EM Evidence Rundown — Issue 35

EM EVIDENCE RUNDOWN — ISSUE 35 — 22 OCTOBER 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 (Claude). All content editorially reviewed.

Full archive and PDF downloads at emevidence.org

Lead: ICare-FASTER (52,601 presentations, 6 New Zealand EDs): point-of-care hs-troponin I in an accelerated pathway cut ED stay by 47 minutes with no change in 30-day MI or cardiac death (0.39% in both arms). Change Tonight: Abbott i-STAT hs-TnI field safety notice: some cartridge lots can give falsely raised troponin from interference; send a laboratory troponin when a raised result does not fit. Paediatric EM (4 items): hypothermia and mortality, reliability of the chest examination in pneumonia, manual detorsion, corridor care in UK EDs; the modified Kocher criteria and paediatric traumatic cardiac arrest are in Quick Hits.

BOTTOM LINE UP FRONT — 22 OCTOBER 2026

ACT ON THIS NOW

ABBOTT FSN i-STAT hs-TnI (Abbott FSN, MHRA-listed): lots beginning A24, A25, B25, B26, C26 or D26 can read falsely high from macrotroponin or heterophile antibodies; if a raised result does not fit the patient, repeat on a laboratory assay or an A26 lot rather than act on serial cartridge results.

WORTH KNOWING ON SHIFT

DIAGNOSTIC Miosis: 1 in 4 opioid-positive patients with a respiratory rate under 10 had normal or dilated pupils (LR− 0.46); treat the respiratory rate, not the pupils.

DIAGNOSTIC HINTS Plus: sensitivity 80.6% for a central cause in acute vestibular syndrome; a “peripheral” result is not a stand-alone stroke rule-out.

REVIEW Necrotising infection: LRINEC 6 or more has 68.2% sensitivity, so a low score does not rule it out.

KNOW FOR NEXT TIME

RCT POC troponin (lead): 47 minutes shorter ED stay (95% CI 33 to 61), safety unchanged; no POC hs-troponin is on NICE’s recommended list.

RCT Intermediate-high-risk PE (PRAGUE-26): catheter-directed alteplase cut the 7-day composite from 6.8% to 0.7%, mostly decompensation; both intracranial bleeds were in that arm.

RCT Vasopressin (VASSPR, n=3,959): starting at noradrenaline 0.1 vs 0.4 µg/kg/min did not change 28-day mortality (31.8% vs 33.0%).

SR/MA Oxygen after ROSC: conservative targets did not lower mortality (RR 0.96) but caused more hypoxaemia (RR 1.65); keep to SpO2 94–98%.

SR/MA Peripheral vasopressors: minor complications in 2.3% (2.6% with noradrenaline) across 33,060 catheters, mostly observational data.

PAEDS Hypothermia: a first temperature of 36°C or less in children carried adjusted odds of death of 20.15, though absolute mortality was about 0.5%.

EM Evidence Rundown, Issue 35. Page 1 of 28

Point-of-care troponin is this week’s thread. ICare-FASTER shows a whole-blood assay can take three-quarters of an hour off an ED stay without missing more infarcts, and a systematic review suggests single-sample rule-out is sensitive across four devices. A field safety notice on a different point-of-care cartridge is a reminder that a fast number can still be a wrong one. PRAGUE-26 reports a large fall in early decompensation with catheter-directed thrombolysis for intermediate-high-risk PE, at the cost of two intracranial bleeds. VASSPR settles nothing about when to add vasopressin, and three resuscitation reviews mostly confirm current RCUK practice. Paediatrics has four items, including a large signal that a cold child in the ED is a sick child.

WHAT’S INSIDE — ISSUE 35

• Lead: ICare-FASTER point-of-care hs-troponin RCT (pathway in Figure 1 below)
• PE: PRAGUE-26 catheter-directed thrombolysis; outpatient management of low-risk PE
• Shock and arrest: VASSPR vasopressin threshold; peripheral vasopressors; oxygen after ROSC
• Diagnosis: miosis and opioids; HINTS Plus; skin and soft tissue infection (BMJ review); sternal fracture
• Pain: opioid use after acute pain
• UK: i-STAT hs-TnI field safety notice; NICE draft on sepsis in pregnancy; MHRA antidepressant warnings; UKHSA Ebola update; RCEM winter survey
• PEM (4): hypothermia, pneumonia examination, manual detorsion, corridor care
• FOAMed: the SSC 2026 dissent; EM Cases on episodic vertigo; REBEL EM on thrombolysis for retinal artery occlusion; EMCrit on the haemodynamically neutral intubation
• Quick Hits (12): including POC troponin accuracy, ventilator vs bag in CPR, toxicological arrest, Rhabdo-RRT, small bowel obstruction, STUMBL, Kocher criteria and paediatric traumatic arrest
• Core Revision: Clinical — tranexamic acid

Contents: 1. Key Trials & Articles — 2. Guidelines & UK Updates — 3. Paediatric Emergency Medicine — 4. FOAMed & Critical Appraisal — 5. Quick Hits — Core Revision — 6. Action Points — Trials to Watch

Figure 1 (lead item). ICare-FASTER pathway (one hospital): a single point-of-care hs-troponin I below 7 ng/L, EDACS under 16 and symptoms for 3 hours or more allows discharge. Laboratory values in square brackets. Source: Than M, et al. Circulation 2026, Figure 2. doi:10.1161/CIRCULATIONAHA.126.080866. © the authors or publisher; reproduced for criticism and review (fair dealing).

EM Evidence Rundown, Issue 35. Page 2 of 28

1 — KEY TRIALS & ARTICLES

LEAD

INFORMING PRACTICE

RCT

FRCEM

ICare-FASTER (Stepped-Wedge Cluster RCT, 6 EDs, 52,601 Presentations): Point-of-Care hs-Troponin I Cut ED Stay by 47 Minutes, With 30-Day MI or Cardiac Death Unchanged (0.39% vs 0.39%)

Than M, Pickering JW, Joyce LR, et al. ICare-FASTER (Improving Care by Faster Risk Stratification Through Use of High-Sensitivity Point-of-Care Troponin in Patients Presenting With Possible Acute Coronary Syndrome in the Emergency Department): A Stepped-Wedge Cluster-Randomized Trial. Circulation. 2026 (published online 19 October 2026). doi:10.1161/CIRCULATIONAHA.126.080866. PMID: 42895015. ACTRN12619001189112. Source route: PDF via BMA Library (supplied by Jake); open access (CC BY-NC-ND). Appraised from full text

Pragmatic stepped-wedge cluster-randomised trial in 6 New Zealand hospitals (February 2023 to January 2025), with each hospital crossing over from control to intervention. All adults tested for possible MI were included: 21,209 control and 31,392 intervention presentations. The comparator was each site’s existing accelerated diagnostic pathway (0 and 2 hour sampling with EDACS) with a central laboratory hs-troponin; the intervention was the same pathway with nurse-run whole-blood Siemens Atellica VTLi hs-troponin I (low-risk threshold 7 ng/L). The primary outcome, ED length of stay, fell by 13% (95% CI 9% to 16%), or 47 minutes (95% CI 33 to 61) from a control mean of 376 minutes. 30-day MI or cardiac death in patients discharged from the ED was 76 of 19,271 (0.39%) vs 50 of 12,906 (0.39%), a difference of 0.01 percentage points (95% CI −0.14 to 0.15).

• 47 min shorter ED LENGTH OF STAY (95% CI 33 TO 61)
• 13% shorter ED LENGTH OF STAY (95% CI 9% TO 16%)
• 0.39% in both arms 30-DAY MI OR CARDIAC DEATH AFTER DISCHARGE
• +0.01 pp ARD FOR SAFETY (95% CI −0.14 TO 0.15)

The pathway is shown in Figure 1 on page 2. CRITICAL APPRAISAL Focused question; registered primary outcome

Yes ACTRN12619001189112; published protocol and statistical analysis plan; ED length of stay

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for a UK ED: Nothing tonight: no point-of-care hs-troponin is on the NICE list. Measure your own blood-draw-to-result time | Confidence in that: moderate | BEEM relevance: 5/7

UK practice: NICE HTG552 (formerly DG40, 2020) recommends 11 laboratory hs-troponin assays for early rule-out of NSTEMI, including single-sample rule-out; the Atellica VTLi point-of-care assay used in ICare-FASTER is not on that list, and NICE recommends further research on the only point-of-care test it assessed (TriageTrue, rec 1.4).

EM Evidence Rundown, Issue 35. Page 3 of 28

INFORMING PRACTICE RCT

PRAGUE-26 (RCT, n=558): Catheter-Directed Alteplase in Intermediate-High-Risk PE Cut the 7-Day Composite From 6.8% to 0.7%, Mostly Decompensation; Both Intracranial Bleeds Were in the CDT Arm

Kroupa J, Radvan M, Mrozek J, et al. Catheter-Directed Thrombolysis in Intermediate-High-Risk Pulmonary Embolism. N Engl J Med. 2026 (published online 31 August 2026). doi:10.1056/NEJMoa2608012. PMID: 42678032. NCT05493163. Funded by the Ministry of Health of the Czech Republic and Charles University. Source route: NEJM author accepted manuscript (supplied by Jake); JournalFeed Article of the Day 21 Oct; Critical Care Reviews. Appraised from full text

Investigator-initiated open-label RCT at 11 Czech tertiary cardiovascular centres (October 2022 to March 2026). 558 haemodynamically stable adults aged 18 to 80 with intermediate-high-risk PE (sPESI 1 or more, RV/LV ratio 0.9 or more, raised troponin or natriuretic peptide; 49.5% transferred in) were randomised. The intervention was conventional catheter-directed thrombolysis, alteplase 1 mg bolus then 1 mg/h per catheter to 20 mg (10 mg if unilateral) over about 9 hours with reduced-intensity heparin; the comparator was anticoagulation alone with UFH or LMWH. The primary outcome, a 7-day composite of death, recurrent PE or cardiorespiratory decompensation, occurred in 2 of 280 (0.7%) vs 19 of 278 (6.8%): risk difference −6.1 percentage points (95% CI −9.2 to −3.0), RR 0.10 (95% CI 0.02 to 0.44). Decompensation drove it (0.7% vs 5.4%); deaths by day 7 were 0 vs 4 (1.4%). BARC 2 or higher bleeding at 7 days was 4.6% vs 5.0%, GUSTO major 1.4% vs 2.2%, and intracranial haemorrhage 2 (0.7%) vs 0.

• 0.7% 7-DAY PRIMARY OUTCOME, CDT (2 OF 280)
• 6.8% 7-DAY PRIMARY OUTCOME, ANTICOAGULATION (19 OF 278)
• NNT 17 CALCULATED FROM ARD −6.1 PP (95% CI 11 TO 34)
• 0.7% vs 0% INTRACRANIAL HAEMORRHAGE (2 VS 0)

Source: PRAGUE-26 author accepted manuscript, Results and Tables 3 and 4. Components add up to more than the composite because one patient can have more than one event. NNT and CI calculated by us. One intracranial bleed followed a hypertensive episode after the infusion, the other a documented LMWH overdose on day 5.

CRITICAL APPRAISAL Focused question; registered primary outcome

Yes NCT05493163; 7-day composite; sample size based on 1.5% vs 6.0%

Design: RCT | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing yet: keep risk-stratifying (sPESI, RV strain, troponin) and involve the PE or cardiology team early; NICE advises against systemic thrombolysis while stable | Confidence in that: moderate Disagreement: The authors argue that earlier catheter-based reperfusion in interventional centres may prevent collapse. JournalFeed (21 October 2026): nearly all the benefit came from soft, clinician-triggered decompensation events judged by an unblinded committee. Critical Care Reviews: no mortality benefit shown; it supports only “selective adoption within experienced pulmonary embolism networks”.

UK practice: NICE NG158 (2020, updated 2023) rec 1.6.3: do not offer systemic thrombolysis to people with PE and haemodynamic stability, with or without RV dysfunction; rec 1.6.2: consider it if haemodynamic instability develops. NG158 makes no recommendation on catheter-directed thrombolysis.

EM Evidence Rundown, Issue 35. Page 4 of 28

INFORMING PRACTICE OBSERVATIONAL

Outpatient Low-Risk PE (Single-Centre Implementation Study, 1,810 Patients): ED Discharge of Low-PESI Patients Rose From 5.1% to 33.3%, and to 75% in a Post Hoc Hestia-Negative Subgroup

Greineder CF, Mefford V, O’Hare C, et al. Adoption of Outpatient Management of Acute Low-Risk Pulmonary Embolism. Ann Emerg Med. 2026 (published online October 2026). doi:10.1016/j.annemergmed.2026.08.738. PMID: 42878715. Source route: PDF via BMA Library (supplied by Jake); open access (CC BY). Appraised from full text

Before-and-after implementation study in one US quaternary academic ED, comparing a preimplementation period (October 2016 to December 2019) with three successive years from March 2021 to March 2024. The intervention had four parts: clinician education, an electronic record nudge triggered by an automated PESI of 85 or less, an order set with the first DOAC dose in the ED and social-work medication help, and a follow-up clinic within 7 to 10 days. Of 1,810 patients with acute PE, 705 (39.0%) had a low PESI; the primary outcome, discharge from the ED, rose from 17 of 331 (5.1%) to 45 of 135 (33.3%) in year III (unadjusted OR 9.24, 95% CI 5.04 to 16.92). In a post hoc analysis restricted to low-PESI patients without modified Hestia criteria it rose from 17 of 189 (9.0%) to 45 of 60 (75%); there were no recurrent VTE or deaths within 30 days of any ED discharge.

• 5.1% LOW-PESI DISCHARGED, BEFORE
• 33.3% LOW-PESI DISCHARGED, YEAR III
• 9.0% to 75% LOW-PESI, HESTIA-NEGATIVE, DISCHARGED (POST HOC)
• 0 of 91 30-DAY VTE OR DEATH AFTER DISCHARGE, YEARS I–III (CALCULATED)

CRITICAL APPRAISAL Exposure and outcome measured reliably

Yes All CTPE patients chart-reviewed, PESI checked by an emergency physician, nudge firing used to confirm case capture

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Audit how many of your low-risk PE patients go home; NICE already supports outpatient care, and the barriers here were workflow (first DOAC dose, prescription, follow-up slot), not evidence | Confidence in that: moderate

UK practice: NICE NG158 (2020, updated 2023) rec 1.2.1: consider outpatient treatment for suspected or confirmed low-risk PE using a validated risk stratification tool; rec 1.2.4: agree a monitoring and follow-up plan with written information, direct contact details for a thrombosis team and out-of-hours advice. The BTS guideline for the initial outpatient management of PE (Howard et al, Thorax 2018;73(Suppl 2):ii1–ii29) says PESI class I/II, sPESI 0 or meeting the Hestia criteria should prompt consideration of outpatient care, that exclusion criteria should be applied when PESI or sPESI is used, and that patients should have a formal review at least once in the first week after discharge. Applying exclusion criteria on top of PESI is the combination this paper arrives at post hoc.

EM Evidence Rundown, Issue 35. Page 5 of 28

INFORMING PRACTICE

RCT

FRCEM

VASSPR (Cluster Crossover RCT, n=3,959): Starting Vasopressin at Noradrenaline 0.1 vs 0.4 µg/kg/min Did Not Change 28-Day Mortality (31.8% vs 33.0%)

Peltan ID, Groat D, Jacobs JR, et al. Effect of Vasopressin Initiation Threshold on Septic Shock Mortality: The VASSPR Cluster Crossover Randomized Strategy Trial. JAMA. Published online 14 October 2026. doi:10.1001/jama.2026.21333. PMID: 42872917. NCT06217562. Source route: PDF via BMA Library (supplied by Jake); PulmCrit Hot Take 15 October 2026. Appraised from full text

Pragmatic, open-label, cluster-randomised, cluster-crossover trial at 13 hospitals in one US health system (Utah and Idaho, February 2024 to July 2025). Every adult given a vasopressor for septic shock was enrolled; hospitals switched monthly between a strategy recommending fixed-dose vasopressin (0.03 units/min) once other vasopressors reached 0.1 µg/kg/min of noradrenaline or equivalent and one recommending it at 0.4 µg/kg/min. The primary outcome, 28-day all-cause mortality, was 31.8% (628/1,975) with the lower threshold vs 33.0% (655/1,984) with the higher threshold (adjusted marginal OR 0.98, 95% CI 0.87 to 1.06; adjusted risk difference −0.4 percentage points, 95% CI −2.7 to 1.3). Kidney replacement therapy-free days, the key secondary outcome, did not differ (median 29 in both; adjusted OR 1.01, 95% CI 0.91 to 1.14). The strategies separated only partly (table); the one safety signal, more cardiogenic shock with the lower threshold, was exploratory and not adjusted for 16 other outcomes.

• 31.8% 28-DAY MORTALITY, LOWER THRESHOLD
• 33.0% 28-DAY MORTALITY, HIGHER THRESHOLD
• −0.4 pp ADJUSTED RISK DIFFERENCE (95% CI −2.7 TO 1.3)
• OR 0.98 ADJUSTED MARGINAL OR (95% CI 0.87 TO 1.06)

Outcome Lower threshold (0.1) Higher threshold (0.4) Effect (95% CI)

Source: Peltan et al, JAMA 2026, Tables 2 and 3. Doses are noradrenaline base or equivalent. Crude mortality difference −1.2 percentage points (calculated); no NNT is given because the result is null.

CRITICAL APPRAISAL Focused question; registered primary outcome

Yes NCT06217562, registered before enrolment; protocol and analysis plan published before database lock; primary outcome 28-day mortality

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for a UK ED: Nothing in the ED: there is no case for starting vasopressin earlier to save lives; keep noradrenaline as the first agent and leave the timing of a second agent to the critical care team | Confidence in that: moderate | BEEM relevance: 4/7 Disagreement: PulmCrit (Farkas, 15 October 2026) agrees there is no mortality benefit but reads the cardiogenic shock signal (7.4% vs 6.3%) as possibly real, from excess afterload, and now calls vasopressin a niche drug for selected patients (refractory vasoplegia, right ventricular failure, tachyarrhythmia). The trial authors call that signal exploratory, unadjusted for multiple testing and not robust in sensitivity analysis.

UK practice: NICE NG253 (2025, updated September 2026) recommendations 1.8.11 to 1.8.13 tell the ED to discuss with the critical care team whether vasopressors should be given and whether to start them peripherally and, if starting them peripherally, to follow local policy on the choice of vasopressor, dose and concentration; NG253 does not mention vasopressin or a dose threshold. The Surviving Sepsis Campaign adult guidelines (2026) strongly recommend noradrenaline first line and “suggest” adding vasopressin on escalating noradrenaline without naming a dose.

EM Evidence Rundown, Issue 35. Page 6 of 28

INFORMING PRACTICE

SR/MA

UK

Peripheral Vasopressors (SR/MA, 49 Mostly Observational Studies, 33,060 Catheters): Minor Complications in 2.3%; 1 Major Event in 29,596 Short Cannulas

ZhangJian SJ, Niu KY, Chen CB, et al. Incidence of Adverse Events in Peripheral Intravenous Vasopressor Use: A Systematic Review and Meta-Analysis. JAMA Netw Open. 2026;9(3):e260710. doi:10.1001/jamanetworkopen.2026.0710. PMID: 41838001. PROSPERO-registered (August 2025; number not given in the PMC text). Source route: PMC full text (PMC12993702); SGEM #522 (30 September 2026). Appraised from full text

Systematic review of adults given vasopressors through peripheral cannulas or midline catheters for hypotension, shock or critical illness (49 studies, 1997 to 2025; 9 in the ED). Only 4 were RCTs; 31 were retrospective and 14 prospective observational studies, and none had a central-line comparison in the main analysis, so the results are single-arm incidence rates. Minor adverse events (pain, swelling, infiltration, extravasation, thrombophlebitis) occurred in 2.3% of catheters (95% CI 1.5% to 3.7%; I² 95.2%), and 2.6% with noradrenaline (95% CI 1.4% to 4.7%). Short peripheral cannulas had 1 major event (tissue necrosis needing fasciotomy) in 29,596 catheters (0.003%; the pooled estimate rounds to 0.0%), while all 30 venous thromboembolisms were in the 4 midline studies (1,126 catheters; pooled 1.4%, 95% CI 0.4% to 5.4%).

• 2.3% MINOR ADVERSE EVENTS, ALL AGENTS (1.5 TO 3.7)
• 1 in 29,596 MAJOR EVENTS, SHORT CANNULAS
• 1.4% VTE WITH MIDLINE CATHETERS (0.4 TO 5.4)
• 59.7% CENTRAL LINE AVOIDED (46.4 TO 71.7), 38 STUDIES

CRITICAL APPRAISAL Clear question; protocol registered

Partly PROSPERO registration in August 2025, after the search began in February 2025; deviations described in the supplement

Design: SR/MA | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing new: it supports starting noradrenaline through a good, visible, proximal cannula under your local policy rather than waiting for a central line | Confidence in that: moderate Disagreement: SGEM #522 (guest Aaron Skolnik) is more cautious than the authors: incidence is not safety, there is no central-line control group, rare events are the ones retrospective studies miss, and the 59.7% central-line avoidance does not prove the peripheral strategy caused it. The SGEM bottom line is to start through a closely monitored short cannula while deciding whether a central line is needed, not to treat peripheral access as a replacement.

UK practice: NICE NG253 (2025, updated September 2026) recommendation 1.8.11 asks the ED to discuss with the critical care team (or, if they are not available, the senior clinical decision maker) whether to start vasopressors peripherally if central access is not available; 1.8.13 says to follow local policy on agent, dose, concentration and monitoring, keep the line and cannula visible and watch for extravasation. NICE found the evidence limited and made a research recommendation on the safety of peripheral vasopressors. The Intensive Care Society guidance on peripheral vasopressor infusion (v1.1, November 2020, interim update September 2022) recommends a cannula “ideally of size 20G or more” sited proximal to the wrist, a second cannula as a contingency, an infusion pump, invasive blood pressure where appropriate, and a standard peripheral noradrenaline concentration of 16 µg/mL; its listed review date (November 2023) has passed. The Surviving Sepsis Campaign adult guidelines (2026) “suggest” starting vasopressors peripherally rather than delaying for central access (conditional, very low certainty) and did not consider midline catheters.

EM Evidence Rundown, Issue 35. Page 7 of 28

INFORMING PRACTICE

SR/MA

FRCEM

Oxygen After ROSC (SR/MA, 11 RCTs, n=4,136): Conservative Targets Did Not Reduce Mortality (RR 0.96) but Increased Hypoxaemia (RR 1.65)

Khan MM, Saboor A, Aqil M, et al. Conservative Versus Higher Oxygen Targets Post-Cardiac Arrest: A Systematic Review, Meta-Analysis, and Trial Sequential Analysis. Acad Emerg Med. 2026;33(10):e70422. doi:10.1111/acem.70422. PMID: 42872895. PROSPERO CRD420261423230. Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Meta-analysis of 11 RCTs (12 reports; 4,136 adults resuscitated from OHCA or IHCA; 2,076 restrictive, 2,060 higher oxygen), including BOX, EXACT, LOGICAL and the PILOT cardiac arrest substudy. Restrictive targets included 30% oxygen for 60 minutes, SpO2 88–92% or 90–94%, and PaO2 9–10 kPa (68–75 mmHg); comparators ranged from 100% oxygen to usual care. All-cause mortality at longest follow-up (primary outcome) did not differ (10 trials; RR 0.96, 95% CI 0.88 to 1.06; I² 44%), nor did unfavourable neurological outcome (7 trials; RR 0.96, 95% CI 0.88 to 1.04). Hypoxaemic episodes were more frequent with restrictive targets (8 trials; 549/1,464 vs 339/1,420; RR 1.65, 95% CI 1.41 to 1.92); ventilator-free days rose by 2.14 (95% CI 0.14 to 4.14) in 2 trials. A mortality benefit in the 33–34°C hypothermia subgroup (RR 0.77, 95% CI 0.62 to 0.96) did not survive trial sequential analysis, and under normothermia the evidence trended towards futility.

Outcome (trials) Restrictive Higher oxygen Effect (95% CI) GRADE (authors)

Source: Khan et al, Table 3 (GRADE summary of findings). Event counts are crude pooled totals.

• 50.3% vs 50.2% MORTALITY, RESTRICTIVE VS HIGHER (CRUDE, CALCULATED)
• RR 0.96 MORTALITY (95% CI 0.88 TO 1.06)
• 37.5% vs 23.9% HYPOXAEMIA (CRUDE, CALCULATED)
• NNH 7 HYPOXAEMIA, CALCULATED FROM CRUDE TOTALS (37.5% VS 23.9%); CI NOT REPORTED

CRITICAL APPRAISAL Clear question; protocol registered

Yes PROSPERO CRD420261423230; RCTs only, comatose ventilated adults after ROSC

Design: SR/MA | Certainty: moderate (GRADE-style, our judgement) | What it changes for a UK ED: Nothing new: titrate to the RCUK range and do not aim lower; the extra hypoxaemia with restrictive targets argues against going below it | Confidence in that: moderate

UK practice: Resuscitation Council UK Post-resuscitation care guidelines (2025) (published 27 October 2025): use 100% (or the maximum available) inspired oxygen immediately after ROSC until SpO2 or PaO2 can be measured reliably, then titrate to SpO2 94–98% or PaO2 10–13 kPa (75–100 mmHg), avoiding hypoxaemia (PaO2 under 8 kPa) and hyperoxaemia. RCUK also warns that pulse oximetry can overestimate saturation in people with darker skin tones. Several restrictive arms in this review (SpO2 88–92% or 90–94%, PaO2 9–10 kPa) sat below that range; the review gives no reason to change it.

EM Evidence Rundown, Issue 35. Page 8 of 28

INFORMING PRACTICE DIAGNOSTIC

Absent Miosis Does Not Rule Out Opioids in ED Respiratory Depression: 1 in 4 Opioid-Positive Patients Had Normal or Dilated Pupils (LR− 0.46, n=468)

Greene S, Hodgson SE, Weber C, et al. Absence of miosis does not exclude opioid exposure in emergency department patients with respiratory depression: a diagnostic accuracy study. Emerg Med J. Published online 20 October 2026. doi:10.1136/emermed-2026-216276. PMID: 42901744. Not prospectively registered. Funded by the Victorian Government and NHMRC (APP2001107). Open access (CC BY 4.0). Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Retrospective diagnostic accuracy study from the Emerging Drugs Network of Australia – Victoria registry (14 metropolitan and four regional EDs, September 2020 to December 2025): 468 patients with suspected illicit drug exposure, a pretreatment respiratory rate below 10 breaths/min and a documented pupil examination, all with comprehensive blood toxicology. The index test was clinically documented constricted or pinpoint pupils; the reference standard was any opioid detected in blood, found in 267 (57.1%). Of the 267 opioid-positive patients, 65 (24.3%; 95% CI 19.6 to 29.8) had normal or dilated pupils, giving sensitivity 75.7% (95% CI 70.2 to 80.4), specificity 52.7% (45.8 to 59.5) and LR− 0.46 (0.36 to 0.59). Results were similar at a respiratory rate below 8 (LR− 0.54, 0.37 to 0.79) and with added hypoxaemia or oxygen need (LR− 0.44, 0.31 to 0.63); counting only heroin or above-threshold morphine or codeine as positive gave LR− 0.33 (0.23 to 0.47).

• Sn 75.7% MIOSIS FOR OPIOID (70.2 TO 80.4)
• Sp 52.7% MIOSIS FOR OPIOID (45.8 TO 59.5)
• LR+ 1.60 MIOSIS PRESENT (1.36 TO 1.88)
• LR− 0.46 MIOSIS ABSENT (0.36 TO 0.59)

Post-test probabilities calculated from the reported primary-cohort LRs; the 20% and 80% priors are illustrative, as we found no published UK estimate for this population.

CRITICAL APPRAISAL Representative spectrum, consecutive enrolment

Partly Purposive registry sampling; 115 of 583 with RR below 10 excluded for undocumented pupils or prior medication; bias direction unknown

Design: Diagnostic | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Do not withhold a titrated naloxone trial from a patient with a respiratory rate below 10 and plausible opioid exposure because the pupils are normal or large; treat the respiratory rate | Confidence in that: moderate

Bias direction: counting any detected opioid as positive likely understates how well miosis performs for true opioid toxicity, yet even the restrictive reference standard left LR− at 0.33, which is not a rule-out. Opioid-positive patients without miosis were less often given naloxone (32 of 65 vs 178 of 202) and more often intubated (23 of 65 vs 28 of 202); the authors caution that these comparisons are confounded and not causal.

UK practice: The RCEM and NPIS best practice guideline, Assessment and Management of Acute Opioid Toxicity in Adults in the Emergency Department (April 2024) lists pupillary miosis in the typical triad but states that the diagnosis of an opioid toxidrome is clinical, that naloxone can be a therapeutic trial in respiratory depression with suspected opioid toxicity, and that IV boluses of 100–200 micrograms every 60 seconds should be given until the respiratory rate is above 10. It notes that co-ingested drugs such as crack cocaine may mask the toxidrome. The Naloxone 400 micrograms/ml solution for injection/infusion SmPC (hameln pharma, eMC, text revised February 2026) lists diagnosis of suspected acute opioid overdose or intoxication as a licensed indication; its usual initial overdose dose (0.4–2 mg IV) is larger than the RCEM/NPIS titration. The guideline advises discussion with a clinical toxicologist via NPIS (TOXBASE (NPIS)) when uncertain.

EM Evidence Rundown, Issue 35. Page 9 of 28

INFORMING PRACTICE

DIAGNOSTIC

FRCEM

HINTS Plus in Isolated Acute Vestibular Syndrome (n=298): Sensitivity 80.6% Against CT and MRI, So a “Peripheral” Result Missed 12 of 62 Central Causes

Coşkun A, Şener A, Oğuz O, et al. Diagnostic accuracy of HINTS Plus test in the differential diagnosis of central and peripheral vertigo: a prospective, cross-sectional study. BMC Emerg Med. 2026;26. doi:10.1186/s12873-026-01675-9. PMID: 42399767. Source route: PubMed; PMC full text (PMC13613708, open access). Appraised from full text

Prospective single-centre study in a Turkish stroke-centre ED (Ankara Bilkent City Hospital, July 2023 to July 2024) of 298 adults with acute vestibular syndrome, spontaneous nystagmus and no focal neurological deficit, selected from 5,223 patients with a preliminary diagnosis of vertigo. One of three trained emergency medicine residents performed HINTS Plus (head impulse, nystagmus, test of skew, finger-snap hearing check), with any central component classed as central; every patient then had non-contrast CT and diffusion-weighted MRI, reported blind to the bedside result. Imaging found a central cause in 62 (20.8%), and HINTS Plus had a sensitivity of 80.65% (95% CI 68.63 to 89.58) and specificity of 82.63% (95% CI 77.18 to 87.24). All 12 false negatives were strokes (8 ischaemic, 4 haemorrhagic), and dropping the hearing component changed little (HINTS alone: sensitivity 79.03%, specificity 83.9%).

• Sn 80.6% HINTS PLUS (95% CI 68.6 TO 89.6)
• Sp 82.6% HINTS PLUS (95% CI 77.2 TO 87.2)
• LR+ 4.64 95% CI 3.42 TO 6.29
• LR− 0.23 95% CI 0.14 TO 0.38

Post-test probabilities calculated from the reported LR+ and LR−. The 5% and 10% priors are illustrative, not measured UK prevalences. CRITICAL APPRAISAL Representative spectrum, consecutive enrolment

Partly Described as consecutive, but 298 of 5,223 screened, spontaneous nystagmus required, focal deficits excluded and only three study residents on rotating shifts could enrol; a harder-than-average spectrum, which the authors say lowers sensitivity

Design: Diagnostic | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing new to adopt: a peripheral HINTS Plus does not exclude stroke in isolated acute vestibular syndrome, and a central result should prompt urgent stroke assessment and MRI | Confidence in that: moderate Disagreement: RCEMLearning cites a study in which HINTS, performed by a neuro-ophthalmologist with specialist equipment, was potentially more sensitive than early MRI. This study, with trained EM residents and early MRI as the reference, found a sensitivity of 80.6%. Both positions agree that accuracy depends on who performs the test.

UK practice: NICE NG128 (2019, last updated 2022) asks EDs to use a validated tool such as ROSIER for suspected stroke (recommendation 1.1.3); its recommendations do not mention vertigo, dizziness or HINTS. The RCEMLearning Vertigo module (HiNTs examination, updated February 2026) says HINTS is only valid in continuous, ongoing vertigo, may be less sensitive when non-specialists perform it, and needs further training before wide adoption. A UCLH-sponsored UK study (HRA summary, IRAS 324423) is comparing clinician-performed HINTS-plus with an automated video-goggle version against brain imaging at 24–72 hours.

EM Evidence Rundown, Issue 35. Page 10 of 28

INFORMING PRACTICE

REVIEW

FRCEM

BMJ State of the Art Review on Skin and Soft Tissue Infection: POCUS Detects Abscess With 94.6% Sensitivity, and LRINEC 6 or More Misses About 1 in 3 Necrotising Infections

Long B, Yadav K, Rech MA, et al. Advances in the diagnosis and management of skin and soft tissue infections. BMJ. 2026;394:e100580 (published 10 September 2026). doi:10.1136/bmj-2026-100580. PMID: 42727947. Source route: PDF via BMA Library (supplied by Jake); LRINEC figures checked against the cited meta-analysis abstract (Fernando et al, Ann Surg 2019, PMID: 29672405). Appraised from full text

A commissioned BMJ State of the Art Review by four authors from US and Canadian emergency medicine departments and a US Veterans Affairs hospital (PubMed and Cochrane to December 2025; no formal quality grading). Four points matter most for a UK ED: 1. Many “cellulitis” admissions are mimics. A scoping review of 8 observational studies found that 1 in 3 patients admitted from the ED with suspected cellulitis did not have it (clinical reference standard), and other studies put misdiagnosis at 30% to 39%. Bilateral or chronic redness makes cellulitis highly unlikely. 2. POCUS when abscess vs cellulitis is unclear. A meta-analysis of 14 studies (2,656 participants) gave POCUS a sensitivity of 94.6% (95% CI 89.4 to 97.4) and specificity of 85.4% (78.9 to 90.2) for abscess, with an appropriate change in management in 10.3%. 3. Necrotising soft tissue infection (NSTI) cannot be excluded by a score or a sign. In a meta-analysis of 23 studies (5,982 participants), LRINEC 6 or more had a sensitivity of 68.2% and specificity of 84.8%, fever a sensitivity of 46.0%, and soft tissue gas on plain film a sensitivity of 48.9%. The review, and the Global Alliance for Infections in Surgery it cites, say LRINEC should not be used to rule NSTI out; suspicion means an urgent surgical opinion. 4. Short courses and simpler drainage. One RCT found cellulitis resolved at 14 days in 98% with both 5-day and 10-day courses. For abscess, packing made no difference to recurrence or need for secondary intervention (meta-analysis of 8 RCTs, n=485). Adjunctive antibiotics after drainage cut treatment failure from 16.1% to 7.7% (4 trials, n=2,406; NNT about 12 from these rates, calculated; the review quotes 14) at the cost of more diarrhoea.

• 94.6% POCUS SENSITIVITY FOR ABSCESS (89.4 TO 97.4)
• 85.4% POCUS SPECIFICITY (78.9 TO 90.2)
• 68.2% LRINEC 6 OR MORE SENSITIVITY FOR NSTI
• 98% vs 98% CELLULITIS RESOLVED, 5 VS 10 DAYS

Likelihood ratios and post-test probabilities calculated by us from the pooled sensitivity and specificity; pre-test probabilities are illustrative, not from the review.

CRITICAL APPRAISAL Who wrote it and for whom

Yes Authors from US and Canadian EM departments and a US VA hospital; BMJ says State of the Art Reviews are written predominantly by US authors

Design: Review | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Use POCUS when abscess vs cellulitis is unclear; never let a LRINEC under 6 stop a surgical referral; do not pack small abscesses; keep NG141 flucloxacillin for 5 to 7 days rather than the US regimens | Confidence in that: moderate Disagreement: The review recommends an MRSA-active agent (trimethoprim-sulfamethoxazole, clindamycin or doxycycline) for purulent infection; NICE NG141 adds MRSA cover only when MRSA is suspected or confirmed.

UK practice: NICE NG141 (Cellulitis and erysipelas: antimicrobial prescribing, 2019) recommends flucloxacillin first line orally (500 mg to 1 g four times a day) for 5 to 7 days, with up to 14 days on clinical assessment (rec 1.2.1, table 1). It asks clinicians to refer anyone with signs of necrotising fasciitis or sepsis (1.1.13), and says nothing on POCUS, LRINEC or drainage technique. UKHSA (originally HPA/Department of Health) guidance on PVL-associated Staphylococcus aureus infections in England (2nd edition, 2008) says small abscesses without cellulitis need drainage, not systemic antibiotics, unless the patient is immunocompromised, an infant or deteriorating.

EM Evidence Rundown, Issue 35. Page 11 of 28

INFORMING PRACTICE OBSERVATIONAL

Sternal Fracture (n=245): Death in 20% With Initial hs-Troponin T Above 17.5 ng/L vs 2% Below, in Unadjusted Single-Centre Data

Yeoh C, Ellis DY, Foo N, et al. Do ECG abnormalities, degree of sternal displacement or troponin level influence outcomes following sternal fracture? Injury. 2026 (available online 2 September 2026);113650. doi:10.1016/j.injury.2026.113650. PMID: 42731895. Source route: PDF via BMA Library (ClinicalKey), supplied by Jake. Appraised from full text

Retrospective audit of 245 patients with radiologically confirmed sternal fracture after blunt trauma at the Royal Adelaide Hospital (November 2020 to March 2024; median age 56, median ISS 18); 22 (9%) died in hospital. Mortality was 20% (17/84) with an initial high-sensitivity troponin T above 17.5 ng/L vs 2% (3/143) at or below it, a cut-off chosen after the fact by Youden index (AUC 0.830, 95% CI 0.746 to 0.915). Mortality was 36% (4/11) with moderate or complete displacement vs 7.7% (18/233) with none or mild, and rose from 1% (1/119) with sinus rhythm to 33% (3/9) with a conduction abnormality. All comparisons are univariable; with 22 deaths the authors did not attempt adjustment.

• 20% MORTALITY, TROPONIN T ABOVE 17.5 NG/L (17/84)
• 2% MORTALITY, TROPONIN T 17.5 OR BELOW (3/143)
• 36% MORTALITY, MODERATE OR COMPLETE DISPLACEMENT (4/11)
• 1% MORTALITY, SINUS RHYTHM (1/119)

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Displacement graded from descriptive CT reports mapped to percentages after the fact; ECGs read by treating consultants; troponin missing in 18 and ECG in 10

Design: Observational | Certainty: very low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing yet; an ECG and troponin after sternal fracture remain reasonable, but do not adopt 17.5 ng/L as a threshold | Confidence in that: moderate

UK practice: NICE NG39 (major trauma: assessment and initial management, 2016) recommends urgent imaging for suspected chest trauma and considering immediate CT in adults without severe respiratory compromise (recs 1.4.3 and 1.4.5); its recommendations do not mention ECG, troponin, blunt cardiac injury or sternal fracture. We found no NICE or RCEM guidance on troponin or ECG thresholds after sternal fracture, so the work-up follows local trauma protocols.

EM Evidence Rundown, Issue 35. Page 12 of 28

INFORMING PRACTICE OBSERVATIONAL

Opioid-Naive Patients Offered an Opioid for Acute Pain (US Cohort, n=1,708): Median 7 Days of Use at 10 MME a Day, 66.9% With Leftovers, but 10% Still Using at 90 Days

Jeffery MM, Bellolio F, Chang N, et al. Opioid Use and Pain Resolution for Acute Pain Among Opioid-Naive Patients. JAMA Netw Open. 2026;9(7):e2621875. doi:10.1001/jamanetworkopen.2026.21875. PMID: 42412431. Source route: PubMed; PMC full text (PMC13343240). Appraised from full text

Prospective cohort study (the FDA-funded Acute Pain Pathways study) of 1,708 opioid-naive patients offered an outpatient opioid prescription for acute pain at 5 US health systems between September 2020 and March 2023, 915 (53.6%) of them recruited in EDs. Patients reported pain and treatment on digital surveys for 180 days; there was no comparison group and all analyses were descriptive. Median time to pain resolution (3 consecutive reports of no pain) was 20 days (IQR 8–88), but 69 days for low back pain (IQR 18 days to not reached) and 74 days for postsurgical pain. Among 1,189 patients who used an opioid, median time to discontinuation was 7 days (IQR 2–31), the median daily dose was 10 MME (IQR 5–15) in those who reported it, an estimated 10.0% (95% CI 7.7% to 12.7%) used opioids for at least 90 days, and 657 of 982 respondents (66.9%) had leftover opioids.

• 7 days MEDIAN OPIOID USE (IQR 2 TO 31)
• 10.0% OPIOID USE 90 DAYS OR MORE (95% CI 7.7 TO 12.7)
• 66.9% LEFTOVER OPIOIDS (657/982)
• 69 days MEDIAN TO RESOLUTION, LOW BACK PAIN

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Self-reported surveys, because linkage to EHR and pharmacy portals proved difficult; “no pain” covered pain from any source and almost a third had pain most days or every day before enrolment, so time to resolution is likely overstated

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing new to adopt; it supports short discharge opioid supplies and telling patients with low back pain that recovery often takes weeks | Confidence in that: moderate

UK practice: RCEM Best Practice Guideline, Management of Acute Pain in Adults in the ED (September 2024) says to consider opioids only when benefit is expected to outweigh risk, to avoid modified-release and compound preparations at discharge, to explain that this is usually a single course, and to limit discharge opioid prescriptions ideally to 3–5 days (up to 7 in specific circumstances). The Faculty of Pain Medicine’s Opioids Aware: opioids and acute pain management (undated web page) says opioids should be part of a multimodal approach, last no longer than pain severe enough to need them, and that the discharge letter should state the dose, amount supplied and planned duration. NICE NG59 (2016, last updated July 2026) says do not routinely offer opioids for acute low back pain (recommendation 1.2.27). NICE NG193 (2021) covers chronic pain lasting more than 3 months and does not apply to acute ED prescribing.

EM Evidence Rundown, Issue 35. Page 13 of 28

2 — GUIDELINES & UK UPDATES

CHANGE TONIGHT

UK

SAFETY ALERT

Abbott i-STAT hs-TnI Cartridges: Field Safety Notice Warns of Falsely Raised Troponin From Interference in Lots Beginning A24, A25, B25, B26, C26 or D26

Abbott Point of Care. Field Safety Corrective Action APOC2026-003: i-STAT hs-TnI cartridge (list number 09P81-25), October 2026. Listed by MHRA in Field Safety Notices 12 to 16 October 2026 (published 19 October 2026), MHRA reference 2026/010/013/601/143. gov.uk; notice (PDF). Source route: MHRA field safety notices list. Appraised from full text

Abbott reports that i-STAT hs-TnI cartridge lots beginning A24, A25, B25, B26, C26 and D26 may be less able to cancel out macrotroponin and heterophile antibody interference, which can give falsely raised troponin I; there is no evidence of falsely low results. Abbott has received 79 reports (about 25 per million cartridges) and 6 reports of unnecessary intervention, cardiac catheterisation without true myocardial injury (about 2 per million). Cartridges can still be used. If a raised result does not fit the clinical picture, retest on a lot beginning A26 (available from mid-October) or with a validated laboratory assay; serial i-STAT results may not separate interference from true elevation.

CRITICAL APPRAISAL Issuer and basis

Yes Manufacturer field safety corrective action, listed by MHRA

Design: Safety alert | Certainty: moderate (GRADE-style, our judgement) | What it changes for a UK ED: If your ED or a referring service uses i-STAT hs-TnI, check the lots and tell clinicians to send a laboratory troponin when a raised point-of-care result does not fit | Confidence in that: high

UK practice: MHRA: “If you receive a field safety notice (FSN) from a manufacturer you must always act on it” (Field Safety Notices, 12 to 16 October 2026). gov.uk. The PRESTO prehospital study in this issue’s Quick Hits used the same i-STAT hs-TnI cartridge.

EM Evidence Rundown, Issue 35. Page 14 of 28

INFORMING PRACTICE

GUIDELINE

UK

NICE Draft on Sepsis in Pregnancy: MEWS Replaces Table-Based Risk Criteria, IV Antibiotics Within 1 Hour at MEWS 8 or More; Comments Close 11 November

National Institute for Health and Care Excellence. Suspected sepsis in pregnant or recently pregnant people: recognition, diagnosis and early management. Draft guideline for consultation (GID-NG10468), 14 October 2026; expected publication 21 January 2027. nice.org.uk/guidance/gid-ng10468. Source route: NICE website check (guidance in consultation). Appraised from full text

The draft aligns the pregnancy sepsis guideline with the national Maternity Early Warning Score (MEWS) in every setting: a score of 8 or more suggests high risk, 5 to 7 medium, 2 to 4 low-medium and 0 to 1 low, and listed concerns (for example severe hypertension of 160/110 mmHg or more, significant vaginal bleeding or reduced urine output) can move a person up a category. At high risk it asks for immediate senior review, a venous gas with lactate plus the sepsis blood panel, broad-spectrum IV antibiotics within 1 hour of the MEWS score on initial ED assessment, and a 250 ml crystalloid bolus (balanced, such as Hartmann’s, preferred) over 10 to 15 minutes, repeated up to 1,000 ml before senior advice. Medium risk with lactate over 2 mmol/L or acute kidney injury is managed as high risk; otherwise antibiotics may be deferred for up to 3 hours (medium) or 6 hours (low-medium) while a diagnosis is sought. The committee made no recommendation on procalcitonin, finding no direct evidence in pregnancy, and asks for a multidisciplinary decision including critical care expertise before peripheral vasopressors.

CRITICAL APPRAISAL Developer and funding

Yes NICE; committee membership and register of interests published with the consultation

Design: Guideline | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing yet for practice; check that ED triage and sepsis screening tools can calculate MEWS for pregnant and recently pregnant patients, and consider a departmental response by 11 November | Confidence in that: moderate

UK practice: Final guideline expected 21 January 2027 (NICE, GID-NG10468). Until then NICE NG255 (suspected sepsis in pregnant or recently pregnant people, November 2025) applies. Its recommendation 1.2.1 treats people who are pregnant, or who have given birth or had a termination or miscarriage in the past 6 weeks, as a high-risk group. For non-pregnant adults, NICE NG253 (updated September 2026, covered in Issue 34) added consider procalcitonin; this draft does not extend that to pregnancy. Comments are by the NICE comments form, through a registered stakeholder organisation or as an individual.

EM Evidence Rundown, Issue 35. Page 15 of 28

ALSO FROM UK BODIES THIS WEEK

MHRA Drug Safety Update: New Boxed Suicide-Risk Warning and Patient Card for All Antidepressants; “PSSD” Wording Added for 10 Drugs

Medicines and Healthcare products Regulatory Agency. Antidepressants: new safety measures for all patients to improve communication of the risk of suicidal thoughts and behaviours and persistent sexual dysfunction. Drug Safety Update volume 20, issue 3, October 2026: 1. Published 15 October 2026. gov.uk. Source route: MHRA Drug Safety Update feed.

A Commission on Human Medicines expert working group (May 2024 to July 2025) reviewed how patient leaflets communicate the risks of suicidal thoughts and behaviours and of sexual dysfunction that can persist after stopping treatment. Leaflets will carry a new boxed warning on suicidal thoughts with signposting to help, a patient card (digital first, then credit-card sized), and the term “post SSRI sexual dysfunction (PSSD)” for citalopram, escitalopram, duloxetine, fluvoxamine, fluoxetine, paroxetine, sertraline, venlafaxine, amitriptyline and nortriptyline; a fuller safety leaflet follows in 2027. MHRA asks prescribers to discuss these risks when starting or changing the dose, and to tell patients to seek urgent help for new or worsening suicidal thoughts, especially early in treatment or after a dose change. MHRA stresses these are not new risks. For a UK ED: Ask about a recent antidepressant start or dose change in anyone presenting with new suicidal thoughts, and report suspected reactions on a Yellow Card.

UKHSA: Bundibugyo Ebola Now Imported Into Kenya (9 Cases) and Nigeria (1 Case); 9,552 Confirmed Cases in the DRC; UK Risk Still Low

UK Health Security Agency. Outbreaks under monitoring: week 42 (week ending 16 October 2026), updated 22 October 2026. gov.uk. Also: High consequence infectious disease: country specific risk, updated 22 October 2026. gov.uk. Source route: gov.uk UKHSA publications check.

As of 18 October 2026 the DRC had reported 9,552 confirmed cases and 4,602 confirmed deaths in 66 health zones across 8 provinces. Kenya confirmed its first imported case on 6 October (a traveller from the DRC via Kampala) and had 9 confirmed cases by 20 October, all linked to that case and found in quarantine. On 20 October Nigeria reported its first confirmed case, in Abuja, in a woman who had recently travelled from Uganda, with 44 contacts being followed up. No imported UK cases have been reported, and UKHSA assesses the risk to the UK population as low; its country-risk table now lists Ebola in Kenya and Nigeria as limited local transmission associated with a case import. For a UK ED: Ask every febrile returning traveller about the DRC, Uganda, Kenya and Nigeria in the past 21 days, isolate, test for malaria urgently and discuss with infection specialists per the ACDP algorithm.

RCEM Winter Survey: 91% of ED Clinical Leads Very or Extremely Concerned; 60% in England Describe Their ED as Actively Unsafe

Royal College of Emergency Medicine. “Perfect conditions for a terrible winter” for NHS as 90% of top A&E doctors worried about winter pressures. Press release, 20 October 2026. rcem.ac.uk. Source route: RCEM news feed.

RCEM surveyed ED clinical leads across the UK from 30 September to 2 October 2026; 163 responded. 91% were very or extremely concerned about winter pressure, only 3% said they were well prepared, 92% rated current pressure high or very high and 88% said strain was greater than last year. Of the 119 respondents in England, 97% said their ED was not well prepared for winter and 60% described their ED as actively unsafe. RCEM says the £150m (DHSC gives almost £140 million) for 14 urgent care schemes announced the weekend before is still not being spent in the right places and calls for better staffing, seven-day working, more beds and investment in discharge. For a UK ED: Nothing new clinically; useful for departmental winter escalation and corridor-care risk discussions.

EM Evidence Rundown, Issue 35. Page 16 of 28

3 — PAEDIATRIC EMERGENCY MEDICINE

INFORMING PRACTICE

OBSERVATIONAL

PAEDS

FRCEM

A First ED Temperature of 36°C or Less in Children (5.3 Million US Encounters): Adjusted Odds of In-Hospital Death 20 Times Higher, Absolute Mortality About 0.5%

Campbell L, Kottapalli A, Hueschen L, et al. Hypothermia and Mortality Among Children in the Emergency Department. Pediatrics. 2026;157(6):e2025075725. doi:10.1542/peds.2025-075725. PMID: 42156016. Source route: scanned PDF via BMA Library inter-library loan (supplied by Jake), read from the page images. Appraised from full text

Retrospective cohort of 5,308,892 ED encounters in children under 18 in the US PECARN registry (12 EDs, 9 of them quaternary children’s hospitals, 2012 to 2021). The exposure was the first recorded ED temperature, grouped as 36°C or less (3.0% of encounters), euthermia (the reference) or 38°C or more; the outcome was in-hospital death, which occurred in 1,918 encounters (0.04%). Mortality was 752 of 161,111 (0.47%, calculated) with a first temperature of 36°C or less vs 910 of 4,432,415 (0.021%, calculated) with euthermia: adjusted OR 20.15 (95% CI 18.00 to 22.55), falling to 8.63 (95% CI 7.37 to 10.12) in the complete-case analysis and to 10.85 (95% CI 9.47 to 12.44) when deaths in the ED were excluded. The association was strongest in infants under 90 days (adjusted OR 82.87, 95% CI 61.13 to 112.34).

• 0.47% MORTALITY, FIRST TEMP 36°C OR LESS (CALCULATED)
• 0.021% MORTALITY, EUTHERMIC (CALCULATED)
• aOR 20.15 95% CI 18.00 TO 22.55 (IMPUTED)
• aOR 8.63 COMPLETE CASE, 95% CI 7.37 TO 10.12

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Outcome from hospital discharge status; thermometer route was not recorded for most visits, so rectal, axillary and tympanic readings are mixed

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing new to adopt; make sure triage treats a temperature under 36°C in a child as a sepsis risk criterion that needs senior review, not as “afebrile” | Confidence in that: moderate

Thresholds and a reporting error: the adjusted OR rose from 7.44 at under 36.5°C to 152.90 at under 35.0°C (Figure 2, journal p8), so the lower the reading, the stronger the signal. Table 3 prints identical univariable and multivariable ORs for both triage (ESI) strata, which looks like a copy error; we have not quoted them.

UK practice: NICE NG254 (suspected sepsis in under 16s, November 2025) lists a temperature under 36°C as a high-risk criterion at any age under 5, and a tympanic temperature under 36°C as a moderate to high risk criterion at 5 to 11 and 12 to 15 years. It also says not to rely on fever or hypothermia alone to rule sepsis in or out. NG254 has replaced NICE NG51 for under-16s. NICE NG143 (fever in under 5s, 2019, updated 2021) advises against routine oral and rectal measurement under 5 (electronic axillary under 4 weeks; electronic or chemical dot axillary, or infrared tympanic, from 4 weeks to 5 years), so UK readings are not directly comparable with this mixed-route US dataset. The study cut-off (36°C or less) differs slightly from NICE (under 36°C).

EM Evidence Rundown, Issue 35. Page 17 of 28

INFORMING PRACTICE

DIAGNOSTIC

PAEDS

Two Clinicians, One Child With Pneumonia (PECARN, 252 Paired Examinations): No Examination Sign Reached Acceptable Reliability; Crackles Had κ 0.21 and Reduced Breath Sounds κ 0.11

Hooli S, Reeder R, Cutler L, et al; PECARN PedCAPS Investigators. Reliability of Physical Examination Findings in Youths Diagnosed With Pneumonia. JAMA Netw Open. 2026;9(8):e2630074. doi:10.1001/jamanetworkopen.2026.30074. PMID: 42623068. Source route: PDF supplied by Jake (18 October 2026; open access, CC BY-NC-ND). Appraised from full text

Planned inter-rater analysis within the prospective PECARN PedCAPS cohort at 7 US academic paediatric EDs (August 2023 to May 2025). Two ED clinicians independently examined each of 252 children aged 3 months to 17 years diagnosed with community-acquired pneumonia, within 60 minutes of each other, and the primary measure was Fleiss κ for each sign, with a lower 95% confidence limit of 0.4 or more prespecified as acceptable. No sign met that threshold: wheeze (κ 0.50, 95% CI 0.39 to 0.62) and retractions or accessory muscle use (κ 0.49, 95% CI 0.37 to 0.60) were the most reliable, while fine crackles (κ 0.21, 95% CI 0.11 to 0.31), rhonchi (κ 0.18, 95% CI 0.08 to 0.28) and decreased breath sounds (κ 0.11, 95% CI 0.02 to 0.20) agreed little beyond chance. Reliability was similar in the 124 children discharged and the 128 admitted.

• 55% RAW AGREEMENT, FINE CRACKLES
• 45% DISAGREEMENT, FINE CRACKLES (CALCULATED)
• κ 0.21 CRACKLES (95% CI 0.11 TO 0.31): FAIR
• κ 0.11 DECREASED BREATH SOUNDS (95% CI 0.02 TO 0.20): SLIGHT

Source: Figure 1 of the paper (selected signs). Each examiner’s assessment is one observation. Landis and Koch: 0.21 to 0.40 fair, 0.41 to 0.60 moderate.

At the bedside, two clinicians disagreed about crackles in about 45 and about decreased breath sounds in about 54 of every 100 children (calculated from raw agreement). There is no reference standard, so no likelihood ratios.

CRITICAL APPRAISAL Representative spectrum, consecutive enrolment

Partly Subset with a second examiner available; more White children (56.7% vs 46.5%) and more admissions (47.2% vs 37.2%) than the rest of the cohort

Design: Diagnostic | Certainty: moderate (GRADE-style, our judgement) | What it changes for a UK ED: Do not base a pneumonia diagnosis or antibiotic decision on crackles or reduced breath sounds alone; weigh fever, respiratory rate and work of breathing, and document exactly what you heard and where | Confidence in that: moderate Disagreement: NICE NG143 (fever in under 5s, 2019, updated 2021) lists “crackles in the chest” as an amber (intermediate risk) feature in febrile under-5s. In this study two clinicians examining the same child agreed on crackles only slightly better than chance (κ 0.21), so the same child could be graded amber by one doctor and not by the next. Note that the study population is older (median 5.7 years) and all had a CAP diagnosis.

UK practice: The BTS guideline on community acquired pneumonia in children (2011, Thorax 66 Suppl 2) says bacterial pneumonia should be considered with persistent or repetitive fever over 38.5°C together with chest recession and a raised respiratory rate [D]. It says chest radiography should not be routine in children thought to have CAP [A], and children not admitted should not have one [A]. NICE NG250 (pneumonia, September 2025) covers children from 1 month, defines a clinical diagnosis as one made from symptoms and signs without a chest X-ray, and has no child-specific chest X-ray recommendation. So in the UK most childhood CAP is diagnosed clinically, and this paper says auscultation is the least reproducible part of that.

EM Evidence Rundown, Issue 35. Page 18 of 28

INFORMING PRACTICE

OBSERVATIONAL

PAEDS

Manual Detorsion by Paediatric Emergency Physicians (n=256, Single US Centre): Successful in 65% of Attempts, and Successful Attempts Reached Blood Flow About 105 Minutes Before Incision With Surgery Alone (About 73 Minutes Across All Attempts)

Rivera T, Lozano J, Maniaci V, et al. Manual Detorsion by Pediatric Emergency Physicians Shortens Ischemia Time in Pediatric Testicular Torsion. Acad Emerg Med. 2026;33(9):e70412. doi:10.1111/acem.70412. PMID: 42740594. Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Retrospective chart review at a US tertiary paediatric centre (January 2013 to January 2025) of 256 patients aged 8 to 21 years with ultrasound-diagnosed testicular torsion. Paediatric emergency physicians attempted manual detorsion in the radiology suite in 141 patients; it succeeded (flow on repeat ultrasound) in 91 of 141 (65%), and 115 went straight to theatre. The primary outcome, time from the start of the first ultrasound to restored flow (successful detorsion) or skin incision (surgery), was a mean of 35.2 vs 140.3 minutes, a difference of 105.1 minutes (95% CI 94.9 to 115.3); a failed attempt did not delay incision (mean 125.3 vs 140.3 minutes, P=0.07; median 114 vs 127 minutes, P=0.048). Testes were saved in 91 of 91 (100%) after successful detorsion, 49 of 50 (98.0%) after a failed attempt and 103 of 115 (89.6%) with surgery alone, but pain had lasted much longer in the surgery-only group (mean 37.4 vs 6.1 hours).

• 35.2 min TO RESTORED FLOW, SUCCESSFUL MANUAL DETORSION
• 140.3 min TO INCISION, STRAIGHT TO SURGERY
• 105.1 min MEAN DIFFERENCE (95% CI 94.9 TO 115.3)
• 65% MANUAL DETORSION SUCCESS RATE

Source: Tables 1 and 2 of the paper. Time zero was the start of the first ultrasound, so total ischaemia time is underestimated in every group.

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Second blinded reviewer; ICC 0.94 (95% CI 0.90 to 0.98) on 50 records for the primary outcome. End points differ: incision comes before surgical untwisting, which shortens the surgical time and biases the difference towards the null

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing yet for pathways; if theatre will be delayed (transfer, no surgeon on site), manual untwisting is a reasonable temporising step that must not delay exploration | Confidence in that: moderate Disagreement: The paper’s pathway is ultrasound first, with detorsion by the ED physician in the scanning room, and the authors suggest community clinicians with ultrasound should try it. GIRFT (2026) puts clinical assessment and the TWIST score first, with exploration within 1 hour of the decision and no delay for ultrasound. In a UK DGH where the scan is the bottleneck, copying the US model could add delay rather than remove it.

UK practice: The GIRFT children and young people testicular torsion pathway (version 2.1, May 2026) asks for review by a surgical decision maker within 60 minutes of ED arrival. It says a TWIST score of 5 or more with under 48 hours of pain mandates exploration, and asks for theatre within 1 hour of the decision. It says surgery should not be significantly delayed for an ultrasound, and it does not mention manual detorsion. The RCEM Learning module on the acute scrotum (2021) describes manual untwisting away from the midline (“opening a book”) as a temporising measure whilst awaiting theatre or where surgery is unavailable, which “does not negate the need for surgery, nor should it delay operating.”

EM Evidence Rundown, Issue 35. Page 19 of 28

INFORMING PRACTICE

OBSERVATIONAL

UK

PAEDS

Corridor Care for Children in 96 UK EDs (UNCORKED, Preprint Data): 15.4% of Admitted Children Had Escalation Area Care, and Half of Children in Escalation Areas Were in Corridors or Other Non-Clinical Spaces

Owens DR, Birse F, Anakebe C, et al; TERN UNCORKED study group. Paediatric escalation area and corridor care in the emergency department: a secondary analysis of the UNCORKED study. Arch Dis Child. Published online 11 October 2026 (letter). doi:10.1136/archdischild-2026-330704. PMID: 42860270. ISRCTN16396025. The published letter has no abstract and we could not read it; all numbers here come from the medRxiv preprint, which may differ from the published letter: Anakebe C, Owens DR, Roland D, et al. medRxiv 6 September 2026 (not peer reviewed, CC BY). doi:10.64898/2026.09.02.26362046. Source route: PubMed sweep; medRxiv full text; RCEM press release 13 October 2026. Abstract only

Unplanned paediatric secondary analysis of UNCORKED, a prospective snapshot study by the Trainee Emergency Research Network in 96 UK type 1 EDs that see children (83 mixed, 13 paediatric-only), at five time points between 3 and 12 March 2025. Children under 16 who were admitted or spent any time in an escalation area were followed to 28 days; the primary outcome was the proportion of children in escalation areas in paediatric-only EDs, because mixed EDs had no paediatric denominator. In paediatric-only EDs, 75 of 1,132 children present (6.6%) were in an escalation area at a snapshot, mostly in a minority of departments. Across all sites, 104 of 675 admitted children (15.4%) had escalation area care during their ED stay, 124 of 244 children in escalation areas (50.8%) were in a non-clinical space such as a corridor, and 8 of 83 mixed EDs (9.6%) reported adults being cared for in paediatric areas.

• 15.4% ADMITTED CHILDREN WITH ESCALATION AREA CARE
• 50.8% OF THOSE IN ESCALATION AREAS WERE IN A NON-CLINICAL SPACE
• 6.6% OF CHILDREN PRESENT IN PAEDIATRIC-ONLY EDS
• 9.6% MIXED EDS PLACING ADULTS IN PAEDIATRIC AREAS

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Sites were given a working definition (“any area not routinely used unless the capacity of the usual ED geographical footprint is exceeded”) and categorised their own areas; no agreed definition existed at the time

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Count how many children in your ED are in non-clinical spaces at your busiest times and escalate it; national guidance already says this should not happen | Confidence in that: moderate

UK practice: NHS England’s Principles for providing patient care in corridors (replaces the September 2024 principles; corridor care definition released 6 March 2026) says corridor care is never acceptable and must be avoided for listed patient groups, including children. The RCEM position statement (December 2024) says corridor care is never acceptable for children or adults. RCEM’s press release on the letter (13 October 2026) describes “around one in seven” children admitted via A&E experiencing corridor care, which may reflect the letter’s own figure rather than the preprint’s 15.4%. The RCPCH Facing the Future standards (5th edition, October 2025) include a chapter on the paediatric emergency care environment.

EM Evidence Rundown, Issue 35. Page 20 of 28

4 — FOAMED & CRITICAL APPRAISAL

INFORMING PRACTICE

REVIEW

UK

The SSC 2026 Dissent (Spiegel, Weingart, Farkas and Colleagues): Drop the 30 mL/kg Default, Stop Chasing Lactate, Stratify Antibiotic Timing; NICE NG253 Already Does Much of This

Spiegel R, Haycock K, Miller A, et al. The 2026 Surviving Sepsis Campaign Guidelines: The dissenting opinion. Am J Emerg Med. 2026;110:411–415. doi:10.1016/j.ajem.2026.09.025. PMID: 42767105. Source route: JournalFeed EM Speed Read 22 October 2026; PDF via BMA Library (supplied by Jake). Appraised from full text

A five-page commentary by seven emergency and critical care physicians (US, Canada and one UK NHS intensivist) rebutting the most influential 2026 Surviving Sepsis Campaign recommendations. They argue that at least 30 mL/kg of crystalloid in 3 hours rests on confounded observational data, citing CLOVERS, CLASSIC and ARISE-Fluid (25 vs 38 mL/kg, no mortality difference); that lactate is a prognostic marker and should prompt a review of antibiotics and source control rather than more fluid; and that dynamic fluid-responsiveness tests ignore fluid tolerance and point-of-care ultrasound. They also argue that MAP is a dependent variable to be read with capillary refill and mottling, and that a 1-hour antibiotic target for sepsis without shock is unsupported, citing two RCTs in which antibiotics given 93 and 73 minutes earlier did not reduce mortality.

Sources: SCCM SSC 2026 recommendation list, NICE NG253 recommendations 1.8.x and the dissent, all fetched or read 23 October 2026.

CRITICAL APPRAISAL Who wrote it and for whom

Yes EM and critical care authors including Scott Weingart and Josh Farkas (PulmCrit); written for bedside clinicians; no competing interests declared, but several authors have published the same critique of earlier SSC versions

Design: Review | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Nothing to change if you follow NG253: it already uses 250 ml boluses with reassessment and risk-tiered antibiotic timing; use the piece to explain why UK practice differs from SSC | Confidence in that: moderate Disagreement: SSC 2026 keeps at least 30 mL/kg in 3 hours, serial lactate and a 1-hour antibiotic target for probable sepsis without shock; the dissent rejects all three. The dissent does not discuss vasopressin timing, which this issue’s VASSPR item addresses.

UK practice: UK EDs follow NICE NG253 (2025, updated September 2026), not SSC. NG253 stratifies by NEWS2, gives 250 ml fluid boluses up to 1,000 ml with reassessment after each (1.8.6 to 1.8.9), says lactate, mottled skin and capillary refill should not guide treatment in isolation, and allows clinicians to consider deferring antibiotics for up to 3 hours at medium risk and 6 hours at low risk, while giving them within 1 hour at high risk (1.8.3, 1.8.17, 1.8.23). On those points NICE is closer to the dissent than to the Surviving Sepsis Campaign adult guidelines (2026).

EM Evidence Rundown, Issue 35. Page 21 of 28

INFORMING PRACTICE FOAMED

REBEL EM on CRAO: Two Small RCTs Show No Significant Visual Benefit From IV Thrombolysis (TenCRAOS 20% vs 24% Vision Recovery With Aspirin)

DiBello J. Clinical Conundrum: What is the Role of Thrombolysis in Central Retinal Artery Occlusion? REBEL EM, 19 October 2026. rebelem.com/crao. Trials: Ryan SJ, et al. TenCRAOS. N Engl J Med. 2026;394(5):442–450. doi:10.1056/NEJMoa2508515. PMID: 41604638. Préterre C, et al. THEIA. Lancet Neurol. 2025;24(11):909–919. doi:10.1016/S1474-4422(25)00308-4. PMID: 41109232. Source route: REBEL EM RSS feed. Appraised from full text

REBEL EM concludes that routine IV thrombolysis is not supported for acute non-arteritic CRAO, despite the 2021 AHA statement that suggested considering it within 4.5 hours on the basis of observational data. In TenCRAOS (78 patients, 16 sites, six countries), vision recovery at 30 days occurred in 20% with tenecteplase 0.25 mg/kg vs 24% with aspirin 300 mg (risk difference −3.7 percentage points, 95% CI −22.0 to 14.7), with more adverse events and one fatal intracranial haemorrhage after tenecteplase. In THEIA (70 patients, France), alteplase did not significantly improve visual acuity compared with aspirin; the trial was likely underpowered. The post recommends rapid diagnosis, ophthalmology and neurology input, stroke work-up and secondary prevention instead.

CRITICAL APPRAISAL Who wrote it and for whom

Yes REBEL EM author, peer reviewed by two named reviewers, for ED clinicians

Design: FOAMed | Certainty: low (GRADE-style, our judgement) | What it changes for a UK ED: Do not offer IV thrombolysis for CRAO outside a trial; treat it as a stroke equivalent and refer urgently to ophthalmology and the stroke team | Confidence in that: moderate Disagreement: The 2021 AHA scientific statement (as described by REBEL EM) suggested considering thrombolysis within 4.5 hours in selected patients; REBEL EM now says the two RCTs do not support routine use.

UK practice: We found no NICE, RCEM or Royal College of Ophthalmologists guidance on thrombolysis for CRAO (searched nice.org.uk, rcem.ac.uk and rcophth.ac.uk on 23 October 2026). Follow local stroke and eye pathways.

EM Evidence Rundown, Issue 35. Page 22 of 28

ALSO WORTH A LISTEN OR A READ

EM Cases Ep 224: A Four-Step Approach to Episodic Vertigo

Lin K, Johns P, Targonsky E, Helman A. Vertigo Part 1: Approach to Episodic Vertigo. Emergency Medicine Cases, Episode 224, October 2026 (posted 20 October 2026). emergencymedicinecases.com/episodic-vertigo. Key reference: Edlow JA, et al. GRACE-3. Acad Emerg Med. 2023;30:442–486. Source route: EM Cases RSS feed.

The episode gives a four-step approach: screen for central and systemic red flags (including new inability to sit, stand or walk unaided, and nystagmus that is vertical or changes direction at rest); classify the syndrome by timing and triggers using the GRACE-3 framework; choose the test that matches the syndrome; then treat and decide disposition. For brief, position-triggered episodes without nystagmus at rest, perform a Dix-Hallpike, waiting 30 to 60 seconds before calling it negative, and treat posterior canal BPPV with an Epley manoeuvre rather than vestibular suppressants. The experts warn that HINTS and HINTS Plus are for continuous vertigo with spontaneous nystagmus and should not be used for episodic positional vertigo; worsening with movement does not distinguish peripheral from central causes. Routine neuroimaging is not advised for typical BPPV without red flags. For a UK ED: Use Dix-Hallpike and Epley for triggered episodic vertigo, and keep HINTS for continuous vertigo with nystagmus.

EMCrit 436: The Haemodynamically Neutral Intubation for the Shocked Patient Who Is Likely to Arrest With RSI

Weingart S. EMCrit 436: When you RSI shocked patients, they tend to die (the hemodynamically neutral intubation part 2). EMCrit, 18 October 2026. emcrit.org. Source route: EMCrit RSS feed.

Weingart describes the haemodynamically neutral intubation: keep the patient breathing spontaneously, because switching to positive-pressure ventilation reduces venous return, and avoid paralysis unless it is unavoidable. Candidates are right heart failure or pulmonary hypertension, tamponade, and patients on high-dose or multiple pressors with a MAP still in the 50s; those whose pressure corrects easily get ketamine-first delayed sequence intubation instead. The approach uses an arterial line, PEEP of 5 to 8 for cardiogenic shock or shunt (none for exsanguinating trauma), 3 minutes of preoxygenation, ketamine 20 to 30 mg then 20 mg aliquots until dissociated, video laryngoscopy with a bougie, and rocuronium with sugammadex ready. After intubation he advises a pressure-support mode at the pre-intubation PEEP and small ketamine pushes, not standard sedation infusions. For a UK ED: Nothing yet as protocol; a reason to identify the very high-risk shocked airway early and plan with senior anaesthetic and ICU colleagues.

EM Evidence Rundown, Issue 35. Page 23 of 28

5 — QUICK HITS

Prehospital 0/1-hour point-of-care hs-troponin (PRESTO, 4 UK ambulance services, n=642): derived i-STAT hs-cTnI cut-offs ruled out type 1 MI in 51.1% of patients with sensitivity 98.8% (95% CI 93.2% to 100.0%); the cut-offs need validation in whole blood drawn by paramedics. Appraised in full in the PHEM Evidence Rundown. Body R, et al. Emerg Med J 2026 (online 20 October). PMID: 42901741

Emergency physician awake intubation (single-centre registry, 10 years): 87 of 1,213 ED intubations (7.2%) were awake; success without critical hypoxaemia, hypotension or arrest in 92% (95% CI 86% to 98%), first-pass success 71%; observational. Parks A, et al. Ann Emerg Med 2026;88(5):621–627. PMID: 42405912

Severe methaemoglobinaemia in Mediterranean sea migrants (Italy, n=82, 2024–2025): median methaemoglobin 44.6% at presentation, attributed to skin absorption of the petrol additive N-methylaniline from fuel-contaminated seawater; 90% received methylene blue, 38 developed delayed haemolytic anaemia and 2 died. Relevant to EDs receiving people arriving by small boat; case series. Bernasconi L, et al. N Engl J Med 2026;395(10):983–991. PMID: 42715562

Anaphylaxis in children under 2 (paediatric ED, retrospective, n=155): 96.1% were food-triggered, skin or mucosal signs were present in 99.4% and cardiovascular features in 5.8%; 87.1% received IM adrenaline (90.4% of them one dose) and 88.4% were discharged home; observational. Shannon C, et al. Pediatr Emerg Care 2026. PMID: 42865096

Point-of-care hs-troponin (systematic review, 11 studies, n=10,115; abstract only): single-sample rule-out sensitivity was 96.3–100% across four devices at very low cut-offs, reported as ranges without pooling or CIs; it supports, but cannot size, the rule-out step behind ICare-FASTER. van Daalen LE, et al. Eur J Emerg Med 2026. PMID: 42894477

Mechanical ventilator vs bag during CPR (Bayesian SR/MA, 8 studies, n=6,847): ROSC 45.0% vs 41.6% with the ventilator (RR 1.20, 95% credible interval 1.04 to 1.42), but almost all ROSC data were observational and the 2 RCTs reporting it (120 patients) were imprecise (RR 1.23, 0.81 to 1.91). Chu YQ, et al. Resuscitation 2026. PMID: 42863592

Cardiac arrest from non-opioid poisoning (ILCOR systematic review, 4 retrospective cohorts): all at critical risk of bias; sodium bicarbonate was associated with lower survival to discharge (adjusted OR 0.16, 95% CI 0.08 to 0.31), most likely confounded, and the authors find no evidence to change standard ALS. Couper K, et al. Resusc Plus 2026. PMID: 42871632

Rhabdo-RRT score (2 Turkish EDs, n=701): at external validation (AUC 0.862, 95% CI 0.802 to 0.922) renal replacement therapy was needed by 5.3% of low-risk (0–4) and 79.4% of high-risk (9 or more) patients; retrospective, and a low score does not replace serial creatinine, potassium and urine output. Comoglu E, et al. Emerg Med J 2026. PMID: 42869638

Small bowel obstruction (10 US EDs, n=4,478): 10.3% needed surgery within 24 h; closed-loop obstruction (OR 5.49, 95% CI 3.95 to 7.61), bowel wall hypoenhancement and peritonism were the strongest predictors in this retrospective cohort. Brower CH, et al. J Emerg Med 2026. PMID: 42636790

STUMBL in older adults with rib fractures (single Melbourne centre, n=476): discrimination for 3-month mortality was modest (AUC 0.60, 95% CI 0.52 to 0.68) and 75.6% scored above 15, so in older patients add a frailty score rather than relying on STUMBL. Bui A, et al. Injury 2026. PMID: 42629302

Modified Kocher criteria in the irritable hip (single US centre, 110 effusions; abstract only): 5 of 7 septic hips scored only 2 or 3, and 16 children with those scores were not aspirated; in this small retrospective series a low score did not exclude septic arthritis. Owens SP, et al. Pediatr Emerg Care 2026. PMID: 42860754

Paediatric traumatic cardiac arrest (SR/MA, 13 studies, 28 survivors; abstract only): in the individual patient data half of survivors had ROSC within 13 minutes and all within 45 minutes; too few survivors to set a stopping time. Erblich R, et al. Pediatr Emerg Care 2026. PMID: 42745353

EM Evidence Rundown, Issue 35. Page 24 of 28

CORE REVISION — ISSUE 35 — 22 OCTOBER 2026

Core Revision — Tranexamic Acid — The Evidence Base

Mechanism, trauma and the time window, PPH, epistaxis and the neurovascular null results — FRCEM

Exam goal: Quote the key tranexamic acid (TXA) trials with their effect sizes, explain why timing matters, and know where TXA helps, where it does not, and the UK dose.

1. THE TRIALS

EM Evidence Rundown, Issue 35. Page 25 of 28

NoPAC (Ann Emerg Med 2021)

496 adults, 26 UK EDs, epistaxis persisting after first aid and vasoconstrictor

Anterior packing 43.7% (TXA) vs 41.3% (placebo), OR 1.107 (0.769–1.594)

Topical TXA on a dental roll no better than placebo. A 2026 meta-analysis (2,760 patients) found more bleeding control within 10 min, OR 2.54 (1.22–5.29), with soaked gauze but not atomised spray

2. MECHANISM AND UK DOSING

Memory aid — “1-10, 1-8, 3; then 2 for 2”: 1 g over 10 min, 1 g over 8 h, within 3 h (trauma); 2 g within 2 h (head injury). Failed: GI bleeding, SAH, PPH prophylaxis in anaemia, and topical TXA in the UK epistaxis trial. FRCEM focus: CRASH-2’s RR 0.91 and after-3-hour RR 1.44; why CRASH-3 was null on its primary outcome; trauma use is off-label. Sources: CRASH-2 (Lancet 2010, 2011); PATCH-Trauma (NEJM 2023; Ann Emerg Med 2026); CRASH-3 (Lancet 2019); ULTRA (Lancet 2021; J Neurol 2026); WOMAN (Lancet 2017); WOMAN-2 (Lancet 2024); HALT-IT (Lancet 2020); NoPAC (Ann Emerg Med 2021); Fatahi et al (Australas Emerg Care 2026); NICE NG39, NG232; Cyklokapron SmPC (eMC).

EM Evidence Rundown, Issue 35. Page 26 of 28

6 — ACTION POINTS

EM Evidence Rundown, Issue 35. Page 27 of 28

TRIALS TO WATCH

Q4 2026

NICE sepsis in pregnancy Draft guideline out for consultation until 11 November 2026; final publication expected 21 January 2027.

2026–2027

LONGER HORIZON

HINTS training in UK EDs Whether ED-performed HINTS Plus can match specialist accuracy, given 80.6% sensitivity in this week’s study.

Jake Turner — Curated with the assistance of AI (Claude). All content editorially reviewed. Feedback form · emevidence.org · emevidence999@gmail.com

EM Evidence Rundown, Issue 35. Page 28 of 28