benefits, but no survival signal.
HFNC is a comfort and work-of-breathing tool, not a mortality drug. Using it routinely for every mildly hypoxic patient ties up resources without changing outcomes — and risks masking a patient who actually needs intubation.
Reserve HFNC for patients with genuinely advanced oxygen needs (e.g. >6 L/min); it eases breathing but won't, by itself, save lives.
Iloprost is a prostacyclin analogue that vasodilates, inhibits platelet aggregation and improves microcirculatory flow in the reperfusing frostbitten limb — countering the thrombosis and vasospasm that kill tissue after rewarming. New data strengthen its role in grade 3–4 frostbite (cyanosis extending proximal to the distal phalanx after rewarming).
Over 250 patients across two Canadian regions that differed in iloprost use but shared a frostbite pathway: iloprost was associated with a lower amputation rate in grade 3 frostbite (30% vs 53%), with a dose-response (higher total dose → fewer amputations). In 40 grade-4 patients it was linked to fewer finger and toe segment amputations.
Historically severe frostbite meant waiting weeks for demarcation, then amputating. Iloprost (and, where available, thrombolysis) can rescue tissue if given early — a genuine shift toward limb salvage.
Give iloprost within 72 hours for severe (grade 3–4) frostbite where available — it saves digits.
Drowning-associated pneumonia (DAP) is classically taught as water-borne gram-negatives, especially Aeromonas. A paediatric ICU series shifts the empiric thinking.
Over 70 children with freshwater DAP in a Chinese PICU: Staphylococcus aureus was the most-isolated organism. Other common pathogens — Mycoplasma pneumoniae, Strep pneumoniae, H. influenzae, Klebsiella, Candida albicans and rhinovirus.
Empiric cover aimed only at Aeromonas and other environmental gram-negatives would miss the commonest isolate. Note that not every drowning needs antibiotics — reserve them for evidence of infection, aspiration of grossly contaminated water, or clinical deterioration — but when you do treat, cover broadly.
When treating suspected drowning pneumonia in a child, cover broadly — including gram-positives (S. aureus) — not just water-borne gram-negatives.
Cold myocardium resists defibrillation and drugs. The traditional teaching: give up to three shocks, then withhold further attempts until the core warms past 30 °C (86 °F), because repeated shocks on a cold heart seemed futile and potentially injurious. New registry data refine the timing.
≈40 arrests with a shockable rhythm and core temperature <30 °C. ROSC was far more likely when defibrillation was delivered during rewarming rather than before it (100% vs 25%). On extracorporeal life support (ECLS), defibrillation carried a fivefold increase in odds of ROSC.
"Not dead until warm and dead" still holds — but this suggests you shouldn't just warm and wait. Continue attempting defibrillation as the patient rewarms, especially on ECLS. Prolonged, high-quality CPR during rewarming is worthwhile in hypothermic arrest.
Attempt defibrillation even below 30 °C, and don't withhold shocks during ECLS rewarming above 25 °C (77 °F). Warm and shock — don't just warm and wait.
Two Crotalinae antivenoms — CroFab and Anavip — differ in carbohydrate content. Anavip carries more galactose-alpha-1,3-galactose (alpha-gal), the mammalian sugar behind alpha-gal syndrome (the tick-bite-acquired red-meat allergy), so it may trigger more reactions in sensitised patients.
Over 200 patients in Arkansas — a state with high alpha-gal sensitisation. Anavip was associated with more adverse drug reactions, including presumed anaphylaxis (22% vs 4%). Basophil activation testing confirmed Anavip more potently activated alpha-gal-sensitised basophils.
Antivenom reactions are usually attributed to the animal-derived protein, but here the sugar is the culprit in a geographically defined population. In alpha-gal country, anticipate and pre-empt anaphylaxis.
In regions with high alpha-gal sensitisation, be ready to manage anaphylaxis after antivenom — especially Anavip.
The YEARS algorithm raises the D-dimer threshold (to 1000 ng/mL) when none of its three items — clinical signs of DVT, haemoptysis, or PE as the most likely diagnosis — are present, sparing many CT scans. Whether it's safe in active cancer, where D-dimer is chronically elevated and PE is common, was unknown.
698 cancer patients with suspected PE, randomised to YEARS vs standard CTPA. The 90-day diagnostic failure rate (missed VTE) was similar and low with YEARS (1.8% vs 5%). Caveats: open-label, patients with life expectancy <3 months excluded, no subgroup analysis by cancer type/stage, and the study stopped early.
Cancer patients get repeated CTPAs with cumulative radiation, contrast nephrotoxicity and incidental findings. A validated rule to safely avoid some scans is valuable — but this needs confirmation before wholesale adoption.
YEARS may be used to exclude PE in active cancer and cut unnecessary CTPAs — promising, but apply with judgement pending validation.
Cyclosporiasis (Cyclospora cayetanensis) is a parasitic infection causing prolonged, relapsing watery diarrhoea, often with profound fatigue and anorexia, from faecally-contaminated fresh produce. A large US outbreak (traced to a distributed lettuce product) is a reminder it hides among undifferentiated "gastro."
Routine stool cultures and standard O&P miss it. Specifically request Cyclospora testing (stool PCR where available, or modified acid-fast staining) and take a careful food and travel history in persistent watery diarrhoea. Unlike most watery diarrhoea, it does not respond to routine antibacterials — the treatment is co-trimoxazole (TMP-SMX).
Because it's self-limited only slowly (weeks) and needs a specific drug and a specific test, missing it means a patient suffers for weeks and a foodborne outbreak goes unrecognised.
Persistent watery diarrhoea + relevant food/travel exposure → ask the lab specifically for Cyclospora and treat confirmed cases with co-trimoxazole.
Marine toxins are a major, under-recognised cause of non-infectious foodborne illness. US surveillance found 402 marine-toxin outbreaks (1280 people) over 2011–2023, nearly all scombroid or ciguatoxin — underscoring the value of asking about recent fish.
Both are clinical diagnoses easily mistaken for allergy or gastroenteritis; the give-away is the meal and the pattern.
Ask about recent fish consumption: histamine rash soon after eating = scombroid (antihistamines); GI upset then paraesthesiae + hot/cold reversal = ciguatera (supportive). Report to public health.
The American Thoracic Society issued clear guidance on choosing between HFNC and noninvasive ventilation (NIV) when low-flow oxygen isn't enough.
The two devices do different jobs: HFNC mainly improves oxygenation and comfort; NIV provides pressure support that reduces work of breathing and clears CO₂. Matching the tool to the failure type avoids both under- and over-support.
Simple rule: HFNC for the hypoxic, NIV for the hypercapnic. Either works for pre-oxygenation; post-extubation, escalate to NIV for high-risk patients.
Acute trigeminal neuralgia (TN) flares cause severe, electric-shock facial pain, and rescue options in the ED have been thin and evidence-light. A small randomised trial gives trial-grade support to IV fosphenytoin (a sodium-channel blocker, the water-soluble prodrug of phenytoin).
22 patients with acute TN. Pain at 2 hours fell more with fosphenytoin than placebo (−6.1 vs −2.4 on a 0–10 scale) after a loading dose then once-daily maintenance for up to four days. More patients achieved >50% pain reduction and had fewer recurrent attacks; it was generally well tolerated.
Opioids work poorly for neuropathic TN pain, and oral carbamazepine takes time to titrate. A rapid IV sodium-channel-blocker bridge fills a real gap — small trial, but a start.
Consider IV fosphenytoin as a rescue for severe breakthrough trigeminal neuralgia — now backed by randomised evidence, not just anecdote.
A joint multi-society (AHA/ACC and others) guideline replaces the old low/submassive/massive language with a graded A–E severity classification that maps to management.
The old terms lumped very different patients together. A clean severity ladder gives ED, medicine and the PE team a shared language for who goes home, who is admitted, and who needs reperfusion. (See the full write-up in the Latest Guidelines tab.)
A–C get anticoagulation; D–E are reperfusion candidates. The "R" flags a normotensive-but-hypoxic patient who is sicker than their BP suggests.
Fresh US estimates of paediatric sepsis using the 2024 Phoenix sepsis criteria (which define sepsis by organ dysfunction — respiratory, cardiovascular, coagulation, neurological — rather than the old SIRS-based definition), from surveillance of over 3.9 million hospitalisations (2016–2023, neonates excluded).
Average annual hospital incidence of sepsis 1.3% (over 18,000 cases/year); in-hospital mortality 10.1% (over 1800 deaths/year), stable over the study period. These update older, definition-dependent estimates.
Roughly 1 in 10 children hospitalised with sepsis dies — a sobering, stable figure that reinforces why rapid recognition and the first-hour bundle genuinely change outcomes.
Paediatric sepsis remains common and deadly (≈10% in-hospital mortality). Early recognition and prompt treatment are where the difference is made.
Children hospitalised with severe community-acquired pneumonia are often kept on prolonged IV antibiotics. PediCAP tested whether switching to oral amoxicillin once improving is as safe.
In this factorial randomised trial of children aged 2 months to 6 years across 13 hospitals in five sub-Saharan African countries, early oral amoxicillin step-down was non-inferior to IV-only therapy: readmission-or-death was about 5.6% (amoxicillin) vs 6.3% (IV-only), with similar adverse events.
Once a child with uncomplicated severe pneumonia is clinically improving and tolerating oral intake, stepping down to oral amoxicillin is safe and shortens IV exposure.
Switch the improving child with uncomplicated severe pneumonia to oral amoxicillin -- no need to keep the drip running.
The Surviving Sepsis Campaign international paediatric guidelines were updated, reaffirming the fundamentals and adding monitoring guidance.
Children compensate then crash: they maintain blood pressure until late, so tachycardia, poor perfusion and lactate — not hypotension — are the early warnings. Escalating to inotropes before the pressure falls saves lives.
Recognise fast, give fluids + antibiotics in the first hour, escalate to adrenaline/noradrenaline early, and steer with lactate, POCUS and ScvO₂.
Balanced solutions (e.g. lactated Ringer's) vs 0.9% normal saline for children in septic shock — does the choice affect hard outcomes, as it may in adults?
Over 8400 children with suspected septic shock. Combined death, renal-replacement therapy and persistent kidney dysfunction at 30 days were similar (balanced 3.4% vs saline 3%). Saline caused more hyperchloraemia and hypernatraemia; balanced fluid caused more raised lactate (lactate is metabolised from Ringer's, transiently nudging levels up). May not generalise to hospital-acquired sepsis or resource-limited settings.
Large-volume saline delivers a big chloride load that can cause hyperchloraemic acidosis; balanced fluids avoid it. Here neither translated into a mortality/renal difference — so either is acceptable, but the physiology still favours balanced fluids for large-volume resuscitation.
Either fluid is acceptable for resuscitating paediatric septic shock — balanced crystalloids simply avoid the chloride/sodium load (and the Ringer's-related lactate bump is benign).
For markedly displaced paediatric distal radius fractures (Salter-Harris I/II or metaphyseal), does surgical reduction and fixation beat casting alone? Children's bones have huge remodelling potential, so the question is whether surgery's extra cost and risk are justified.
The CRAFFT trial: 750 children aged 4–10. At 3 months, surgery gave slightly better function, pain, quality-of-life and cosmesis scores — but the differences were small, below the minimally important threshold, and had disappeared by 12 months.
The young growing skeleton remodels angulation impressively, so a well-reduced cast often ends up equivalent to surgery within a year — avoiding anaesthesia, hardware and a second operation for removal.
Casting alone by an experienced clinician is reasonable in children under 10 with these fractures. Continue to support prompt reduction and casting; surgery isn't mandatory for displacement alone.
We tend to assume toddlers "bounce back" from head knocks and can't report symptoms anyway. This prospective study challenges that.
235 young children (median age 2.4 years, 98% GCS 15) diagnosed with concussion in the ED (by GCS plus ≥1 of: palpable skull fracture, headache, altered mental status, or loss of consciousness). At one month, 28% had persistent parent-reported symptoms — significantly more than injured or community controls. The strongest predictor was ED symptom severity.
Just like older children and adults, young children can have prolonged post-concussion symptoms, and the sicker they look at presentation, the higher the risk. This should shape parental counselling and follow-up, not just discharge-and-forget.
Young children get persistent post-concussion symptoms too — the more severe the ED presentation, the higher the risk. Counsel parents and safety-net accordingly.
How much displacement of a paediatric medial epicondyle avulsion fracture warrants surgery has long been debated (traditional cut-offs of 2–5 mm are poorly evidenced). The SCIENCE trial compared immobilisation with surgical fixation.
334 children (24% with an associated reducible elbow dislocation). At 12 months, function, sports/performing-arts participation, pain and quality of life were similar between groups; the surgical group had more complications.
This supports non-operative management for most closed, displaced medial epicondyle fractures — but surgery is still needed for specific indications.
Non-surgical treatment suits most closed, displaced medial epicondyle fractures. Reserve surgery for a fragment entrapped in the joint, ulnar nerve dysfunction, irreducible dislocation, or elbow instability.
Extended VTE prophylaxis after hip or knee replacement is standard, but the simplest effective regimen is debated. This NEJM trial compared aspirin alone against a short rivaroxaban course followed by aspirin.
After total hip or knee arthroplasty, aspirin alone was non-inferior to rivaroxaban-then-aspirin for preventing symptomatic venous thromboembolism, with no clinically relevant difference in bleeding.
Aspirin monotherapy is a cheap, simple and effective option for extended thromboprophylaxis in these elective orthopaedic patients.
Plain aspirin is enough for VTE prophylaxis after hip or knee replacement -- no rivaroxaban lead-in required.
Transurethral bladder catheterisation (TUBC) for a sterile urine sample in non-toilet-trained infants often fails when the bladder is empty — a distressing, sometimes contaminated "dry tap." Using point-of-care ultrasound to confirm the bladder holds urine before you catheterise should improve success.
Meta-analysis of 3 randomised trials (337 children). POCUS-guided timing raised first-attempt success (89.7% vs 72.5%) and slashed dry taps (3.6% vs 23.9%).
A 10-second bladder scan spares the child repeated attempts, spares the family distress, and reduces the risk of a contaminated or failed sample driving unnecessary repeat testing or antibiotics.
Scan the bladder before you catheterise a young child whenever ultrasound is available — far fewer failed, distressing attempts.
Hydrogen sulfide (H₂S) — the rotten-egg "sewer gas" — poisons mitochondrial cytochrome oxidase, blocking aerobic metabolism much like cyanide, causing rapid collapse and lactic acidosis. A poison-centre series informs antidote thinking.
168 patients: those working underground or in confined spaces were more likely to develop severe toxicity. Early sodium nitrite was associated with potentially lower odds of death (not statistically significant). Sodium nitrite induces methaemoglobin, which binds sulfide (as it does cyanide), pulling it off the mitochondria.
H₂S can drop a worker instantly ("knockdown") and endanger rescuers who enter the same space. Recognising the sudden-collapse-in-a-confined-space pattern — and treating early with sodium nitrite plus 100% oxygen and decontamination — may be lifesaving.
Suspect H₂S in a worker with sudden collapse, lactic acidosis and shock in a confined space; give high-flow oxygen and consider sodium nitrite — and keep rescuers safe.
Illicit substances — especially opioids — are an increasing cause of fatal poisoning in children and adolescents, from exploratory toddler ingestions to adolescent use.
US surveillance: death from poisoning in those aged 1–19 rose 12.7% from 2020–2023. In 2023, 1913 died from illicit overdoses, and 79.5% involved fentanyl — potent enough that a tiny amount is lethal to a small child.
Fentanyl now dominates paediatric overdose deaths. Naloxone should be accessible in homes with any opioids, families counselled on safe storage, and any child with unexplained CNS/respiratory depression treated with naloxone and observed.
Fentanyl drives paediatric overdose deaths — keep naloxone accessible, counsel families on storage, and think opioid in any unexplained paediatric CNS depression.
A reminder that everyday cleaning products remain a leading cause of paediatric poisoning injury — a preventable, high-volume problem.
Over 240,000 children aged ≤5 treated in US EDs for cleaning-product exposure (2007–2022); ~60% from ingestion. The worst offenders: bleach, detergents, detergent packets and spray bottles. Concentrated liquid detergent packets are especially dangerous — bright, squishy, mouth-sized, and capable of causing caustic injury, CNS depression and airway compromise.
This is a design and storage problem as much as a medical one. Reinforce child-resistant packaging, out-of-reach storage, and never decanting cleaners into drink bottles.
Reinforce safe storage and child-resistant packaging — detergent packets in particular remain a high-injury, high-severity item.
Cannabinoid hyperemesis syndrome (CHS) is a clinical diagnosis — recurrent severe vomiting in a chronic heavy cannabis user, classically relieved by hot showers. The Rome V criteria refine the older Rome IV definition.
Compulsive hot bathing/showering is supportive but not required.
CHS patients often endure repeated expensive work-ups before the diagnosis is made. In the acute setting, topical capsaicin and antiemetics/benzodiazepines/haloperidol help; standard antiemetics often don't — and the definitive cure is stopping cannabis.
Recurrent vomiting + heavy chronic cannabis + relief with hot showers = CHS. Treat acutely (capsaicin, antiemetics), and the cure is cannabis cessation.
Activated charcoal (AC) adsorbs many toxins within minutes of contact, but "routine" GI decontamination hasn't reduced morbidity or mortality, and the classic 1-hour window is arbitrary. New multi-society recommendations refine when AC actually helps.
AC isn't a reflex for every overdose (and is dangerous if the airway is unprotected — aspiration risk). Targeting it to specific agents and timings maximises benefit and minimises harm.
AC isn't dead — target it to specific agents, consider a longer window or repeat dosing for drugs that linger, and never give it to a patient who can't protect their airway.
Kratom (Mitragyna speciosa) is a botanical whose alkaloids (mitragynine, 7-hydroxymitragynine) act at opioid and other receptors, giving stimulant effects at low dose and opioid-like effects at high dose. Concentrated commercial products differ markedly from raw leaf and can cause toxicity and dependence.
Over 14,000 US poison-centre exposures (2015–2025) — reports plateaued through 2024 then surged in 2025 (over 3400). Cases involving multiple substances (especially opioids) were far more likely to be serious or need hospitalisation.
Kratom can cause agitation, tachycardia, seizures, hepatotoxicity and, in overdose or with co-ingestants, opioid-type respiratory depression that responds to naloxone. It's widely sold as a "natural supplement," so patients may not volunteer it.
Ask specifically about kratom in unexplained agitation, seizures or opioid-like presentations — and watch for the co-ingestants that make it dangerous.
Lead is a classic neurotoxin, but cumulative exposure is increasingly linked to atherosclerosis and cardiovascular death at a population scale — likely via oxidative stress, hypertension and endothelial dysfunction.
Over 42,000 individuals (US NHANES 1999–2014): higher estimated bone lead (a marker of cumulative lifetime exposure) was associated with increased CVD mortality — the highest-burden group had an 88% rise in adjusted mortality risk. Modelling attributes ~3.5 million deaths globally in 2023 to elevated bone lead (~6% of all deaths), making lead the eighth leading global mortality risk factor.
There is no safe blood-lead level. This reframes lead as a cardiovascular risk factor and a major, under-appreciated driver of global mortality — arguing for occupational and environmental screening and mitigation.
Lead is a cardiovascular risk factor, not just a neurotoxin — screen for and mitigate occupational/environmental exposure.
In acquired methaemoglobinaemia, an oxidant drug/chemical oxidises haem iron (Fe²⁺→Fe³⁺), which can't carry oxygen and shifts the dissociation curve left — causing refractory cyanosis with a normal PaO₂ and a characteristic saturation gap (low SpO₂ despite normal arterial oxygen; chocolate-brown blood). A 10-year series maps the causes and severity.
169 patients from Thailand: commonest causes were dapsone (30%) and the herbicide propanil (27%, mostly self-harm). ~10% had G6PD deficiency (which matters for treatment). Outcomes: fatality 11%, 40% needed ICU, 34% intubation, 89% admitted.
It's easily missed — the SpO₂ often reads stuck around 85% and doesn't improve with oxygen. Treatment is methylene blue (which regenerates haemoglobin via NADPH), but it's ineffective and can be harmful in G6PD deficiency — so check when feasible.
Think methaemoglobinaemia in cyanosis unresponsive to oxygen with a saturation gap — dapsone and propanil lead the causes. Treat with methylene blue (caution in G6PD deficiency).
The over-the-counter antidiarrhoeal loperamide is a peripheral opioid that normally doesn't cross into the brain — but in large "supratherapeutic" doses (taken for opioid-like effects or to self-treat withdrawal), it blocks cardiac sodium and potassium channels, causing dangerous ventricular conduction disturbance without the classic opioid toxidrome.
Systematic review of ~130 cases: QRS and QT prolongation, idioventricular rhythm and ventricular tachycardia, reported at daily doses as low as 20–40 mg. Syncope occurred in almost half; severe cases had polymorphic (torsades) or sustained monomorphic VT.
Because it doesn't look like opioid toxicity, the danger is missed. Naloxone may partially reverse CNS effects but not the cardiac channel blockade; management is supportive with ECG monitoring, and sodium bicarbonate / magnesium / overdrive for the dysrhythmias.
Cardiac-monitor supratherapeutic loperamide exposures — watch for wide-QRS and QT prolongation, not just opioid effects.
An epidemiological signal: self-reported psilocybin (magic-mushroom) use rose in US states that decriminalised it for supervised therapeutic use (Oregon 2020, Colorado 2022).
National survey data 2018–2025: mean prevalence rose from 2→4.5% (Oregon) and 2.7→4.3% (Colorado) — a larger rise than in a weighted composite of control states.
Expect more hallucinogen-related presentations where use is liberalised — acute anxiety/"bad trips," accidental injury, and occasionally persistent perceptual or psychiatric effects. Management of acute intoxication is a calm environment, reassurance and benzodiazepines for severe agitation.
Anticipate more hallucinogen presentations where psilocybin is decriminalised — relevant to ED epidemiology and public-health policy.
Ethylene glycol (antifreeze) poisoning is deadly if missed — it's metabolised to glycolic and oxalic acid, causing a high-anion-gap metabolic acidosis and acute kidney injury. The gold-standard gas chromatography is slow and not widely available, so diagnosis often relies on indirect clues (osmolar gap, oxalate crystals). An enzymatic assay (glycerol dehydrogenase-based, approved for veterinary use) can run 24/7 with results in hours.
Over 400 samples analysed by both methods: sensitivity 89%, specificity 83% versus chromatography. Reassuringly, all four false-negatives were below the 20 mg/dL treatment threshold — i.e. it didn't miss anyone needing treatment.
Fast confirmation lets you start (or stop) fomepizole and arrange dialysis without waiting hours for a send-out — and avoids treating on osmolar-gap guesswork alone.
An enzymatic ethylene glycol assay is a useful, fast screening tool — especially where chromatography isn't quickly available — to guide fomepizole and dialysis decisions.
Starting buprenorphine in the ED for opioid use disorder (OUD) reduces all-cause and opioid-related mortality and improves engagement in treatment. This trial tested whether a single extended-release injectable dose works as well as starting sublingual buprenorphine — attractive because it removes the daily-adherence and follow-up hurdle.
≈2000 ED patients with untreated OUD and withdrawal. Extended-release injectable buprenorphine gave similar treatment engagement at 7 days (41% vs 39%) and 30 days (44% vs 45%). Precipitated withdrawal was rare (<1%) in both arms despite a high prevalence of fentanyl (76%).
A one-off injection that covers the patient for weeks bridges the gap when reliable follow-up is uncertain — a practical way to convert an ED visit into sustained OUD treatment.
A single injectable buprenorphine start in the ED is a safe, feasible option for OUD — especially valuable when follow-up adherence is in doubt.
For a large or symptomatic acute epidural haematoma (EDH) — the classic "talk and deteriorate" arterial bleed that is highly survivable if evacuated in time — the best surgical approach was uncertain: craniotomy (remove the bone flap, evacuate, replace the bone) vs craniectomy (leave the bone out to allow for swelling).
Randomised trial, 120 patients with acute traumatic EDH and transtentorial herniation (median injury-to-surgery 4 h). Favourable 6-month outcomes were similar (craniotomy 79% vs craniectomy 84%), as were cerebral infarction and 30-day mortality — but postoperative intracranial haemorrhage was more common with craniectomy (36% vs 13%).
Pure EDH usually isn't accompanied by the diffuse brain swelling that justifies leaving the bone out — so replacing the flap (craniotomy) avoids the extra bleeding and a second cranioplasty, even when the patient herniated.
Supports craniotomy for evacuating EDH, even with herniation — reserve craniectomy for cases with significant oedema or other intracranial pathology.
End-tidal CO₂ (ETCO₂) reflects not just ventilation but pulmonary perfusion and cardiac output — so a low value in a trauma patient can be an early flag of a haemorrhagic, low-output state before blood pressure falls. A meta-analysis quantifies its prognostic value.
Over 2400 trauma patients: a low ETCO₂ on initial presentation was associated with higher mortality (22% vs 12%) and greater transfusion need (7% vs 1%). The "low" threshold ranged 26–35 mmHg across studies but was <30 mmHg in most.
Capnography is cheap, continuous and non-invasive. A low ETCO₂ should prompt a hunt for occult bleeding and factor into activating massive transfusion — used alongside vital signs, shock index and lactate, not in isolation.
Use capnography as an early severity flag in trauma — a low ETCO₂ (<30 mmHg) should raise suspicion of occult haemorrhage, alongside vitals, shock index and lactate.
Clearing the paediatric cervical spine is hard — children are hard to examine, and CT carries real radiation risk to the developing thyroid and brain. The PECARN cervical spine injury (CSI) rule was derived to guide imaging; the question was whether it needs separate age-specific versions (infants differ anatomically from teenagers) or whether one rule fits all.
Planned subanalysis of the 22,430-child derivation cohort. Age-specific rules (0–<2, 2–<8, 8–<16, 16–<18) performed similarly to the single unified PECARN CSI rule; the unified rule's negative predictive value was 99.9% (CSI prevalence 1.9%) — equalling or exceeding every age-specific version.
One rule is simpler to teach, remember and apply than four — and it safely identifies which children of any age need C-spine imaging, sparing the rest from unnecessary CT.
The single PECARN CSI rule can be applied to infants and children of all ages to decide on cervical-spine imaging — no separate age-specific rules needed.
The trend in trauma is transfusing early and balanced. Carrying separate red cells, plasma and platelets on an ambulance is logistically hard; low-titre group O whole blood (LTOWB) delivers all components in one bag. Two pragmatic trials tested whether EMS should switch to it.
LTOWB isn't shown to be superior, but it is equivalent and far simpler — one product to store, check and transfuse — which in a bleeding patient at the roadside is a real operational advantage.
LTOWB is a reasonable, logistically simpler option for EMS prehospital transfusion — equivalent outcomes with one product instead of three.
Fibrinolysis remains the fallback reperfusion strategy for STEMI when timely primary PCI is unavailable, and intracranial haemorrhage (ICH) is its most feared complication. Clinicians need to know who is at highest bleeding risk so dosing can be tempered.
Analyses of tenecteplase-treated STEMI populations consistently link ICH to older age, higher admission blood pressure and higher (full vs weight-reduced) tenecteplase dose; in the STREAM programme, excess ICH in patients >=75 years disappeared once the dose was halved. Exact effect sizes vary by cohort, so the direction of association is the reliable message.
It reinforces halving the tenecteplase dose in older patients and aggressive pre-lytic blood-pressure control before pharmaco-invasive therapy.
Before you give lytics for STEMI, cut the dose in the elderly and get the blood pressure down.
Mechanical thrombectomy transformed large-vessel stroke care, and there was hope the benefit would extend to smaller, more distal occlusions. Robust randomised evidence in that group had been lacking.
The DISCOUNT randomised trial (22 French stroke centres) was stopped early after 163 patients on the DSMB's recommendation, for futility and an increased rate of symptomatic intracranial haemorrhage with thrombectomy. It aligns with the neutral DISTAL and ESCAPE-MeVO trials.
Routine thrombectomy for medium- or distal-vessel occlusion is not supported and may cause net harm; best medical therapy remains the default outside a trial.
Don't reflexively pursue thrombectomy for a distal or medium-vessel clot -- the bleeding risk outweighs the benefit.
After ISAR-REACT 5, guidelines leaned toward prasugrel over ticagrelor for ACS, but whether a health-system-wide default switch improves outcomes was untested at scale.
This registry-based, stepped-wedge cluster-randomised trial switched seven Swedish regions' default P2Y12 inhibitor across 17,095 ACS patients undergoing PCI. Death, MI or stroke was similar (about 11.1% vs 11.8%, adjusted OR ~0.90), while major bleeding was modestly lower with the prasugrel policy.
A default prasugrel strategy does not reduce ischaemic events versus ticagrelor but may reduce bleeding, so either agent is reasonable and tolerability/cost can guide the choice.
Prasugrel and ticagrelor come out even on ischaemic events after PCI -- pick on bleeding risk, cost and adherence.
Patients with atrial fibrillation who also have an ACS need both anticoagulation and antiplatelet therapy, and the safest P2Y12 partner for a DOAC has been uncertain. Guidelines favour clopidogrel, but potent agents are often used.
The EPIDAURUS randomised trial was stopped early (602 of a planned 1,474 patients) on safety grounds: a potent P2Y12 inhibitor (prasugrel or ticagrelor) added to a DOAC increased bleeding without reducing ischaemic events compared with clopidogrel.
In AF patients on a DOAC after ACS/PCI, clopidogrel should be the P2Y12 inhibitor of choice; potent agents add bleeding without ischaemic payoff.
For AF-plus-ACS on a DOAC, pair it with clopidogrel -- not prasugrel or ticagrelor.
QTc prolongation with antidepressants is a recurring safety worry, but whether it translates into real cardiovascular harm has been unclear. This BMJ individual-participant network meta-regression pooled double-blind randomised trials.
Escitalopram and amitriptyline produced the largest average QTc increases (on the order of ~5-9 ms), varying with individual patient factors. Across a large pooled dataset of trials, no increase in early major adverse cardiovascular events was detected versus placebo.
For most patients starting these antidepressants the near-term cardiovascular risk is low, but caution and ECG awareness remain sensible in those with baseline QT risk factors or other QT-prolonging drugs.
A modest QTc bump from escitalopram or amitriptyline rarely means early cardiac events -- reserve extra vigilance for high-risk patients.
About this summary. These entries are an independent educational digest of practice-changing emergency-medicine evidence and guidelines from the recent literature review, re-written and expanded in our own words for teaching and revision. It is a study aid, not a substitute for the primary trials, full guidelines, or local protocols. Always confirm doses, thresholds and indications against current references and your institution's policy before applying them to a patient.
Dr D Academy · Emergency & PG exam prep · Free for all learners.