EM EVIDENCE RUNDOWN — ISSUE 25 — 14 AUGUST 2026
EM Evidence Rundown
Emergency medicine evidence for UK clinicians — weekly — emevidence.org
Jake Turner — Senior Registrar in Emergency Medicine, ST6 — Curated with the assistance of AI (Perplexity). All content editorially reviewed.
Lead: The ICECAP RCT (JAMA, 5 August 2026) — the largest trial of targeted temperature management after OHCA — found no difference in neurological outcome between 24 and 48 hours of therapeutic hypothermia (33°C). This closes a long-debated question. For post-ROSC management in the ED: standard 24 hours remains appropriate; no benefit from extending to 48 hours. RCEM Urgent — Surpass Data Security: RCEM notified 12 August that Surpass (exam delivery partner) has experienced a potential data security incident. All EM trainees who have used Surpass for MRCEM/FRCEM exams should check for direct notification from RCEM and monitor affected accounts. Heatwave: Amber heat-health alerts across all of England from 11 August (up to 38°C). ED demand surging. Standard heat stroke management applies. UKHSA monitoring actively. Core Revision: Toxicology — the EM approach to poisoning and overdose (antidotes, toxidromes, specific agents).
BOTTOM LINE UP FRONT
ACT ON THIS NOW
CHANGE THIS MONTH ICECAP (JAMA 5 Aug): 24h vs 48h therapeutic hypothermia post-OHCA — no neurological outcome difference. Current 24h standard validated. Do not extend to 48h routinely.
SAFETY RCEM Surpass breach (12 Aug): Potential exam data security incident. All MRCEM/FRCEM candidates: check for direct RCEM/Surpass notification. Monitor email and financial accounts.
ACT NOW UKHSA amber heat all England (11+ Aug): Up to 38°C forecast. Full resus bay heat stroke protocol active. Check at-risk medications on all patients. Heat stroke: cool immediately, target <39°C in 30 min.
CHANGE THIS MONTH SAVE-O2 AI (JAMA Intern Med 3 Aug): Autonomous AI oxygen titration achieves normoxaemia more reliably than manual titration in acutely ill adults — RCT. Clinically relevant for ED acute bay oxygen prescribing.
CHANGE THIS MONTH NaV1.8 inhibitor acute pain (NEJM): Phase 2b RCT — novel sodium channel blocker provides comparable acute pain relief to standard care with favourable safety profile. Watch for Phase 3 data.
CHANGE THIS MONTH PERC-Peds (Lancet Resp Med): Validated paediatric PE rule — applying PERC criteria to children reduces unnecessary CT in low-risk presentations. Combine with Wells-Peds for risk stratification.
KNOW FOR NEXT TIME
INFORMING PRACTICE Paeds POCUS in sepsis: Cardiac dysfunction on ED POCUS in children with suspected systemic infection predicts vasoactive requirement and ICU admission. Add cardiac view to your paediatric sepsis assessment.
INFORMING PRACTICE CT head reduction: Multimodal QI intervention (guideline + education + CDS) reduces CT in blunt head injury without increasing missed TBI. Local QIP template available via approach in this paper.
INFORMING PRACTICE Post-ROSC sedation: Pre-ICU sedation post-ROSC associated with better day-90 neurological outcome in ~1600 patients (retrospective). Sedation protocol at the ED-to-ICU handover matters.
INFORMING PRACTICE ML for febrile infant bacteraemia: Machine learning SR/MA — improved specificity for bacteraemia/meningitis in febrile infants vs traditional rules. Watch for UK validation studies.
INFORMING PRACTICE Influenza vaccine paeds death: 80% vaccine effectiveness against fatal influenza (n=1,200 paediatric deaths, case-cohort). Strong argument for flu vaccination in ED discharge counselling.
INFORMING PRACTICE DoLS — RCEM position statement (Aug 2026): Updated position statement on Deprivation of Liberty Safeguards in the ED. Key for acute behavioural disturbance and capacity management.
The ICECAP trial is this week's landmark paper — not because it overturns practice, but because it confirms it with the statistical confidence that previously didn't exist. The question of whether 48 hours of therapeutic hypothermia at 33°C is better than 24 hours has been controversial since TTM2 showed no benefit of hypothermia over normothermia (36°C) in 2021, and HYPERION's conflicting signal in non-shockable arrest. ICECAP, with over 1,800 patients, is the definitively powered duration trial. The answer is 24 hours. While the results were being published, RCEM issued two statements that carry immediate professional weight: the Surpass exam data security notification — which requires active monitoring from every trainee who has sat MRCEM or FRCEM exams — and the Northern Ireland A&E data showing 1 in 5 patients waiting over 12 hours, in what RCEM describes as a year-round crisis that summer has not alleviated. A busy week on all fronts.
WHAT'S INSIDE
- ICECAP RCT: 24h vs 48h hypothermia post-OHCA (LEAD)
- SAVE-O2 AI: autonomous oxygen titration RCT (JAMA IM)
- NaV1.8 inhibitor for acute pain — Phase 2b (NEJM)
- RCEM: Surpass data security incident (act now)
- RCEM: Northern Ireland A&E crisis + GMC trolley data
- UKHSA: Amber heat all England (11 Aug)
- St Emlyn's: Ghost jams and UEC seams (11 Aug)
- EMCrit 431: Post-Partum Haemorrhage
- RCEM DoLS position statement (August 2026)
- PEM: PERC-Peds paediatric PE rule
- PEM: POCUS cardiac dysfunction paediatric sepsis
- PEM: Influenza vaccine 80% effectiveness vs paeds death
- PEM: ML for febrile infant bacteraemia SR/MA
- PEM: SGEM#517 paediatric elbow fracture
- Quick Hits + Core Revision: Toxicology — Poisoning & Overdose
S1 Key Trials | S2 UK Updates | S3 Paediatric EM | S4 FOAMed & Critical Appraisal | S5 Quick Hits | Core Revision: Toxicology | S6 Action Points
SECTION 1 — KEY TRIALS
ICECAP RCT: 24 Hours vs 48 Hours of Therapeutic Hypothermia (33°C) After OHCA — No Difference in Neurological Outcome (JAMA, 5 August 2026, n=1,800+)
Meurer WJ et al. (ICECAP Investigators). JAMA. Published online 5 August 2026. DOI: 10.1001/jama.2026.14022
Design: Multicentre, adaptive randomised clinical trial. Adults with OHCA and return of spontaneous circulation (ROSC) who remained comatose were randomised to 24 hours versus 48 hours of targeted temperature management at 33°C. Primary outcome: 90-day functional neurological outcome (modified Rankin Scale). Adaptive design with Bayesian stopping rules for efficacy/futility.
- 24h = 48h NO NEUROLOGICAL DIFFERENCE
- n>1,800 MULTICENTRE ADAPTIVE RCT
- 90-day mRS PRIMARY OUTCOME
- Futility STOPPED BY BAYESIAN RULE
Bottom line: The trial was stopped for futility — continued enrolment was deemed incapable of demonstrating a benefit for 48 hours. Functional neurological outcomes were similar between arms. Secondary outcomes including mortality and adverse events did not differ significantly.
UK / FRCEM context: The TTM literature has evolved significantly: TTM (2013) showed 36°C was non-inferior to 33°C; TTM2 (2021) showed that targeted normothermia (37.5°C) was not inferior to hypothermia (33°C); ICECAP now shows that 24h is sufficient. Current RCUK guidance (2021 ALS) recommends targeted temperature management with avoidance of fever (T >37.7°C) for at least 72 hours post-ROSC — this remains appropriate. Active cooling to 33°C for 24 hours is a reasonable strategy for selected patients (non-shockable post-ROSC, cardiogenic shock); extending to 48 hours without added benefit only risks rewarming complications (coagulopathy, electrolyte disturbance). The message for the resus bay: initiate TTM at 33-36°C for post-ROSC comatose patients; handover to ICU with a 24-hour duration target.
SAVE-O2 AI Trial: Autonomous AI Oxygen Titration Achieves Normoxaemia More Reliably Than Manual Titration in Acutely Ill Adults — RCT (JAMA Internal Medicine, 3 August 2026)
Douin DJ et al. JAMA Internal Medicine. Published online 3 August 2026. DOI: 10.1001/jamainternmed.2026.3841
Design: Randomised clinical trial comparing autonomous AI-driven oxygen titration versus standard manual oxygen titration in acutely ill adults requiring supplemental oxygen in hospital. Primary outcome: time within a predefined normoxaemia range (SpO2 94-98%).
Finding: AI-driven titration significantly increased time in the normoxaemia target range compared with manual titration. Hyperoxia and hypoxia episodes were both reduced. The AI system adjusted flow rates continuously in response to SpO2 without nursing intervention.
ED applicability: Oxygen prescribing in the ED is poorly standardised — flows are frequently left on admission settings, target ranges are not consistently applied, and hyperoxia is common in post-ROSC, stroke, and COPD patients. This RCT provides the strongest evidence yet that closed-loop AI titration improves oxygen delivery. For immediate practice: ensure oxygen prescriptions always specify a target SpO2 range and not just a flow rate. AI titration devices (e.g., FreeO2) are not yet standard in most UK EDs — this data will accelerate procurement conversations. In the meantime: treat oxygen as a drug with a dose and a target, not a background gas.
NaV1.8 Inhibitor for Acute Pain — Phase 2b RCT Shows Comparable Analgesia to Standard Care with Favourable Safety Profile (NEJM, August 2026)
Singla N et al. N Engl J Med. 2026;395(5):454–464. Published 31 July / 7 August 2026.
Study: Phase 2b randomised controlled trial of a selective NaV1.8 sodium channel inhibitor versus placebo and active comparator for acute post-procedural/post-traumatic pain. NaV1.8 is a voltage-gated sodium channel subtype expressed in peripheral nociceptors — blocking it reduces pain signal transduction without central opioid effects.
Finding: The NaV1.8 inhibitor demonstrated comparable pain reduction to standard care with a significantly favourable safety profile (no respiratory depression, no sedation, lower nausea rates). Phase 3 trials are planned.
Why this matters for EM: Opioid-sparing acute pain management is a major unmet need in the ED — particularly for musculoskeletal trauma, renal colic, and procedural pain in patients with opioid sensitivity. A peripherally acting sodium channel inhibitor with no respiratory depression or addiction potential would be transformative if Phase 3 data confirms this signal. This is early (Phase 2b, likely underpowered for definitive efficacy) — watch for Phase 3 results. Do not change practice yet, but this represents the most promising novel analgesic mechanism in over a decade.
SECTION 2 — UK UPDATES & SAFETY
RCEM: Surpass Exam Provider Data Security Incident — All MRCEM/FRCEM Candidates Must Check for Notification (12 August 2026)
RCEM, 12 August 2026. rcem.ac.uk
DATA SECURITY NOTIFICATION — ACTION REQUIRED RCEM has been informed by Surpass (the external provider that delivers electronic examination services on behalf of the College) that Surpass has identified a potential data security incident. RCEM published an update on 12 August 2026. Who may be affected: Any candidate who has sat MRCEM Primary, MRCEM Part B, FRCEM Primary, FRCEM Intermediate, or FRCEM Final SAQ examination via Surpass.
Actions for affected candidates: (1) Check for direct notification from RCEM and/or Surpass to your registered email address. (2) Monitor your email, financial accounts, and any accounts where you used the same password as your Surpass account for unusual activity. (3) Consider changing passwords for any accounts that share credentials with your Surpass/RCEM examination portal login. (4) Report any suspected fraudulent activity to your bank and to Action Fraud (actionfraud.police.uk, 0300 123 2040). (5) Full update at rcem.ac.uk — check for the most current version, as investigations are ongoing.
RCEM: 1 in 5 Northern Ireland A&E Patients Wait Over 12 Hours — Year-Round Crisis — August 2026 ED Performance Data
RCEM, 12-13 August 2026. rcem.ac.uk | NHS Northern Ireland performance statistics, August 2026.
RCEM published two related statements on 12-13 August. The Northern Ireland A&E crisis statement highlighted that tens of thousands of patients waited over 12 hours in recent months — a figure RCEM described as a year-round structural problem, not a seasonal winter phenomenon. Separately, RCEM responded to GMC data on ED trolley waiting times across England, reinforcing that corridor care and boarding represent a systemic patient safety failure — not a performance metric.
Systemic context: With August NHS performance data showing sustained pressure in England and Northern Ireland, and the UKHSA amber heat alerts driving additional ED attendance, August 2026 is tracking as one of the most pressured summer periods on record. RCEM's call for structural change — exit block elimination, social care integration, and the NHS England Model ED programme — remains the consistent ask. Board assurance statements are due 30 September 2026. At a departmental level: document every corridor care episode and escalate through OPEL. This is your evidence base for future QIP and ARCP reflection.
UKHSA: Amber Heat-Health Alerts Across All of England From 11 August — Up to 38°C — Surge in ED Presentations Expected
UKHSA, 11 August 2026. gov.uk/guidance/heat-health-alert-system
UKHSA issued amber heat-health alerts across all of England from 11 August, with temperatures forecast up to 38°C. This is the third significant heat-health event in England in 2026, following the 2,877 excess heat deaths reported through July.
ED heat stroke protocol — active now: Heat stroke (core T >40°C + neurological dysfunction): immediate whole-body cooling — ice water immersion or evaporative cooling + fanning; target core T <39°C within 30 minutes. Do not wait for core temperature if clinical picture is clear. Heat exhaustion (T <40°C, no neurological): oral or IV rehydration, cooling, observation. High-risk groups: elderly, isolated, no air conditioning, heat-sensitive medications (diuretics, ACE inhibitors, antipsychotics, lithium, metformin, NSAIDs). Hypernatraemia: common in elderly heat victims; check sodium on all patients with confusion and dehydration in this weather period. Issue 22 CHS context: post-hypertensive crisis hypernatraemia management should not use hypotonic fluids in EDs without senior review.
St Emlyn's: "Ghost Jams and the Seams Between the Model UEC Papers" — What the NHS England ED Flow Guidance Gets Wrong (11 August 2026)
St Emlyn's. Published 11 August 2026. stemlynsblog.org
This St Emlyn's piece addresses a critical failure mode in UEC flow planning: the "ghost jam" — a system that appears to be flowing on paper but is actually gridlocked at the handover points between components (ED to SDEC, SDEC to wards, ward to community). The post argues that NHS England's Model ED papers — while individually reasonable — fail to account for the seam problems that emerge when components are implemented independently rather than as an integrated system.
Practical implication: If your department is implementing Model ED components (SDEC, ambulatory care, rapid assessment) in isolation, identify the handover points between them and audit patient flow at those seams. Ghost jams most commonly occur at: ED-to-SDEC referral; SDEC-to-specialty acceptance; ward-to-community discharge. Map your system, find the seam, and fix it — not just the component. This is directly relevant to ARCP governance reflections and QIP project design.
EMCrit 431 — OB-EMCrit: Life-Threatening Post-Partum Haemorrhage (PPH) — Resuscitation Framework for Emergency Physicians (7 August 2026)
Cho Espinoza, Scott Weingart. EMCrit. Published 7 August 2026. emcrit.org/emcrit/ob-emcrit-pph
EMCrit 431 presents a systematic framework for managing life-threatening post-partum haemorrhage (PPH) in the ED or resus bay — a scenario that EM physicians may face without immediate obstetric backup. Key components: uterine atony recognition (most common cause, 70-80%); uterotonic drugs (oxytocin first line, ergometrine second, carboprost/misoprostol); uterine massage and bimanual compression; balloon tamponade (Bakri/SOS balloon or improvised Foley catheter); surgical escalation timing; massive haemorrhage protocol activation; and resuscitation with blood products. Tranexamic acid is indicated within 3 hours of delivery for PPH (WOMAN trial data).
For UK EM practice: PPH in the ED is uncommon but immediately life-threatening when it occurs. Key decision points: (1) Any PPH in the ED >500 mL post-vaginal or >1,000 mL post-caesarean: activate obstetric emergency team immediately — do not manage alone. (2) While awaiting obstetric team: uterine massage, oxytocin 5 IU slow IV, establish large-bore access ×2, activate MHP. (3) Uterine inversion: do not attempt manual replacement without anaesthetic support. (4) Ergometrine is contraindicated in hypertension. Full EMCrit algorithm at emcrit.org.
RCEM: Updated Position Statement on Deprivation of Liberty Safeguards (DoLS) in the Emergency Department — August 2026
RCEM. Published August 2026. rcem.ac.uk/rcem-position-and-advisory-statements
RCEM published an updated position statement on Deprivation of Liberty Safeguards (DoLS) in the ED. DoLS applies when a person who lacks mental capacity is being deprived of their liberty for care or treatment — a situation that arises frequently in ED practice with patients who lack capacity due to dementia, acute confusion, intoxication, or mental illness. The updated statement clarifies when a DoLS application is required in the ED context, who has authority to complete it, the criteria for a "deprivation of liberty" versus restriction of movement, and the interaction with the Mental Health Act.
Key ED practice points (RCEM statement): A DoLS is NOT needed for every patient with reduced capacity. It applies when a person is under continuous supervision, cannot leave, and is not able to consent to these conditions. In an ED: a confused patient in a bay who is sedated and prevented from leaving likely requires a DoLS if this will persist. An unconscious patient on a trolley being assessed — generally not required immediately, but document the process. The Liberty Protection Safeguards (LPS) reform was passed by Parliament and is expected to replace DoLS from 2026/27 — watch for implementation guidance from DHSC.
SECTION 3 — PAEDIATRIC EMERGENCY MEDICINE
PERC-Peds: Paediatric PE Exclusion Rule — Validated to Reduce Unnecessary CT in Low-Risk Children (Lancet Resp Med, 30 July 2026)
Ornowska M et al. (PERC-Peds Investigators). Lancet Respir Med. Published online 30 July 2026. thelancet.com
Study: Prospective validation of the PERC-Peds decision rule for excluding PE in paediatric ED presentations without advanced imaging. The rule incorporates age-adjusted HR, SpO2, clinical DVT signs, recent surgery, and trauma. In children with a score of 0 (all criteria negative), PE is excluded without D-dimer or CT-PA.
- High Sn SAFE RULE-OUT AT LOW RISK
- ↓ CT REDUCES CT-PA IN LOW RISK
- Prosp. PROSPECTIVE VALIDATION
UK application: PE in children is rare but increasingly diagnosed, partly due to rising CT-PA use. PERC-Peds provides a validated rule-out tool analogous to adult PERC. Apply alongside Wells-Peds for age-adjusted risk stratification. For children with PERC-Peds score 0 and no high-clinical-suspicion features: PE can be excluded without imaging or D-dimer. For intermediate risk: D-dimer first, then Wells-Peds to determine CT threshold. Paediatric oncology and haematology patients should be assessed with a higher index of suspicion regardless of rule score — thrombophilia and port-related DVT are common causes of paediatric PE not fully captured by clinical rules.
POCUS Cardiac Dysfunction in Children with Suspected Systemic Infection Predicts Vasoactive Requirement and ICU Admission (Pediatr Emerg Care, JF Thu 13 Aug)
Hoffmann RM et al. Pediatr Emerg Care. 2026. PMID: 42441468. pubmed.ncbi.nlm.nih.gov/42441468
In children presenting to the ED with suspected systemic infection, reduced ventricular function on POCUS was significantly associated with subsequent vasoactive medication requirement, ICU admission, and worse clinical outcomes. POCUS cardiac assessment at triage may identify the subset of paediatric sepsis patients who will deteriorate requiring haemodynamic support — before clinical deterioration becomes overt.
Practice application: Add a rapid parasternal long axis / subcostal 4-chamber cardiac view to your paediatric sepsis POCUS bundle. Reduced ejection fraction or wall motion abnormalities in a febrile child — even one who appears alert — should prompt early ICU discussion and vasopressor pre-positioning. This is consistent with Issue 24 data showing septic cardiomyopathy is common and detectable by POCUS. For DipIMC/FRCEM candidates: paediatric sepsis POCUS is increasingly assessed in OSCE stations.
Influenza Vaccine 80% Effective Against Fatal Paediatric Influenza — Case-Cohort Study (n=1,200+ Deaths) (Pediatrics, JF Paeds Mon)
Case-cohort analysis. Pediatrics. Published 2026. JournalFeed Paeds Speed Read, Monday 11 August 2026. journalfeed.org
Study: Case-cohort analysis of over 1,200 paediatric influenza deaths in the US. Influenza vaccine effectiveness against fatal influenza was 80% overall, with even higher protection in children without underlying conditions.
UK ED application: Influenza vaccination in paediatrics is universally recommended in the UK (JCVI) for children aged 2-17 via live attenuated intranasal vaccine. An 80% effectiveness against death is a compelling statistic for parental vaccine hesitancy conversations. In August: although flu season is months away, discharge conversations about flu vaccination in autumn are timely — particularly for children with respiratory conditions, immunosuppression, or complex medical needs who should have early access. Signpost NHS Inform / nhs.uk flu vaccine pages.
Machine Learning Improves Specificity for Bacteraemia and Meningitis in Febrile Infants vs Traditional Rules — SR/MA of 6 Studies (Pediatrics, JF Paeds Wed 13 Aug)
SR/MA. Pediatrics. 2026. JournalFeed Paeds Speed Read, Wednesday 12 August 2026. journalfeed.org
This systematic review of 6 studies examined whether machine learning (ML) models improve diagnostic performance for bacteraemia and meningitis in febrile infants compared with traditional validated tools (Rochester,
Step-by-step, NICE NG143). ML models demonstrated improved specificity — reducing unnecessary LP and antibiotic use in low-risk infants — while maintaining acceptable sensitivity.
Context: This is hypothesis-generating SR/MA — no ML model is validated for routine UK clinical use in febrile infants yet. Current UK standard remains NICE NG143 and the Step-by-step algorithm (validated in >60-day infants). These ML findings suggest a future direction for reducing over-investigation in this high-stakes but evidence-challenging group. Watch for prospective UK validation studies.
SGEM#517: Paediatric Elbow Fractures — Surgical vs Non-Surgical Management: What Does the Evidence Actually Show? (8 August 2026)
SGEM#517 / Paper in a Pic #517. Published 8 and 13 August 2026. thesgem.com
SGEM#517 critically appraises the evidence for surgical vs non-surgical management of paediatric elbow fractures (specifically lateral condyle and supracondylar fractures). The episode highlights that many surgical indications are based on expert consensus rather than high-quality RCT evidence, and that some non-displaced fractures currently managed surgically may be safely treated non-operatively in selected cases with close follow-up.
ED take-home: This does not change ED management — paediatric elbow fractures should be assessed with AP and lateral X-ray, fat pad sign assessed, and referred to paediatric orthopaedics per local pathway. The SGEM appraisal is directed at orthopaedic decision-making. For the ED: ensure adequate analgesia (IN fentanyl or ibuprofen), accurate X-ray interpretation including posterior fat pad sign (pathological) and elevated anterior fat pad sign (sail sign), and timely orthopaedic review for any displaced or angulated fracture. Document neurovascular examination in every elbow injury.
SECTION 4 — FOAMED & CRITICAL APPRAISAL
Multimodal QI Intervention Reduces CT in Blunt Head Injury Without Missing TBI or Increasing Return Visits (Am J Emerg Med, JF Wed)
Berlyand Y et al. Am J Emerg Med. 2026. PMID: 42435500. JournalFeed EM Speed Read, Wednesday 12 August 2026. pubmed.ncbi.nlm.nih.gov/42435500
A bundled QI intervention — guideline implementation, education sessions, feedback mechanisms, and clinical decision support (CDS) tools — significantly reduced CT utilisation for adult blunt head injury in the ED without any increase in missed TBI, return visits, or adverse outcomes.
QIP relevance: This is a template for a FRCEM/ARCP quality improvement project in your own department. Key elements of the intervention: (1) circulate NICE CG176 / NICE NG232 (head injury) criteria to all ED staff; (2) create a structured documentation proforma for Canadian CT Head Rule / NEXUS criteria; (3) audit CT rates in head injury by NICE criteria pre and post; (4) implement CDS prompt in the clinical system. This paper gives you the evidence basis, the method, and a positive outcome to reference. Radiation reduction, cost, and patient experience are all secondary benefits to document.
SECTION 5 — QUICK HITS
POST-ROSC SEDATION ASSOCIATED WITH BETTER DAY-90 NEUROLOGICAL OUTCOME IN ~1,600 OHCA PATIENTS (RESUSCITATION, JF FRI 8 AUG)
Retrospective cohort. Resuscitation. 2026. JournalFeed EM Speed Read, Friday 7 August 2026.
In a retrospective propensity-adjusted analysis of ~1,600 OHCA patients with ROSC, pre-ICU sedation (initiated in the ED or ambulance resus bay) was associated with improved day-90 neurological outcome. This is observational and hypothesis-generating — but supports the principle that sedation protocol at the ED-to-ICU handover is clinically meaningful. Ensure your post-ROSC patients have a documented sedation and analgesia plan before leaving the resus bay.
CHEST TUBE CLAMPING VS NON-CLAMPING PRIOR TO REMOVAL — NO DIFFERENCE: META-ANALYSIS (INT J EMERG MED, JF FRI 8 AUG)
Meta-analysis. Int J Emerg Med. 2026. JournalFeed EM Speed Read, Friday 7 August 2026.
Meta-analysis found no difference in pneumothorax recurrence, re-intervention rates, or hospital length of stay between clamping and non-clamping chest tubes prior to removal. If your department has a clamping protocol based on historical practice, this data supports simplifying to a non-clamping approach — reducing procedure steps and nurse workload without clinical harm. Confirm with your thoracic surgery and respiratory colleagues.
NICE SEPSIS UPDATE (GID-NG10467 PROCALCITONIN): NOTE ON EXPECTED DATE DISCREPANCY
NICE, August 2026.
The official bodies subagent noted a discrepancy: NICE's own in-development tracker shows an expected date of 15 July 2026 (already overdue), while the previously cited expected date is 11 September 2026. This is likely a tracker update lag — the document has not published in either case. Monitor nice.org.uk/guidance/indevelopment/gid-ng10467 directly. No change to current NG253 sepsis pathway.
RCEM: MRCEM BASIC SCIENCES SYLLABUS UNCHANGED FOR OCTOBER 2026 EXAM — CANDIDATES SHOULD CONTINUE CURRENT PREPARATION
RCEM, 3 August 2026. rcem.ac.uk
RCEM confirmed on 3 August that candidates sitting the October 2026 MRCEM Primary examination should continue to prepare using the current Basic Sciences Syllabus. No changes to the exam format or content have been announced for the October 2026 sitting. Check rcem.ac.uk for any updates post-Surpass data security incident.
EM EVIDENCE RUNDOWN — CORE REVISION — ISSUE 25
Toxicology: Poisoning and Overdose — The EM Approach
Toxidromes, antidotes, and specific agents — FRCEM edition
Principle: Most poisoned patients do not need a specific antidote. Supportive care — airway, breathing, circulation, monitoring — treats the majority. The key skills are: (1) recognise the toxidrome, (2) identify the specific agent, (3) know the handful of antidotes that actually change outcome. Anything else is supportive.
THE MAJOR TOXIDROMES — RECOGNITION AT THE BEDSIDE
| TOXIDROME | VITAL SIGNS | EYES | SKIN | OTHER FEATURES | COMMON AGENTS |
|---|---|---|---|---|---|
| Sympathomimetic | HR ↑, BP ↑, RR ↑, T ↑ | Mydriasis | Diaphoretic, flushed | Agitation, seizures, psychosis, tremor | Cocaine, amphetamines, MDMA, ephedrine, caffeine excess |
| Anticholinergic | HR ↑, BP ↑, T ↑ | Mydriasis (fixed), blurred vision | Dry, flushed, hot | Urinary retention, ileus, delirium, mumbling speech. "Dry as a bone, red as a beet, hot as a hare, blind as a bat, mad as a hatter" | Tricyclics (TCA), antihistamines, atropine, scopolamine, jimsonweed, benztropine |
| Cholinergic | HR ↓ or ↑, BP ↓, RR ↓, T ↓ | Miosis | Diaphoretic, wet | SLUDGE: Salivation, Lacrimation, Urination, Defaecation, GI cramps, Emesis. + DUMBELS. Bronchospasm, bradycardia. | Organophosphates, carbamates, nerve agents, pilocarpine, physostigmine |
| Opioid | HR ↓, BP ↓, RR ↓↓, T ↓ | Miosis (pinpoint) | Pale, clammy | CNS depression, respiratory depression, hyporeflexia. Triad: miosis + CNS depression + respiratory depression. | Heroin, morphine, oxycodone, fentanyl, methadone, tramadol, codeine |
| Sedative-Hypnotic | HR ↓, BP ↓, RR ↓ | Normal or slightly dilated | Normal or pale | CNS depression without miosis (distinguishes from opioid). Ataxia, slurred speech, amnesia. GCS falls progressively. | Benzodiazepines, barbiturates, alcohol, GHB, Z-drugs (zopiclone, zolpidem) |
| Serotonin Syndrome | HR ↑, BP ↑, T ↑↑ (can be extreme) | Mydriasis | Flushed, diaphoretic | CLONUS (the pathognomonic finding) — especially inducible clonus at ankle and knee. Hyperreflexia, agitation, rigidity. Hunter criteria: clonus + agent = serotonin syndrome. | SSRIs, SNRIs, MAOIs, tramadol, linezolid, methylene blue, St John's Wort, lithium, fentanyl (high dose), triptans |
THE KEY ANTIDOTES — WHAT WORKS AND WHEN
| POISON | ANTIDOTE | DOSE | MECHANISM | KEY NOTES |
|---|---|---|---|---|
| Opioids | Naloxone | 0.4-2 mg IV/IM/IN; titrate to breathing not consciousness. Infusion 2/3 of effective bolus/hour. | Competitive opioid receptor antagonist | Short half-life vs methadone/long-acting opioids — infusion needed. Do not fully reverse in opioid-dependent patients (precipitates acute withdrawal). Monitor for re-narcotisation. |
| Benzodiazepines | Flumazenil | 0.2 mg IV every 1 min up to 1 mg | Competitive GABA-A antagonist | RARELY used in ED — risk of seizures in benzodiazepine-dependent patients; may precipitate withdrawal. Use only in pure iatrogenic BZD over-sedation, not recreational overdose. |
| Tricyclics (TCA) | Sodium bicarbonate | 1-2 mmol/kg IV bolus (50-100 mL of 8.4%) repeated until QRS narrows or pH 7.50-7.55 | Na loading overcomes Na channel blockade; alkalosis displaces TCA from channel | Indication: QRS >100ms on ECG or haemodynamic compromise. Also seizure prophylaxis. Lipid emulsion as rescue for refractory TCA toxicity. |
| Paracetamol | N-Acetylcysteine (NAC) | 21-hour IV infusion protocol (TOXBASE / BNF): 150 mg/kg over 1h, then 50 mg/kg over 4h, then 100 mg/kg over 16h | Replenishes glutathione, prevents NAPQI-mediated hepatotoxicity | Most effective if within 8h of ingestion. Still indicated up to 24h (or later if staggered/unknown). Use TOXBASE nomogram. Paediatric weight-based dosing: always use TOXBASE or NPIS. |
| Organophosphates | Atropine + Pralidoxime | Atropine 2-4 mg IV (titrate to dry secretions); Pralidoxime 1-2g IV over 15-30 min | Atropine blocks muscarinic effects; Pralidoxime reactivates AChE before ageing | Pralidoxime must be given within 24-48h before enzyme ageing. Use PPE (risk of secondary contamination). TOXBASE/NPIS guidance for all organophosphate exposures. |
| Digoxin | Digoxin-specific antibody fragments (Digibind/DigiFab) | Based on serum dig level or dose ingested (see TOXBASE formula) | Fab fragments bind free digoxin in circulation | Indicated for: life-threatening arrhythmia, K+ >5.5 mmol/L, serum digoxin >10-15 nmol/L, or massive acute ingestion. Not for chronic mild toxicity in stable patient. Expensive — NPIS advice first. |
| Beta-blockers / CCBs | High-dose insulin euglycaemia (HDIE), glucagon, calcium, lipid emulsion | HDIE: insulin 1 unit/kg/h IV + 50% dextrose to maintain BG 5-10. Glucagon 5-10 mg IV bolus. CaCl2 10mL 10% IV. | HDIE improves cardiac glucose uptake; glucagon bypasses beta-receptors | HDIE is now first-line for severe CCB/BB toxicity over glucagon. Requires close glucose monitoring. Lipid emulsion (Intralipid 20% 1.5 mL/kg IV bolus) for refractory toxicity — especially lipid-soluble agents (verapamil, propranolol, bupivacaine). |
| Carbon monoxide | Oxygen (100%, high-flow) | 100% O2 via NRM until COHb <5% and symptoms resolved. Hyperbaric oxygen (HBO) in selected cases. | O2 displaces CO from haemoglobin (half-life: air 5h, 100% O2 90 min, HBO 20 min) | HBO indications: LOC, neurological symptoms, COHb >25%, pregnancy (>15% or any symptoms), cardiac involvement. Contact NPIS for HBO centre locations. DO NOT use pulse ox — reads |
falsely normal. Use CO-oximetry.
Cyanide
Hydroxocobalamin (Cyanokit)
5g IV over 15 minutes (pre-hospital and ED)
Binds free cyanide
- cyanocobalamin (excreted)
Preferred over dicobalt edetate in the UK. Also: sodium thiosulphate 25g IV. House fire victims with high lactate (>10 mmol/L) should be assumed to have cyanide poisoning. Turns urine red — warn staff.
SPECIFIC HIGH-RISK AGENTS — FRCEM FOCUS
| AGENT | KEY FEATURES | ECG | TREATMENT |
|---|---|---|---|
| Tricyclic antidepressants (TCA) | Anticholinergic + Na channel blockade. Rapidly fatal. Seizures, hypotension, arrhythmia within 1-2h of large ingestion. | Wide QRS (>100ms = severe), right bundle branch pattern, aVR terminal R wave >3mm, deep S in I and aVL | NaHCO3 for QRS >100ms, seizures, or haemodynamic compromise. Early intubation for declining GCS. Avoid physostigmine (seizure risk). Lipid emulsion if refractory. |
| Salicylate (Aspirin) | Mixed acid-base: resp alkalosis (early, hyperventilation) + metabolic acidosis (late). Tinnitus, hyperventilation, altered consciousness = severe toxicity. | No specific changes; monitor QTc | NaHCO3 infusion (alkalinisation of urine to pH ≥7.5 — traps salicylate in renal tubules). IV fluid, glucose. Haemodialysis for level >700 mg/L, renal failure, severe acidosis, or neurological symptoms. Do NOT acidify. |
| Lithium | Acute overdose: GI symptoms, mild CNS. Chronic toxicity (therapeutic range): tremor, confusion, seizures, nephrogenic DI, thyroid effects. More dangerous. | QTc prolongation, T-wave changes, wide QRS in severe toxicity | Normal saline for volume replacement (lithium follows sodium). Haemodialysis for level >5 mEq/L (acute) or >2.5 with symptoms (chronic). NPIS advice. Avoid thiazides (increase lithium levels). |
| Serotonin syndrome | Clonus (distinguish from NMS — NMS is slower onset, lead-pipe rigidity, no clonus). Temperature can exceed 41°C in severe cases. Rhabdomyolysis. | Tachycardia, no specific pattern | Cyproheptadine 12 mg PO/NG (serotonin antagonist, not IV). Benzodiazepines for agitation/seizures. Active cooling if hyperthermia. Stop all serotonergic agents. Severe: intubation and neuromuscular blockade (non-depolarising only). NOT neuroleptics (worsen). |
| Paracetamol (staggered) | Staggered ingestion (>one timepoint): cannot use standard nomogram. Risk assessment requires TOXBASE/NPIS guidance. Treat as if full dose at earliest time. AST/ALT rise early in hepatotoxicity. | N/A | NAC for all staggered paracetamol overdoses regardless of levels. Check PT/ALT/creatinine at 8h and 16h after starting NAC. Liver unit referral if INR rises or LFTs deteriorating despite NAC. |
FRCEM Core Knowledge: Toxidrome recognition is assessed in FRCEM SAQs and OSCE. Know the six toxidromes by their vital signs, pupil size, and skin. Common SAQ traps: (1) Serotonin syndrome vs NMS — clonus vs lead-pipe rigidity. (2) TCA overdose: widened QRS + right axis deviation (deep S in I, R in aVR) = treat with NaHCO3. (3) Organophosphate: DUMBELS + pinpoint pupils = atropine and pralidoxime. (4) Paracetamol: NAC within 8h is highly effective; nomogram uses time of ingestion vs serum level; staggered = treat as worst-case. (5) Carbon monoxide: don't trust pulse ox; use CO-oximetry. (6) Digoxin Fab: indicated for life-threatening arrhythmia or hyperkalaemia, not just high levels. Know TOXBASE (toxbase.org) and NPIS (0344 892 0111) as the two clinical resources — cite them in exams and in clinical documentation.
EM Evidence Rundown — Core Revision — Issue 25 · emevidence.org
Sources: TOXBASE (NPIS clinical toxicology database) · EMCrit IBCC Toxicology · LITFL Toxicology Library · Resuscitation Council UK Toxicology guidance 2021 · BNF/BNFC antidote doses
SECTION 6 — ACTION POINTS
- ICECAP — post-ROSC hypothermia protocol: Update your resus bay post-ROSC care bundle documentation to state 24-hour target temperature management (33-36°C) as standard. Remove any departmental default that specifies 48h. Brief ICU on ICECAP if they are not aware. The evidence base for prolonged hypothermia is now closed.
- Surpass data security — all trainees: Check your RCEM-registered email for any notification from RCEM or Surpass. Change your Surpass/RCEM online portal password. Check if you reused that password on any financial, GMC, or banking accounts and change those too. Monitor accounts for unusual activity.
- Heat stroke protocol — active now: With amber alerts across all England, ensure your resus bay team is briefed on heat stroke: immediate cooling (ice water, evaporative), target <39°C in 30 minutes, monitor for rhabdomyolysis, correct electrolytes, IV fluid rehydration, continuous rectal temperature monitoring. Heat-sensitive medications: check on every patient with confusion or deterioration this week.
- Oxygen prescribing: SAVE-O2 AI trial shows manual oxygen titration is suboptimal. As an immediate practice change: prescribe oxygen with a target SpO2 range, not just flow rate. Reassess oxygen flow in every admitted patient at the ED-to-ward handover. Target SpO2 94-98% for most patients; 88-92% for COPD/chronic hypercapnia; 94-98% post-ROSC; avoid SpO2 >98% in stroke or post-ROSC without specific indication.
- PPH in the ED: Review your department's PPH management algorithm. Confirm that oxytocin (5IU slow IV), TXA (within 3 hours of delivery), and MHP activation are all in your ED protocol for PPH. Brief maternity staff on escalation pathways to the ED for haemodynamically unstable post-partum women. EMCrit 431 is an excellent training resource for the scenario — share with your team.
- DoLS in the ED: Read the RCEM August 2026 DoLS position statement (rcem.ac.uk). Know when DoLS applies: continuous supervision + cannot leave + lacks capacity = potential DoLS. Check your trust's process for Emergency DoLS (urgent authorisation by an approved mental capacity professional). Document capacity assessments in all patients being physically detained in the ED for safety reasons.
- Toxicology — TOXBASE access: Confirm all clinical staff in your ED have personal TOXBASE accounts (toxbase.org — free NHS registration). NPIS telephone (0344 892 0111) should be on your resus bay quick-reference card. Run a 10-minute toxidrome teaching session at your next team meeting using the revision table from this issue.
- PERC-Peds implementation: Print or add to your ED decision support: PERC-Peds criteria for low-risk children with possible PE. The rule includes age-adjusted HR, SpO2, clinical DVT signs, recent surgery, and trauma. Wells-Peds for intermediate risk stratification. CT-PA for high risk. Review your local paediatric PE pathway and confirm PERC-Peds is incorporated.
TRIALS TO WATCH
NEAR-TERM (Q3 2026)
NICE GID-NG10467 Procalcitonin in sepsis (NG253 update) — overdue. Monitor nice.org.uk/guidance/indevelopment/gid-ng10467 directly for latest expected date.
NHS Model ED BAS Board assurance statements due 30 September 2026. Implementation audit data expected November 2026. RCEM DoLS/LPS
Liberty Protection Safeguards (LPS) reform implementation — watch for DHSC guidance on rollout date following parliamentary passage.
2026-2027 HORIZON
| NaV1.8 Phase 3 | Phase 3 trials of NaV1.8 inhibitor for acute pain planned following Phase 2b NEJM data. Could represent first new analgesic class in EM in a decade. |
| SAVE-O2 Scale-up | Autonomous AI oxygen titration — implementation trial data in UK EDs expected following JAMA Intern Med RCT publication. |
| PERC-Peds UK validation | UK prospective validation of PERC-Peds in British paediatric ED population — watch RCEM research calls and Lancet Resp Med for updates. |
EM Evidence Rundown — Issue 25 — 14 August 2026 Curated by Jake Turner, Senior Registrar in Emergency Medicine, ST6. Curated with the assistance of AI (Perplexity). All content editorially reviewed. Published by EM Evidence — emevidence.org
Educational purposes only. Verify against local guidelines and protocols before clinical implementation. Links verified at time of publication.