Anaesthetics & ICU Evidence Rundown — Issue 10 (November 2026)

ANAESTHETICS & INTENSIVE CARE MEDICINE

Anaesthetics & ICU Evidence Rundown

Issue 10 · November 2026 · UK Edition

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 — renal replacement in the ICU: in ICRAKI (French multicentre RCT, 554 randomised, 535 analysed), starting with intermittent haemodialysis rather than CRRT in ventilated or vasopressor-treated AKI did not meet non-inferiority for major adverse kidney events by day 90 (78.1% vs 71.6%). Crude 90-day mortality, a secondary outcome, was 63.1% vs 53.5% (HR 1.30, 95% CI 1.04 to 1.63) in a trial stopped early, so the difference is likely overestimated. Change this month: MHRA DSI/2026/014 — where possible, give echinocandins such as caspofungin at least 6 hours after CRRT is initiated or after routine filter circuit set changes, without delaying or pausing CRRT solely to give the drug. The alert does not address first or loading doses in new infection: seek microbiology advice. UK: MHRA CRRT and echinocandin alert • UK malignant hyperthermia epidemiology 1988–2025 • PHIND near-patient ARDS subphenotyping in 30 UK and Irish ICUs. Feature: why syndrome RCTs keep failing — the “fruit-peeler” argument tested against the evidence on ARDS and sepsis subphenotypes. Core Revision: Malignant hyperthermia — recognition and dantrolene (Final FRCA).

BOTTOM LINE UP FRONT — NOVEMBER 2026

ACT ON THIS NOW

SAFETY CRRT and echinocandins: PAN membranes can adsorb caspofungin, most of all on a new filter. Where possible, give echinocandins at least 6 hours after CRRT is initiated or a routine filter circuit set change; do not delay or pause CRRT solely to give the drug; for first or loading doses, which the alert does not address, seek microbiology advice.

WORTH KNOWING ON SHIFT

RCT Renal replacement (ICRAKI): keep starting with CRRT in ventilated or vasopressor-dependent AKI; intermittent haemodialysis first was not shown to be non-inferior, and crude mortality was higher (secondary outcome, trial stopped early).

OBS Tumour lysis syndrome: rasburicase within 12 hours was associated with less AKI needing KRT or death (32.7% vs 42.0%; adjusted OR 0.67, 95% CI 0.52 to 0.88; 1,276 US adults).

RCT Dobutamine in septic shock surgery: added to noradrenaline during source control, it cut new or worsening AKI from 34.7% to 22.8% (adjusted RR 0.64, 95% CI 0.49 to 0.83), with no significant effect on mortality.

OBS Malignant hyperthermia (UK): modelled incidence 7.8 per million general anaesthetics; previous uneventful anaesthetics do not exclude MH.

PAEDS Clear fluid fasting: aspiration was 0.03% under both liberal and conventional policies across 580,197 children (RR 0.77, 95% CI 0.40 to 1.47; very low certainty).

KNOW FOR NEXT TIME

RCT Isoflurane sedation (SAVE-ICU): no lower hospital mortality than intravenous sedation (42.5% vs 43.8%; RR 0.98, 95% CI 0.84 to 1.14), with fewer delirium- and coma-free days.

SR/MA Daily sedation interruption (Cochrane): probably reduces mortality versus no interruption (RR 0.88, 95% CI 0.77 to 0.99); no benefit shown over protocolised light sedation.

RCT PELICAN: cardiac index-guided fluid and dobutamine in elective major abdominal surgery did not reduce 7-day complications (44.3% vs 46.9%).

RCT High-flow nasal oxygen after VATS: lower odds of a worse pulmonary complication grade (OR 0.50, 95% CI 0.28 to 0.91) in 60- to 80-year-olds, driven by grade 2 events, with no change in length of stay.

SR/MA Prehabilitation: programmes combining aerobic, pulmonary and functional exercise ranked best for complications, at very low certainty.

OBS Caesarean GA (Europe, n=9,989): 14.8% had GA, almost half after conversion from neuraxial anaesthesia.

OBS Videolaryngoscopy: tube delivery was difficult or failed in 6.3% of intubations despite a good glottic view; record how the tube went in, not only the view.

OBS CT angiography for death by neurological criteria: circulatory arrest was shown in 67.4% of scans within 12 hours of brainstem areflexia vs 98.7% after 12 hours (n=118).

SR/MA ARDS endotypes: 4 of 11 secondary analyses of RCTs found a significant treatment-endotype interaction, but only one, liberal fluids in hyperinflammatory ARDS (90-day mortality 40% vs 50%), was rated high credibility.

OBS ARDS subphenotypes (PHIND, UK): a 1-hour near-patient assay classed 18% of 490 patients as hyperinflammatory, with 60-day mortality of 51% vs 28% (RR 1.8, 95% CI 1.4 to 2.4); not yet shown to guide treatment.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 1 of 28

November’s lead is ICRAKI, a French trial that asked whether ventilated or vasopressor-dependent patients with acute kidney injury could start on intermittent haemodialysis instead of CRRT. Non-inferiority was not shown, and in a trial stopped early mortality was higher with intermittent dialysis, which supports what most UK ICUs already do. Around it sit two more renal items: an MHRA alert on PAN filter membranes adsorbing echinocandins, and a target-trial emulation of early rasburicase in tumour lysis syndrome. Sedation gets a pair, the SAVE-ICU isoflurane trial and the updated Cochrane review of daily sedation interruption. In theatre, two JAMA trials of haemodynamic optimisation point in different directions, and there are full appraisals of prehabilitation, paediatric clear fluid fasting, general anaesthesia for Caesarean delivery and UK malignant hyperthermia epidemiology. The FOAMed section tests a critical care blog’s claim that syndrome trials are bound to fail against the evidence on ARDS subphenotypes, including the UK PHIND study. The Core Revision covers malignant hyperthermia crisis management.

WHAT’S INSIDE — ISSUE 10

• Lead: ICRAKI, intermittent haemodialysis vs CRRT in critically ill AKI (NEJM RCT)
• ICU: early rasburicase in tumour lysis syndrome; SAVE-ICU inhaled isoflurane sedation; Cochrane daily sedation interruption; CT angiography timing for death by neurological criteria
• Guidelines and UK: MHRA CRRT and echinocandin alert; UK malignant hyperthermia epidemiology; French ICU atrial fibrillation guideline
• Airway and regional: glottic view vs ease of videolaryngoscopic intubation; PROSPECT shoulder arthroplasty
• Perioperative: intraoperative dobutamine in septic shock; PELICAN cardiac index-guided therapy; high-flow nasal oxygen after VATS; prehabilitation network meta-analysis; paediatric clear fluid fasting; GA for Caesarean delivery in Europe
• FOAMed and appraisal: why syndrome RCTs keep failing; ARDS endotype review; PHIND
• Quick Hits (26): including CHIPS cold-stored platelets, capnography for nasogastric tubes, haemoadsorption, polymyxin B, POCUS in shock and IV iron
• Core Revision: malignant hyperthermia

CHANGE TONIGHT CHANGE THIS MONTH SAFETY ALERT GUIDELINE INFORMING PRACTICE UK FINAL FRCA

Contents: 1. Key Anaesthetics & ICU Articles — 2. Guidelines & UK Updates — 3. Airway & Regional — 4. Perioperative & Patient Safety — 5. FOAMed & Critical Appraisal — 6. Quick Hits — Core Revision — 7. Action Points — Trials to Watch

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 2 of 28

1 — KEY ANAESTHETICS & ICU ARTICLES

LEAD

INFORMING PRACTICE

RCT

FINAL FRCA

Intermittent Haemodialysis vs CRRT as the First Modality in Shocked or Ventilated AKI (ICRAKI, French RCT, 554 Randomised, 535 Analysed): Non-Inferiority Not Shown; Crude 90-Day Mortality, a Secondary Outcome, Was 63.1% vs 53.5% (HR 1.30) in a Trial Stopped Early

Gaudry S, Boubaya M, Louis G, et al; ICRAKI Trial Group. Intermittent or continuous renal-replacement therapy in acute kidney injury. N Engl J Med. 2026;395:1677–1690. doi:10.1056/NEJMoa2608397. PMID: 42873070. ClinicalTrials.gov NCT06032884. Source route: HAL open-archive copy (hal-05800989, CC BY 4.0), the authors’ manuscript with NEJM-formatted tables and figures, not the publisher-typeset version. Appraised from full text

In 38 French ICUs, adults with AKI and a recognised indication for renal-replacement therapy who were receiving, or had received, invasive ventilation or catecholamines were randomised to intermittent haemodialysis (IHD) or CRRT as the initial modality for days 0 to 3. The primary outcome was a major adverse kidney event by day 90 (death, RRT dependence or creatinine more than 25% above baseline), with a 7.5-point non-inferiority margin for IHD. Enrolment stopped on the monitoring board’s advice at 554 of a planned 1,000 patients because of a survival difference favouring CRRT; in the 535-patient ITT population the primary outcome occurred in 78.1% vs 71.6% (adjusted ARD 6.28 points, 95% CI −1.22 to 13.79), so non-inferiority was not shown and CRRT was not significantly superior. Crude death by day 90, a secondary outcome, occurred in 63.1% vs 53.5% (HR 1.30, 95% CI 1.04 to 1.63), while hypotension was similar (68.8% vs 66.5%).

• 78.1% vs 71.6% MAJOR ADVERSE KIDNEY EVENT BY DAY 90, IHD VS CRRT (PRIMARY)
• ARD +6.28 points PRIMARY, ADJUSTED (95% CI −1.22 TO 13.79); MARGIN 7.5 POINTS
• 63.1% vs 53.5% CRUDE DEATH BY DAY 90 (SECONDARY OUTCOME), HR 1.30 (1.04 TO 1.63)
• NNH 10 CRUDE, CALCULATED (95% CI 6 TO 77); SECONDARY OUTCOME, TRIAL STOPPED EARLY, LIKELY OVERESTIMATED

Source: Gaudry et al, Table 2. Risk ratios adjusted for stratification factors; Kaplan–Meier day-90 mortality 63.3% vs 53.6%; no multiplicity adjustment. CRITICAL APPRAISAL Registered primary; analysis plan

Yes NCT06032884; protocol and statistical plan published; primary outcome and 7.5-point margin match the Methods

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing new for most units, which already start with CRRT; do not move ventilated or vasopressor-dependent patients to intermittent haemodialysis first to save CRRT capacity | Confidence in that: moderate Main concerns: stopping early on a mortality signal without a formal boundary likely overstates the size of the mortality difference, as the authors acknowledge; more chronic kidney disease, emergency surgery and catecholamine use at ICU admission in the IHD arm could overstate harm from IHD; IHD delivered slightly below its target dose may overstate harm from IHD done well Strengths: contemporary multicentre trial restricted to patients with a documented RRT indication, with vital status known for 99.3% at day 90

UK practice: NICE NG148, Acute kidney injury (2019, updated October 2024) covers when to start renal-replacement therapy and whom to involve but makes no recommendation on modality. The UK Renal Association AKI guideline (5th edition, 2011) recommends choosing the modality by the patient’s clinical status, staff expertise and availability (grade 1B). The Renal Association and Intensive Care Society guideline on RRT for critically unwell adults during COVID-19 (October 2020) notes that CRRT predominates in UK ICUs and that, before the pandemic, IHD was available in fewer than 10%.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 3 of 28

INFORMING PRACTICE OBSERVATIONAL

Rasburicase Within 12 Hours of Tumour Lysis Syndrome (STOP-TLS Emulated Target Trial, 36 US Hospitals, n=1,276): AKI Needing Kidney Replacement Therapy or Death 32.7% vs 42.0% (Adjusted OR 0.67)

Shenoy T, Sanchez-Almanzar D, Dias JA, et al. Early treatment with rasburicase and risk of kidney replacement therapy and death in adults with tumour lysis syndrome: emulated target trial. BMJ. 2026;394:e100040. doi:10.1136/bmj-2026-100040. PMID: 42492944. No registration reported. Source route: publisher PDF supplied by Jake (bmj.com). Appraised from full text

Retrospective multicentre cohort (2014–2023) of 1,276 adults with laboratory and clinical tumour lysis syndrome (TLS) and a high tumour burden, analysed as an emulated trial of rasburicase within 12 hours of TLS onset (705 patients, given at a median 5.0 hours) versus delayed or no rasburicase (571, of whom 193 received it after 12 hours). Patients with recent rasburicase, a life-threatening indication for kidney replacement therapy (KRT) at onset, a moribund state or known G6PD deficiency were excluded; most had a haematological malignancy and about one in six were in ICU at onset. After inverse probability weighting, the primary outcome of AKI needing KRT or death in hospital occurred in 32.7% vs 42.0% (adjusted OR 0.67, 95% CI 0.52 to 0.88), and 90-day mortality was 45.6% vs 54.3% (adjusted OR 0.71, 95% CI 0.54 to 0.94). Three treated patients (0.4%) had a rasburicase-related adverse event.

• 32.7% vs 42.0% AKI NEEDING KRT OR DEATH IN HOSPITAL, EARLY VS DELAYED OR NO RASBURICASE (WEIGHTED)
• aOR 0.67 PRIMARY OUTCOME (95% CI 0.52 TO 0.88)
• 11.2% vs 17.0% AKI NEEDING KRT ALONE, AOR 0.62 (0.44 TO 0.87)
• 45.6% vs 54.3% 90-DAY MORTALITY, AOR 0.71 (0.54 TO 0.94)

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Manual chart abstraction at 36 sites; clinical TLS adjudicated by site investigators; starting KRT is itself a clinician decision

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing new: it supports giving rasburicase promptly in established TLS, which UK safety alerts already require; check that ICU and outreach teams can obtain it without delay | Confidence in that: moderate Main concerns: confounding by clinical trajectory: rasburicase may be delayed or withheld in patients judged likely to need KRT anyway, which would overstate benefit; events during the 12-hour grace period and late rescue rasburicase sit in the comparator arm, likely overstating benefit; excluding 548 patients given prophylactic rasburicase and early deaths selects a narrower group (direction unknown) Strengths: large, granular multicentre cohort with a target-trial protocol and consistent sensitivity analyses (adjusted OR 0.45 to 0.69)

UK practice: The Fasturtec SmPC (eMC, revised January 2023) licenses rasburicase for treatment and prophylaxis of acute hyperuricaemia in haematological malignancy with a high tumour burden at the start of chemotherapy, at 0.20 mg/kg/day, and contraindicates it in G6PD deficiency; this cohort mostly used fixed 3 or 6 mg doses and included solid tumours. The NHS England National Patient Safety Alert NatPSA/2025/005/NHSPS (September 2025) requires organisations providing emergency department and cancer services to act on harm from delayed rasburicase. The NHS Specialist Pharmacy Service (updated September 2025) states that the updated BSH tumour lysis syndrome guideline (2025) recommends rasburicase be available for administration within 1 hour of prescription.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 4 of 28

INFORMING PRACTICE

RCT

FINAL FRCA

Isoflurane-Based Inhaled Sedation in Ventilated Adults With Hypoxaemic Respiratory Failure (SAVE-ICU, n=786 Randomised): No Reduction in Hospital Mortality (42.5% vs 43.8%)

Jerath A, Slessarev M, Martin C, et al. Isoflurane-Based Inhaled Sedation in Mechanically Ventilated Adults With Acute Hypoxemic Respiratory Failure: The SAVE-ICU Randomized Clinical Trial. JAMA. 2026 (published online 14 October 2026). doi:10.1001/jama.2026.18471. PMID: 42872958. NCT04415060. Source route: PDF via BMA Library; supplements not read. Appraised from full text

Pragmatic, open-label superiority RCT in 14 Canadian ICUs (September 2020 to March 2025): adults on intravenous sedation for less than 72 hours and ventilated for acute hypoxaemic respiratory failure (PaO2:FiO2 below 300 mm Hg [40 kPa], or COVID-19) were randomised to an isoflurane-based inhaled strategy (sevoflurane at 2 sites) with intravenous adjuncts as needed, or to usual intravenous sedation. Of 786 randomised, 780 formed the primary analysis population (6 randomised in error) and 769 had the primary outcome (11 withdrew consent). All-cause hospital mortality did not differ: 161/379 (42.5%) vs 171/390 (43.8%), RR 0.98 (95% CI 0.84 to 1.14), absolute difference −1.37 points (−8.37 to 5.64). Only 88.5% of the inhaled group received the intervention, and 49.9% to 55.4% also received intravenous sedation on any day in the first 14 days. The inhaled group received more refractory hypoxaemia adjuncts (63.9% vs 55.0%; difference 8.89 points, 1.95 to 15.83), had fewer delirium- and coma-free days (−0.95, −1.69 to −0.22) and had all 5 serious adverse events (raised intracranial pressure in 2, prolonged asystole, propofol infusion syndrome, and cardiovascular instability with seizure) and all 5 abnormal liver function adverse events (5/385 vs 0/395).

• 42.5% vs 43.8% HOSPITAL MORTALITY (RR 0.98, 95% CI 0.84 TO 1.14)
• ARD −1.4 points 95% CI −8.4 TO 5.6; NO NNT AS NO DIFFERENCE SHOWN
• 63.9% vs 55.0% HYPOXAEMIA ADJUNCTS; 1 EXTRA PATIENT PER 12 (CALCULATED, 95% CI 7 TO 52)
• 5 vs 0 SERIOUS ADVERSE EVENTS LEADING TO DISCONTINUATION

CRITICAL APPRAISAL Focused question; registered primary outcome

Yes NCT04415060 with a published protocol; hospital mortality primary throughout

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing: no mortality case for switching ventilated hypoxaemic patients to isoflurane; where inhaled sedation is used for difficult sedation, monitor CO2, rhythm and liver tests | Confidence in that: moderate Main concerns: about half the inhaled arm also had intravenous sedation, biasing any isoflurane-specific effect towards the null; staff new to the device in an open-label trial may explain more coma and adjunct use (learning curve), which would overstate harm of the agent itself; deep sedation in 55.5% to 90.5% of patients on any day and no dose data mean results may not transfer to light-sedation practice (direction unknown) Strengths: concealed allocation, an objective primary outcome and near-complete follow-up in the largest pragmatic trial of inhaled ICU sedation

UK practice: NICE MTG65, now HealthTech guidance HTG607 (Sedaconda ACD-S for sedation with volatile anaesthetics in intensive care, 2022) recommends Sedaconda ACD-S as a cost-saving delivery device when isoflurane or sevoflurane is being considered (1.1) and calls for research on which patients benefit more than with standard care (1.2). The generic UK Isoflurane 100% inhalation vapour SmPC (eMC, Piramal, revised March 2019) lists induction and maintenance of anaesthesia only (4.1) and warns of hepatic injury up to fatal necrosis (4.4), consistent with the liver function signal here. Read alongside the Cochrane daily sedation interruption review below: how deeply patients are sedated may matter more than the agent.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 5 of 28

INFORMING PRACTICE

SR/MA

FINAL FRCA

Daily Sedation Interruption vs No Interruption (Cochrane Update, 23 RCTs, n=5,987): Mortality RR 0.88, About 1 Fewer Death per 35 Adults; Moderate Certainty by the Authors’ GRADE, Low by Ours

Rose L, Williamson DR, McCullagh IJ, et al. Daily sedation interruption versus no daily sedation interruption for critically ill adults and children requiring invasive mechanical ventilation. Cochrane Database Syst Rev. 2026;10:CD009176. doi:10.1002/14651858.CD009176.pub3. PMID: 42841340. Registration CD009176 (protocol 2011). Source route: Cochrane Library full text (PDF); supplementary analyses not read. Appraised from full text

Cochrane review update (search to 8 October 2025) of 23 RCTs (18 adult, 5 paediatric; 5,987 participants, 4,910 after cluster adjustment) comparing daily sedation interruption (DSI) with protocolised sedation (9 trials), usual clinician-directed sedation (9) or analgesia-first or no-sedation strategies (5). DSI probably reduced mortality (RR 0.88, 95% CI 0.77 to 0.99; I² 0%; 18 trials, 4,057 participants; moderate certainty), from 245 to 216 per 1,000 adults. It may shorten ventilation by 14% (1% to 25%; about 1 day in adults; I² 88%; low certainty) and probably shortened ICU stay by 11% (2% to 20%) and reduced tracheostomy (RR 0.68, 0.55 to 0.85). Accidental extubation did not clearly differ (RR 0.88, 0.66 to 1.19; low certainty), and DSI probably increased total morphine dose by 7% (0% to 14%).

• 216 vs 245 ADULT DEATHS PER 1,000 (RR 0.88, 95% CI 0.77 TO 0.99)
• NNT 35 DEATH, CALCULATED FROM THE SUMMARY-OF-FINDINGS RISKS (95% CI 18 TO 500)
• 58 vs 85 TRACHEOSTOMY PER 1,000 (RR 0.68, 0.55 TO 0.85); NNT 38, CALCULATED (27 TO 77)
• −14% VENTILATION DURATION (95% CI −1% TO −25%), ABOUT 1 DAY IN ADULTS; LOW CERTAINTY

CRITICAL APPRAISAL Clear question; protocol registered; changes stated

Yes Cochrane protocol 2011; children added, two subgroups dropped and restraint added after the protocol, all reported

Design: SR/MA | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing for units already running a daily sedation hold within a light-sedation protocol; a unit with neither should add a daily hold | Confidence in that: moderate Main concerns: mixed mortality time points and a heavily weighted no-sedation comparator likely overstate the benefit over current UK protocolised light sedation; unblinded clinicians decided extubation and discharge, which likely overstates the ventilation and length-of-stay gains; most trials predate current propofol- and dexmedetomidine-based practice, likely overstating benefit today Strengths: Cochrane methods with a full search, GRADE and no statistical heterogeneity for mortality (I² 0%) across 18 trials Disagreement: The 2022 SCCM PANDEM guideline conditionally recommends against adding DSI to protocolised sedation (as cited in the review), whereas this review concludes DSI probably reduces mortality against all non-DSI comparators; its comparator subgroup analysis could not settle the difference.

UK practice: NICE CG103 (delirium: prevention, diagnosis and management, 2010, updated January 2023) recommends CAM-ICU or ICDSC for delirium assessment in critical care (1.6.1) but makes no recommendation on sedation interruption. Paired with SAVE-ICU above, the stronger signal is for daily review of sedation depth rather than for any particular sedative agent.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 6 of 28

INFORMING PRACTICE

OBSERVATIONAL

FINAL FRCA

Timing of CT Angiography for Death by Neurological Criteria (Polish Cohort, n=118): Circulatory Arrest Confirmed in 67.4% When Scanned Within 12 h of Brainstem Areflexia vs 98.7% After 12 h

Bohatyrewicz R, Sawicki M, Chamier-Ciemińska K, et al. Optimum timing of cerebral computed tomography angiography for determination of death using neurological criteria: a retrospective sensitivity cohort study. Anaesthesia. 2026 (published online 21 October 2026). doi:10.1111/anae.70438. PMID: 42906612. Source route: open-access full text (publisher PDF, CC BY-NC); supporting appendices S1 and S2 not read. Appraised from full text

Single-centre retrospective cohort from Szczecin, Poland (August 2012 to May 2022): 118 ICU patients with devastating brain injury and initial signs of brainstem areflexia (fixed dilated pupils, loss of cough and gag), all of whom eventually had cerebral circulatory arrest confirmed. The first CTA, read against the Polish four-point criteria, confirmed arrest in 103/118 (87.3%, 95% CI 79.6% to 92.5%), and 14 of the 15 scans showing flow were done within 12 hours of areflexia onset. Confirmation rose from 29/43 (67.4%, 51.3% to 80.5%) at 12 hours or less to 74/75 (98.7%, 91.8% to 99.9%) after 12 hours, and all 15 repeat scans confirmed arrest. The authors recommend waiting at least 12 hours, and repeating an inconclusive CTA at least 12 hours later.

• 29/43 (67.4%) CTA CONFIRMED ARREST AT 12 H OR LESS (95% CI 51.3 TO 80.5)
• 74/75 (98.7%) CONFIRMED AFTER 12 H (95% CI 91.8 TO 99.9)
• 4/7 (57.1%) CONFIRMED WHEN SCANNED UNDER 6 H (95% CI 20.2 TO 88.2)
• n=118 ALL WITH ARREST EVENTUALLY CONFIRMED, SO SPECIFICITY CANNOT BE ESTIMATED

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Onset of areflexia dated retrospectively from records by one clinician; scans verified by one senior radiologist with neuroradiological expertise; no blinding (authors)

Design: Observational | Certainty: very low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing yet: when ancillary CTA is indicated, a scan showing flow soon after areflexia should prompt a later repeat; no fixed 12 h wait is set by UK guidance | Confidence in that: moderate Main concerns: a post hoc 12 h threshold likely overstates how sharp the change is; the 36 patients already confirmed by Doppler or angiography before CTA likely inflate confirmation rates, and their timing group is not reported; with no patients without arrest, the harm of waiting (deterioration, lost donation) and any false positives cannot be measured (direction unknown) Strengths: a clearly defined clinical start point and complete eventual confirmation in every patient, with open data

UK practice: Under the Academy of Medical Royal Colleges Code of Practice for the diagnosis and confirmation of death (2025 update, in effect from 1 January 2025), ancillary investigations are not routinely required (6.55) but are needed when the full examination including apnoea testing is impossible or confounders cannot be excluded (6.64), and are recommended after decompressive craniectomy (6.36). The Code sets no interval for ancillary tests, only that their timing relative to clinical testing be considered (6.57), and clinical testing waits at least 6 hours after the last observed brainstem reflex in primary brain injury (6.22). The UK multidisciplinary consensus guideline on CT angiography for death by neurological criteria (hosted by FICM, January 2023; endorsed by BSNR, ICS, RCR and others) uses four-vessel criteria on the 60-second scan and gives no hours-based timing.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 7 of 28

2 — GUIDELINES & UK UPDATES

CHANGE THIS MONTH

SAFETY ALERT

UK

FINAL FRCA

MHRA Device Safety Information DSI/2026/014: Where Possible, Give Echinocandins at Least 6 Hours After CRRT Is Initiated or After Routine Filter Circuit Set Changes, Without Delaying or Pausing CRRT, Rather Than Avoiding PAN Membranes

Medicines and Healthcare products Regulatory Agency. Continuous renal replacement therapy (CRRT): consider delayed administration of concurrent echinocandins, such as caspofungin, to reduce the risk of adsorption onto new polyacrylonitrile (PAN)-based filter membranes. DSI/2026/014. 13 October 2026. gov.uk. No DOI or PMID (regulatory notice). Source route: gov.uk page (re-fetched 2 November 2026), with the Medicines UK caspofungin DHPC (7 October 2026) and the EMA DHPC (October 2025); the 35-patient retrospective study is Arrestier et al, Infect Dis Ther 2025 (PMID: 40616768). Appraised from full text

The MHRA warns that PAN-based CRRT membranes may adsorb caspofungin and cause ineffective antifungal treatment, with the greatest risk when a new membrane or filter circuit starts, because adsorption falls as the membrane saturates. Where possible, antifungal treatments, including echinocandins and similar drugs, should be given at least 6 hours after CRRT has been initiated or following routine filter circuit set changes, and CRRT should not be delayed or paused solely to give the drug; the alert does not address first or loading doses in new infection, and for patients already on an echinocandin it says delaying the next dose or the start of CRRT should be considered only if clinically viable, so seek microbiology advice. Filter adsorption should be considered when a patient on CRRT responds poorly to any concurrent treatment, specialist pharmacist, infectious diseases, microbiology or mycology advice should be sought, and providers should update local protocols and SOPs. The MHRA departs from the European PRAC advice to avoid PAN membranes or switch antifungal, which it judged not feasible in UK units and likely to harm patients for whom echinocandins are first-line.

• 6 hours or more WHERE POSSIBLE, AFTER CRRT HAS BEEN INITIATED OR FOLLOWING ROUTINE FILTER CIRCUIT SET CHANGES, BEFORE AN ECHINOCANDIN DOSE
• Do not pause CRRT CRRT SHOULD NOT BE DELAYED OR PAUSED SOLELY TO GIVE THE DRUG
• 4 echinocandins CASPOFUNGIN, MICAFUNGIN, REZAFUNGIN AND ANIDULAFUNGIN ARE ALL POTENTIALLY AFFECTED (MEDICINES UK DHPC)

CRITICAL APPRAISAL Developer; stakeholders

Yes MHRA, with the UK Kidney Association, NHS England National Patient Safety team and the devolved administrations

Design: Safety alert | Certainty: very low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Update ICU CRRT and antifungal SOPs this month: where possible, time echinocandin doses at least 6 hours after CRRT has been initiated or following routine filter circuit set changes; do not delay or pause CRRT solely to give the drug; the alert does not address first or loading doses, so seek microbiology advice; treat poor response as possible filter adsorption | Confidence in that: moderate Main concerns: the underlying evidence is low grade and conflicting, so the size of any real drug loss is unknown (direction unknown); the alert does not address first or loading doses, so a timing rule could delay a first echinocandin dose in candidaemia, a harm it does not quantify; other membranes and drugs are not excluded, so a focus on PAN and caspofungin could understate the risk Strengths: pragmatic UK-specific advice, developed with the UK Kidney Association, that keeps first-line echinocandins available Disagreement: The European PRAC (EMA DHPC, October 2025) advises avoiding PAN-based membranes in patients on caspofungin by using another membrane or another antifungal; the MHRA judged this not feasible in UK units and recommends timing doses instead.

UK practice: This is the UK regulator’s own advice, and the Medicines UK caspofungin DHPC (7 October 2026) adds that clinicians should verify the CRRT membrane type, particularly when starting echinocandin therapy, and that the MHRA has asked manufacturers to update caspofungin product information. The MHRA asks for incidents to be reported through Yellow Card in England, Wales and Northern Ireland, and to IRIC in Scotland.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 8 of 28

INFORMING PRACTICE

OBSERVATIONAL

UK

FINAL FRCA

Malignant Hyperthermia in the UK, 1988–2025 (Leeds MH Unit, 375 Index Cases With Full Histories): Modelled Prevalence of Clinical MH Risk 1 in 38,800, and Each Triggering Anaesthetic Carried a 0.46 Risk in Those Who Reacted

Aboelsaod EM, Daly C, Fisher N, et al. Epidemiology of malignant hyperthermia in the UK from 1988 to 2025: implications for prevalence, mode of inheritance, relative risk associated with RYR1 genotypes, and in vitro contracture test phenotyping. Br J Anaesth. 2026 (article in press). doi:10.1016/j.bja.2026.08.048. PMID: 42850084. Source route: PDF via Jake (BJA; open access, CC BY-NC 4.0); supplementary tables not available. Appraised from full text

Retrospective analysis of the database of the Leeds MH Investigation Unit, the UK national diagnostic service, with Bayesian modelling anchored to the 150 IVCT-confirmed reactions of 1988–93 and an estimated 3.37 million UK general anaesthetics a year, plus UK Biobank RYR1 allele frequencies. The modelled incidence was 7.8 MH reactions per million general anaesthetics (1 in 128,000; 95% credibility interval 7.2 to 8.5 per million), and the prevalence of people at risk of a clinical reaction 1 in 38,800 (1 in 34,600 to 1 in 43,700), against a combined prevalence of pathogenic or likely pathogenic RYR1 variants of 1 in 900. In 375 index cases with complete anaesthetic histories, the risk of reacting was consistent with a constant 0.46 per triggering anaesthetic (375 reactions in 809 exposures; 95% CI 0.42 to 0.50): 161 (42.9%) reacted at their first exposure and one only at the twelfth. The modelled number of reactors per family (median 1.70) and a more than 125-fold spread in risk across 26 RYR1 variants argue against simple autosomal dominant inheritance, so genotyping and IVCT probably over-diagnose clinical risk.

• 1 in 128,000 MODELLED MH REACTIONS PER GENERAL ANAESTHETIC, 1988–93 (7.8 PER MILLION; 95% CRI 7.2 TO 8.5)
• 1 in 38,800 MODELLED PREVALENCE OF CLINICAL MH RISK (95% CRI 1 IN 34,600 TO 1 IN 43,700)
• 0.46 RISK OF REACTING WITH EACH TRIGGERING ANAESTHETIC, INDEX CASES (95% CI 0.42 TO 0.50)
• 214/375 INDEX CASES WITH AT LEAST ONE UNEVENTFUL EXPOSURE BEFORE REACTING (57.1%, CALCULATED)

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Numerator: IVCT-confirmed reactions at the single national unit with 95% ascertainment assumed; denominator extrapolated from England 1989 procedure data

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing changes in theatre: people with an MH-associated RYR1 variant or positive IVCT, and their untested relatives, should still be managed as MH-susceptible, and a history of uneventful anaesthetics does not exclude MH | Confidence in that: high Main concerns: the prevalence depends on investigator-set ascertainment and denominator assumptions; if ascertainment was lower than 95%, prevalence is underestimated; analysis by the national unit’s own team (the senior author is president of the European MH Group; one author is a paid adviser to Norgine), direction unknown; an older, less diverse cohort means variant-specific risks may not transfer to today’s UK population (direction unknown) Strengths: IVCT-confirmed reactions from 1988 to 2025 from a single national referral centre linked to UK population genomic data

UK practice: Under the UK MH Registry guidance on testing for MH (web page, copyright 2023), suspected index cases and their relatives are tested through the Leeds MH Investigation Unit, by DNA testing or by muscle biopsy contracture testing (a two-night stay), and a negative gene result must be confirmed by biopsy before someone is considered not at risk. The Association of Anaesthetists Malignant hyperthermia 2020 guideline covers how to refer a patient after a suspected reaction; this paper changes the counselling around incidental RYR1 findings, not that referral pathway.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 9 of 28

INFORMING PRACTICE

GUIDELINE

FINAL FRCA

French Intensive Care Society Guideline on Atrial Fibrillation in ICU: 18 Recommendations, 16 of Them Expert Opinion, Advising Against Routine Anticoagulation for New-Onset AF

Labbé V, Aissaoui N, Asfar P, et al. Guidelines for the management of atrial fibrillation in adult patients admitted to an intensive care unit. Ann Intensive Care. 2026;16:100159. doi:10.1016/j.aicoj.2026.100159. PMID: 42894798. Source route: PMC full text (PMC13676918; open access, CC BY); supplementary PICO file not read. Appraised from full text

Société de Réanimation de Langue Française expert recommendations (modified GRADE, Delphi voting): 24 questions produced 18 recommendations, all with strong agreement, of which 2 are GRADE 2+ and 16 expert opinion; 6 questions had no recommendation. For new-onset AF (NOAF) it suggests systematic transthoracic echo, rhythm control in cardiogenic shock (electrical cardioversion if ventilated), amiodarone as first-line pharmacological cardioversion (amiodarone or propafenone in sepsis without severe LV dysfunction, GRADE 2+), and beta-blockers for rate control unless there is cardiogenic shock needing vasopressors or inotropes, or ventricular systolic dysfunction, when amiodarone is preferred. It suggests no therapeutic anticoagulation for NOAF under 48 hours without cardioversion and not routinely at 48 hours or more, unfractionated heparin over LMWH if used, and pausing anticoagulation for pre-existing AF during acute illness unless thrombotic risk is very high.

• 18 RECOMMENDATIONS FROM 24 QUESTIONS; 6 WITH NO RECOMMENDATION
• 16 of 18 EXPERT OPINION; 2 GRADE 2+
• 3% to 11% IN-HOSPITAL ARTERIAL THROMBOTIC EVENTS WITH NOAF (CITED COHORTS)
• 19% MAJOR BLEEDING IN THE FAST COHORT OF VENTILATED SEPTIC PATIENTS WITH NOAF

R = SRLF recommendation; numbers in the NICE column are NG196 recommendations. CRITICAL APPRAISAL Developer, funding and conflicts

Partly Funded by the French Intensive Care Society; independent steering committee without conflicts; 4 of 14 authors declared industry grants or fees

Design: Guideline | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing tonight: a reasonable template for a unit AF protocol, but withholding heparin from new-onset AF departs from NICE and needs a documented consultant decision | Confidence in that: moderate Main concerns: 16 of 18 recommendations rest on expert opinion, so their direction of error is unknown; per-expert searches of two databases may have missed evidence (direction unknown); anticoagulation advice rests on confounded sepsis cohorts (direction unknown) Strengths: ICU-specific, with transparent grading and explicit gaps where no recommendation is possible Disagreement: SRLF suggests not starting therapeutic anticoagulation for NOAF under 48 hours without cardioversion (R15A), whereas NICE NG196 offers heparin at initial presentation to people with new-onset AF on no or subtherapeutic anticoagulation (1.8.8).

UK practice: NICE NG196 (atrial fibrillation: diagnosis and management, April 2021, updated June 2021) is not ICU-specific (its scope does not mention critical care) and uses CHA2DS2-VASc to guide anticoagulation (1.2, 1.6). The 2023 Scandinavian (SSAI) ICU guideline also weakly recommends against routine therapeutic anticoagulation for NOAF; units adopting this approach should agree it with cardiology.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 10 of 28

3 — AIRWAY & REGIONAL

INFORMING PRACTICE

OBSERVATIONAL

FINAL FRCA

Video Classification of Intubation Score Reproducible Between Operator and Observer (Spanish Multicentre Cohort, n=5,302; Krippendorff’s α 0.86, 95% CI 0.85 to 0.87): Tube Delivery Was Difficult or Failed in 6.3% Despite a View Above 75%, and Easy in 24.5% Despite a View Below 25%

Fernández-Vaquero MÁ, Charco-Mora P, Gómez-Ríos MÁ, et al, on behalf of the VCISpain Trial Investigators Group. Association between glottic view and tracheal intubation during videolaryngoscopy: a prospective observational study. Anaesthesia. 2026 (published online 2 October 2026). doi:10.1111/anae.70417. PMID: 42823939. ClinicalTrials.gov NCT06537531. Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Prospective observational study in 44 Spanish hospitals (September 2024 to September 2025) of adults (ASA 1–4; 98.1% in theatre) intubated with a videolaryngoscope chosen by the operator. The operator and an observer each scored the Video Classification of Intubation (VCI): blade geometry (Macintosh-type 57.8%, hyperangulated 42.2%), percentage of glottic opening (POGO) at tube delivery, and tube delivery (easy: first attempt with standard technique; difficult: an adjunct or more than one attempt; failed: change of blade geometry or operator). For the primary objective, inter-rater reliability in 5,302 paired assessments was high (Krippendorff’s α 0.86, 95% CI 0.85 to 0.87; full agreement 90.6%). On operator ratings, delivery was difficult or failed in 252/4,025 (6.3%) with POGO above 75% but easy in 37/151 (24.5%) with POGO below 25%; hypoxaemia (SpO2 below 90% for 10 s or more) occurred in 1.1%, oesophageal intubation in 0.1%, and five patients (0.1%) could not be intubated with videolaryngoscopy.

• 6.3% DIFFICULT OR FAILED (252/4,025) DESPITE POGO ABOVE 75%
• 24.5% EASY (37/151) DESPITE POGO BELOW 25%
• 30.5% FAILED (46/151) WITH POGO BELOW 25%
• α 0.86 OPERATOR–OBSERVER AGREEMENT (95% CI 0.85 TO 0.87)

Source: Fernández-Vaquero et al, Table 3 (airway operator ratings). “Difficult” includes any adjunct other than a stylet with a hyperangulated blade; the definition does not separate planned from unplanned adjuncts, so any adjunct with a Macintosh-type blade counts as difficult.

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Scored in real time by unblinded operator and observer; POGO disagreed in 9.4%; number of attempts not recorded

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing in technique; in airway notes and handovers, record blade type, view and how the tube went in (adjunct, attempts, change of device or operator), not view alone | Confidence in that: moderate Main concerns: unblinded real-time self-rating with an observer in the room likely overstates inter-rater agreement; only 151 poor views and 22 ED or ICU intubations, so findings may understate difficulty in UK emergency and ICU airway management; counting any adjunct use, planned or not, as “difficult” may overstate the gap between a good view and a difficult intubation Strengths: large prospective multicentre cohort with paired independent ratings for 5,302 of 5,333 documented intubations

UK practice: The Difficult Airway Society 2025 guidelines for management of unanticipated difficult tracheal intubation in adults (Ahmad et al, Br J Anaesth 2026; PMID: 41203471) list documentation among their 65 recommendations and prioritise success at the first attempt. This study supports documenting the whole videolaryngoscopic intubation: a good POGO on its own should not be recorded as an easy airway when the tube needed an adjunct, further attempts or a change of blade.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 11 of 28

INFORMING PRACTICE

GUIDELINE

FINAL FRCA

PROSPECT Recommendations After Total Shoulder Arthroplasty (Systematic Review, 30 RCTs and 7 Reviews): Single-Shot Interscalene Block With Paracetamol, an NSAID and Dexamethasone 8–10 mg; Not Liposomal Bupivacaine or Routine Catheters

Mrazek I, Desai N, Sauter AR, et al, on behalf of the PROSPECT Working Group of ESRA. Pain management after total shoulder arthroplasty: a systematic review and procedure-specific postoperative pain management (PROSPECT) recommendations. Anaesthesia. 2026 (published online 22 October 2026). doi: 10.1111/anae.70427. PMID: 42910655. PROSPERO CRD420251156103. Source route: PDF supplied by Jake (open access). Appraised from full text

Systematic review of English-language RCTs and systematic reviews of interventions for pain after total shoulder arthroplasty (four databases to 2 June 2026; primary outcome pain score, with more than 1 point out of 10 taken as clinically relevant); 30 RCTs and seven reviews were included, trials were not pooled, and recommendations were agreed by modified Delphi within the ESRA PROSPECT Working Group. Recommended: paracetamol plus an NSAID started before or during surgery and continued afterwards, a single intra-operative IV dose of dexamethasone 8–10 mg, rescue opioids, and a single-shot interscalene block with bupivacaine or ropivacaine (moderate evidence, strong recommendation), with local infiltration analgesia (LIA) using bupivacaine or ropivacaine as the alternative (moderate evidence, strong recommendation). Not recommended: continuous interscalene catheters (two conflicting trials), liposomal bupivacaine (six of eight interscalene trials found no statistically significant or clinically relevant benefit) and opioid-free protocols.

• 30 + 7 RCTS AND SYSTEMATIC REVIEWS INCLUDED (1,236 RECORDS SCREENED)
• Strong SINGLE-SHOT INTERSCALENE BLOCK; MODERATE EVIDENCE
• 6 of 8 LIPOSOMAL BUPIVACAINE INTERSCALENE TRIALS WITHOUT A SIGNIFICANT OR RELEVANT BENEFIT
• >1/10 PAIN DIFFERENCE TAKEN AS CLINICALLY RELEVANT

CRITICAL APPRAISAL Developer, funding, conflicts

Partly ESRA PROSPECT Working Group, unrestricted ESRA grant; senior author is an Associate Editor of Anaesthesia; two authors report industry honoraria

Design: Guideline | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Little: it supports single-shot interscalene block with multimodal analgesia as the default, LIA when a block is not possible, and no routine catheters or liposomal bupivacaine | Confidence in that: moderate Main concerns: no meta-analysis and consensus weighting of small trials, so recommendation strength rests on panel judgement (direction unknown); trials too small to assess harms, so a strong block recommendation based on efficacy alone may understate risk in patients with poor respiratory reserve; controls often lacked optimised multimodal analgesia, which overstates the benefit of single added interventions Strengths: registered, procedure-specific review searched to June 2026 that separates plain from liposomal local anaesthetic Disagreement: NICE NG157 (2020) could not recommend a specific approach for shoulder replacement; PROSPECT (2026), with trials to June 2026 and expert consensus, strongly recommends a single-shot interscalene block.

UK practice: NICE NG157 (joint replacement (primary): hip, knee and shoulder, 2020) recommendation 1.3.3 asks clinicians to discuss general, regional, local infiltration and nerve-block options with people having elective shoulder replacement, but the committee could not recommend specific options and made research recommendations instead. PROSPECT therefore fills a gap in UK guidance for this operation.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 12 of 28

4 — PERIOPERATIVE & PATIENT SAFETY

INFORMING PRACTICE

RCT

FINAL FRCA

Intraoperative Dobutamine Added to Noradrenaline During Source Control for Septic Shock (39 Chinese Hospitals, n=584): New or Worsening AKI Fell From 34.7% to 22.8%, With No Significant Difference in Mortality or Kidney Replacement

Chinese Perioperative Septic Shock Investigators for the Chinese Peri-Operative Organ Protection Group (Chu L, Xie G, Shi R, et al). Intraoperative dobutamine plus norepinephrine to prevent postoperative acute kidney injury in septic shock: a randomized clinical trial. JAMA. 2026 (published online 14 October 2026). doi: 10.1001/jama.2026.21322. PMID: 42872914. ClinicalTrials.gov NCT06462313. Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Open-label RCT in 39 tertiary hospitals in China: 584 adults with Sepsis-3 septic shock having surgery for source control under general anaesthesia were randomised to dobutamine started at 5 μg/kg/min plus noradrenaline, or noradrenaline alone, with noradrenaline titrated to a MAP of 65 mmHg in both arms, during surgery only. The primary outcome was new or worsening KDIGO acute kidney injury (AKI) within 7 days, judged by blinded assessors. AKI occurred in 66/290 (22.8%) vs 102/294 (34.7%) (adjusted RR 0.64, 95% CI 0.49 to 0.83; unadjusted difference −11.9 points, −19.2 to −4.7), with less stage 2 or 3 AKI in post hoc analyses (37/290, 12.8% vs 69/294, 23.5%, calculated from Table 2). Mortality at 28 days (16.2% vs 20.4%; RR 0.75, 0.55 to 1.04) and kidney replacement within 7 days (9.0% vs 10.9%; RR 0.82, 0.52 to 1.27) did not differ significantly, and tachycardia over 100/min was commoner with dobutamine (41.0% vs 29.9%; RR 1.42, 1.16 to 1.74).

• 22.8% vs 34.7% AKI WITHIN 7 DAYS, DOBUTAMINE VS NORADRENALINE ALONE
• ARD −11.9 points 95% CI −19.2 TO −4.7 (UNADJUSTED)
• NNT 9 TO PREVENT ONE AKI, CALCULATED (95% CI 6 TO 22)
• NNH 9 TACHYCARDIA >100/MIN, CALCULATED (95% CI 5 TO 29)

CRITICAL APPRAISAL Registered primary; analysis plan

Yes NCT06462313; AKI within 7 days is the primary in the abstract and methods; staging and 30 robustness analyses labelled post hoc

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing yet: keep dobutamine for septic shock with suspected low cardiac output; this single trial does not justify routine intraoperative dobutamine during source control until it is replicated | Confidence in that: moderate Main concerns: open label with clinician-titrated noradrenaline and fluids, so co-interventions could favour the dobutamine arm (overstates benefit); AKI is a surrogate; deaths and kidney replacement were not significantly reduced, so patient-centred benefit is unproven; cardiac function was not used to select patients (advanced monitoring only when available), so who benefits is unclear (direction unknown) Strengths: large multicentre trial with concealed allocation, blinded outcome assessment and no loss to follow-up

UK practice: The Dobutamine 5 mg/ml solution for infusion SmPC (eMC, text revised August 2024) licenses dobutamine in adults for inotropic support in cardiac decompensation due to depressed contractility, and in septic shock as an addition to dopamine where ventricular function is disturbed; routine use without evidence of low output sits outside that wording. NICE NG148 (acute kidney injury, 2019, last updated October 2024) asks for an AKI risk assessment before surgery and names emergency surgery with sepsis as a risk factor (1.1.13), so the trial population is one UK teams already flag as high risk.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 13 of 28

INFORMING PRACTICE

RCT

FINAL FRCA

PELICAN: Keeping Cardiac Index at the Preoperative Resting Value During Elective Major Abdominal Surgery (12 European Centres, n=1,095) Did Not Reduce 7-Day Complications (44.3% vs 46.9%)

Saugel B, Meidert AS, Meyhoff CS, et al; PELICAN Trial Group. Individualized cardiac index–guided hemodynamic therapy in high-risk patients undergoing elective major abdominal surgery: the PELICAN randomized clinical trial. JAMA. 2026 (published online 19 October 2026). doi:10.1001/jama.2026.18574. PMID: 42896534. ClinicalTrials.gov NCT05648279. Source route: PDF via BMA Library (supplied by Jake). Appraised from full text

Single-blind RCT in 12 university centres in Austria, the Czech Republic, Denmark, Germany and Spain: 1,111 high-risk adults aged 45 or over having elective major abdominal surgery were randomised (1,095 analysed) to intraoperative therapy (crystalloid, 500 mL boluses and dobutamine by algorithm) keeping cardiac index at or above each patient’s awake preoperative resting value, measured by bioreactance, or to routine care with cardiac index recorded but hidden from clinicians. The primary outcome, a 7-day composite of AKI, myocardial injury, severe infection, non-fatal cardiac arrest or death, occurred in 243/549 (44.3%) vs 256/546 (46.9%) (RR 0.94, 95.1% CI 0.83 to 1.08; difference −2.6 points, −8.5 to 3.3), although the intervention group got more fluid (median 3,500 vs 2,838 mL) and more dobutamine (64.1% vs 0.5%). Unlike the septic-shock trial above, where dobutamine during emergency source control reduced AKI, adding fluid and dobutamine to protect cardiac index in elective patients did not reduce organ injury; in an exploratory, unadjusted secondary analysis, 90-day mortality was 5.7% vs 2.8% (RR 2.06, 95% CI 1.13 to 3.78).

• 44.3% vs 46.9% 7-DAY COMPOSITE, INDIVIDUALISED VS ROUTINE
• ARD −2.6 points 95.1% CI −8.5 TO 3.3; TRIAL POWERED FOR 10 POINTS
• 62% vs 23% MEDIAN TIME WITH CARDIAC INDEX AT OR ABOVE RESTING VALUE
• 5.7% vs 2.8% 90-DAY DEATH, EXPLORATORY AND UNADJUSTED (ARD 3.0 POINTS AS PRINTED IN TABLE 3, 95% CI 0.6 TO 5.4)

CRITICAL APPRAISAL Registered primary; analysis plan

Partly NCT05648279 and published protocol; statistical analysis plan finalised 20 March 2026, after enrolment ended in October 2025

Design: RCT | Certainty: moderate (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing new: do not add a protocol that targets preoperative resting cardiac index with fluid and dobutamine; continue cardiac output monitoring as a “consider” for major or high-risk surgery | Confidence in that: high Main concerns: composite weighted towards asymptomatic troponin rises, which can mask benefit or harm in rarer, more important components (direction unknown); analysis plan finalised after the trial ended and after an interim look that favoured the intervention (1-sided P=.046), though the final result is null (small risk of selective reporting); the 90-day mortality signal is one of many unadjusted secondary outcomes and may be chance; it means harm cannot be ruled out, not that harm is shown Strengths: large multicentre trial with blinded control monitoring, good protocol separation and complete primary outcome data

UK practice: NICE NG180 (perioperative care in adults, 2020) says to consider cardiac output monitoring for people having major or complex or high-risk surgery (1.4.5); this recommendation replaced NICE MTG3 (CardioQ-ODM, 2011). PELICAN does not change that recommendation, but it gives no support for a resting-cardiac-index target with routine dobutamine on top of monitoring.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 14 of 28

INFORMING PRACTICE

RCT

FINAL FRCA

Prophylactic High-Flow Nasal Oxygen After VATS Lung Resection in Adults Aged 60–80 (Single-Centre RCT, 223 Randomised, 210 Analysed): Lower Odds of a Worse Pulmonary Complication Grade (OR 0.50), Driven by Grade 2 Events, With No Difference in Oxygenation or Length of Stay

Lin R, Huang Y, Yu G, et al. Prophylactic high-flow nasal oxygen vs. conventional oxygen for prevention of postoperative respiratory complications: a randomized trial. Crit Care. 2026;30:528. doi:10.1186/s13054-026-06205-4. PMID: 42864673. ChiCTR2200063161. Source route: PMC full text (PMC13660337; open access, CC BY-NC-ND). Appraised from full text

Open-label single-centre RCT in Fuzhou, China: 223 adults aged 60–80 having VATS lobectomy or segmentectomy for stage I–IIIa non-small cell lung cancer were randomised (210 analysed) to high-flow nasal oxygen (HFNO, about 40 L/min and FiO2 40%, median 67 hours) or nasal cannula oxygen from extubation, both titrated to SpO2 95% or more (93% in COPD or obesity). The primary outcome was the worst postoperative pulmonary complication (PPC) grade in hospital on a modified Kroenke scale, analysed by ordinal regression: HFNO lowered the odds of a worse grade (OR 0.504, 95% CI 0.280 to 0.907), driven by grade 2 events (16.3% vs 29.2%) rather than grade 3 or higher (7.7% vs 10.4%). In the secondary binary comparison, grade 2 or higher PPCs occurred in 25/104 (24.0%, calculated; the paper prints 23.1%) vs 42/106 (39.6%) (difference −15.6 points, 95% CI −28.0 to −3.2). PaO2/FiO2 on day 2, 6-minute walk distance and median length of stay (5 days in both groups) did not differ.

• OR 0.50 WORSE PPC GRADE, ORDINAL PRIMARY (95% CI 0.28 TO 0.91)
• 24.0% vs 39.6% GRADE 2 OR HIGHER, SECONDARY BINARY (24.0% CALCULATED; PAPER PRINTS 23.1%; ARD −15.6)
• NNT 7 GRADE 2 OR HIGHER, CALCULATED (95% CI 4 TO 32)
• 7.7% vs 10.4% GRADE 3 OR HIGHER PPCS (NOT TESTED SEPARATELY)

CRITICAL APPRAISAL Registered primary; analysis plan

Partly ChiCTR2200063161 before enrolment; sample size based on binary PPC incidence (45.2% vs 23.8%), not the ordinal primary

Design: RCT | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing yet: do not start routine prophylactic HFNO after VATS; it remains a rescue or escalation option for hypoxaemia | Confidence in that: moderate Main concerns: open-label grading of soft lower-grade complications by unblinded staff likely overstates benefit; the paper prints 23.1% for 25/104; the risk difference of −15.6 matches 24.0%, so we used the counts; patients who declined follow-up after readmission were excluded, and readmission may reflect complications (direction unknown) Strengths: randomised trial in an older thoracic population with daily PPC grading and a prespecified severity outcome

UK practice: The BTS guideline for oxygen use in adults in healthcare and emergency settings (2017) recommends a target of 88–92% for patients at risk of hypercapnic respiratory failure, adjusted to 94–98% if the PaCO2 is normal; the trial’s 93% target in COPD is higher. The BTS clinical statement on high flow nasal therapy in adults (draft for consultation, 20 October to 20 November 2026) covers acute and chronic respiratory failure, secretion management and palliation; its scope does not include prophylactic postoperative use.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 15 of 28

INFORMING PRACTICE

SR/MA

FINAL FRCA

Exercise Prehabilitation Network Meta-Analysis (96 RCTs, n=7,396): Programmes Combining Aerobic, Pulmonary and Functional Exercise Ranked Best for Complications and Length of Stay, at Mostly Low or Very Low Certainty

McIsaac DI, Gillmore TC, Al-Bayati M, et al. Relative efficacy of exercise prehabilitation interventions and their components: a systematic review and network meta-analysis of randomised controlled trials. Br J Anaesth. 2026 (article in press). doi:10.1016/j.bja.2026.08.052. PMID: 42901859. PROSPERO CRD420251034396. Source route: PDF via Jake (BJA; open access, CC BY 4.0); appendices not available. Appraised from full text

Network meta-analysis of 96 RCTs (7,396 adults; 40.6% of trials orthopaedic, 30.2% oncology) of at least 7 days of exercise prehabilitation before elective surgery, comparing aerobic, strength, pulmonary and functional exercise, alone or combined, with usual care and each other, for complications, length of stay, quality of life and physical recovery. Combined aerobic, pulmonary and functional exercise ranked first for complications (OR 0.16, 95% CI 0.05 to 0.60; very low certainty) and length of stay (mean difference −3.4 days, 95% CI −5.7 to −1.0; moderate certainty), but this arm held only 54 patients from 3 trials; pulmonary exercise alone, from 12 trials, also reduced complications (OR 0.40, 95% CI 0.24 to 0.67). Aerobic plus strength exercise ranked first only for quality of life (SMD 0.67, 95% CI 0.07 to 1.27; low certainty), no programme improved physical recovery, and without control event rates no NNT can be derived.

• OR 0.16 COMPLICATIONS, AEROBIC + PULMONARY + FUNCTIONAL VS USUAL CARE (95% CI 0.05 TO 0.60; VERY LOW CERTAINTY)
• −3.4 days LENGTH OF STAY, SAME COMPARISON (95% CI −5.7 TO −1.0; MODERATE CERTAINTY)
• OR 0.40 COMPLICATIONS, PULMONARY EXERCISE ALONE (95% CI 0.24 TO 0.67; 12 TRIALS)
• 74/96 TRIALS AT HIGH RISK OF BIAS (77.1%)

Source: McIsaac et al, Results and Figure 3. Figure 3 prints the first complications CI as 0.05 to 0.50; the abstract and text give 0.05 to 0.60, which we use. Absolute risks are not reported in the main paper.

CRITICAL APPRAISAL Question, protocol and search

Yes PROSPERO CRD420251034396; six databases from inception to 4 April 2025, trial registries, no language restriction, duplicate screening

Design: SR/MA | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing to mandate: keep aerobic exercise central to prehabilitation and consider adding respiratory or inspiratory muscle training | Confidence in that: moderate Main concerns: 77% of trials at high risk of bias, likely overstating benefit; the top-ranked combination rests on 54 patients in 3 trials, so its OR 0.16 is imprecise and likely overstated; complications were often pulmonary-only, which may overstate the advantage of pulmonary components Strengths: registered protocol, comprehensive search and CINeMA ratings with prediction intervals Disagreement: The UK guidelines make aerobic plus strength exercise the core prescription, but in this network that combination did not significantly reduce complications (OR 0.53, 95% CI 0.19 to 1.48) or length of stay; pulmonary-containing programmes did better, whereas the guidelines recommend inspiratory muscle training only before lung cancer surgery.

UK practice: The Macmillan prehabilitation for people with cancer guidelines (October 2025, with NHS England, CPOC and the NIHR) recommend combined aerobic and strength exercise for patients having cancer surgery, with the aerobic component at moderate to high intensity where appropriate, and inspiratory muscle training with aerobic and strength exercise before lung cancer surgery (moderate-quality evidence, strong recommendations). They advise at least 2 weeks of exercise, with programmes longer than 4 weeks more effective.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 16 of 28

INFORMING PRACTICE

OBSERVATIONAL

FINAL FRCA

General Anaesthesia for Caesarean Delivery in 31 European Countries (ACCESS, n=9,989): GA Rate 14.8% Overall, 54.0% in Category 1 and 10.9% in Elective Cases, With Almost Half of GA Following Conversion From Neuraxial

Weiniger CF, Lustig A, Afshari A, et al; ACCESS Study Group. Incidence of general anaesthesia according to urgency of surgery: a multicentre prospective cohort study of 9989 Caesarean deliveries in Europe. Br J Anaesth. 2026 (article in press). doi:10.1016/j.bja.2026.08.046. PMID: 42855372. Source route: PDF via Jake (BJA; open access, CC BY-NC-ND 4.0); supplementary material not available. Appraised from full text

Prospective snapshot cohort of every Caesarean delivery in 361 centres across 31 countries, each centre collecting for a self-chosen 2-week period between September 2023 and April 2025; the UK contributed Wales (55 cases) and Northern Ireland (315), but not England or Scotland. Anaesthetists recorded urgency by the four-category Lucas classification, and the primary outcome was the incidence of GA by urgency, described without hypothesis testing. GA was used in 14.8% (1,474/9,989; 95% CI 14.0 to 15.5): 54.0% of category 1, 13.8% of category 2, 11.3% of category 3 and 10.9% of elective cases. Initial GA accounted for 759 cases (maternal request was the commonest recorded reason in elective cases, 110/234), and conversion from neuraxial to GA for 715 of 9,228 (7.7%), most often for failed neuraxial anaesthesia (259/715); national GA rates ranged from 3.6% (Wales) to 49.0% (Serbia), with Northern Ireland at 6.0%.

• 14.8% GA, ALL CAESAREAN DELIVERIES (95% CI 14.0 TO 15.5)
• 54.0% GA, CATEGORY 1 (384/711)
• 10.9% GA, ELECTIVE (503/4,626)
• 6.1% CONVERSION TO GA AFTER NEURAXIAL, ELECTIVE (269/4,392)

Source: Weiniger et al, Table 2. Conversion counts include conversions for maternal request with a working block and for emergent obstetric reasons; 2.6% of neuraxial cases (239/9,228) received supplementary drugs to avoid conversion.

CRITICAL APPRAISAL Exposure and outcome measured reliably

Partly Anaesthetist-completed form with standard urgency categories, but local categorisation varied and accuracy was not independently verified

Design: Observational | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing new: it gives a European benchmark for auditing your unit’s GA and conversion rates by urgency category, with conversion in elective cases the clearest target for review | Confidence in that: moderate Main concerns: self-selected, research-active centres reporting for 2 weeks may have lower GA rates than average, so the figures may understate GA use; “conversion” mixes failed blocks with maternal request and obstetric emergencies, which overstates neuraxial failure, while pain managed with supplementary drugs is not counted, which understates inadequate blocks; UK data come only from Wales and Northern Ireland (370 cases), so they do not describe England Strengths: prospective all-comer cohort of nearly 10,000 deliveries using one standard urgency classification

UK practice: RCoA GPAS Chapter 9, obstetric population (2025) makes the lead obstetric anaesthetist responsible for auditing the service against agreed standards, including anaesthetic complication rates (1.11), and asks units to keep guidelines on GA for Caesarean section and on managing failed regional anaesthesia, including pain during Caesarean section (2.42). It also recommends that the anaesthetist be told the urgency category and indication at the earliest opportunity (5.20), the classification ACCESS used.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 17 of 28

INFORMING PRACTICE

SR/MA

PAEDS

FINAL FRCA

Liberal (Under 2 h) vs Conventional Clear Fluid Fasting in Children, SR/MA (33 Studies, n=580,197): Pulmonary Aspiration 0.03% Under Both Policies (RR 0.77, 95% CI 0.40 to 1.47), Very Low Certainty

Nagase T, Takahashi K, Takagi T, et al. Liberal vs. conventional pre-operative clear fluid fasting policies in children: a systematic review and meta-analysis of randomised and non-randomised studies. Anaesthesia. 2026 (published online). doi:10.1111/anae.70441. PMID: 42922413. PROSPERO CRD420251004030. Source route: PDF in the candidates folder (publisher version); supporting tables S1–S2 not available. Appraised from full text

Systematic review and meta-analysis of 33 studies (16 RCTs, 11 observational studies and 6 quasi-experimental, quality improvement or before-after studies, published 2005–2025) of children under 18 having general anaesthesia or sedation, comparing clear fluid policies of less than 2 hours (including 1-hour and 0-hour policies) with 2 hours or more. The primary outcome was pulmonary aspiration (with hypoxaemia or pneumonia, as each study defined it) and the secondary outcome vomiting. Pooled aspiration incidence was 0.03% (95% CI 0.01 to 0.09%) under each policy, and in 14 comparative studies (475,399 children) the risk ratio was 0.77 (95% CI 0.40 to 1.47); vomiting did not differ significantly (RR 0.66, 95% CI 0.39 to 1.12). Only one RCT, in children with cyanotic congenital heart disease (3/170 vs 3/174), recorded any aspiration, so the comparison rests on observational cohorts, chiefly EUROFAST (38.4% of the weight).

• 0.03% vs 0.03% POOLED ASPIRATION INCIDENCE, LIBERAL VS CONVENTIONAL (95% CI 0.01 TO 0.09% FOR EACH)
• 1.2 vs 1.9 per 10,000 CRUDE ASPIRATION, 35/298,658 LIBERAL VS 34/176,741 CONVENTIONAL (CALCULATED, UNADJUSTED)
• RR 0.77 ASPIRATION, 14 COMPARATIVE STUDIES (95% CI 0.40 TO 1.47)
• RR 0.66 VOMITING, 8 COMPARATIVE STUDIES (95% CI 0.39 TO 1.12)

CRITICAL APPRAISAL Question and protocol

Yes PROSPERO CRD420251004030; under 18, GA or sedation, studies from 1995 to September 2025

Design: SR/MA | Certainty: very low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing new: it supports the existing UK 1-hour clear fluid policy for healthy children having elective GA, and does not extend that reassurance to children with delayed gastric emptying | Confidence in that: moderate Main concerns: exposure is the institutional policy, not the actual fasting time, which biases towards the null because many children on liberal policies still fast for hours; the comparison depends on two large prospective multicentre cohorts (EUROFAST and Beck carry 64% of the weight), so era and centre confounding is possible (direction unknown); aspiration was variably defined and often a secondary outcome, which likely understates incidence in both groups Strengths: combines randomised and very large observational datasets to estimate a rare event, with GRADE by design

UK practice: The APAGBI, ESPA and ADARPEF joint consensus statement on clear fluid fasting (Thomas et al, Paediatr Anaesth 2018; PMID: 29700894) recommends that, unless there is a clear contraindication, all children able to take clear fluids be allowed and encouraged to drink them up to 1 hour before elective general anaesthesia. The APAGBI Hot Topic on ‘sip til send’ (November 2025) says national work is under way to inform an updated APAGBI position statement and advises individual assessment for children with delayed gastric emptying (including those on GLP-1 agonists), significant reflux, severe obesity, neurological impairment or emergency surgery.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 18 of 28

5 — FOAMED & CRITICAL APPRAISAL

METHODS FEATURE FINAL FRCA

Why Syndrome RCTs Keep Failing: The Fruit-Peeler Problem, and What the Evidence Actually Shows

A methods feature built around a critical care blog post. Opinion, the blog’s and ours, is labelled as such; every trial number below is from the primary source.

The argument. FOAMED In “The Fruit-Peeler RCT” (The Thoughtful Intensivist, Substack, 15 July 2026), Rafael Olivé Leite argues that sepsis and ARDS criteria are cause-agnostic. His analogy, credited to a post by “Dr Lynn” on X (we take this to be Lawrence Lynn, author of the review below), is a trial that enrols anything that is “a fruit” and “round” and peels them all: the apple looks good, the orange is harmed, and “depending on the case mix, the average effect may be beneficial, neutral, or harmful.” From this he concludes that such results lack transportability, that “all sepsis and ARDS RCTs ever performed are fruit-peeler RCTs”, and that “the field would never find a beneficial treatment for sepsis and ARDS, and the prediction holds.” He also criticises the PROVE Network for “15 RCTs in 15 years”. Our view: this is a polemic by a named author, not a study, and its core point about case mix is reasonable. Its strongest claim, that syndrome trials never find benefit, is the one to test. The formal version. Lynn’s single-author review in Patient Safety in Surgery (2026; PMID: 42464342; PMC full text) sets out the same idea mathematically. A trial gated on syndrome criteria estimates a “mean of causally unrelated means”: an average of disease-specific effects weighted by whatever mix of diseases a trial happens to enrol. If that mix shifts between trials, the same treatment can appear to help, do nothing or harm (his “cause-mixture paradox”), which he distinguishes from Simpson’s paradox because statistical adjustment alone cannot resolve it. Two points matter for readers. First, this is not independent support for the blog: Lynn is the source of the analogy, the blog author writes that he has argued it for a few years and Lynn for many more, and the review contains no new patient data (Lynn declares that he owns a company developing medical-record and time-series software). Second, Lynn is less absolute than the blog: he writes that the framework “does not suggest that all cause-agnostic trials are invalid”, and that pooling may be justified when different diseases truly converge on a shared physiological pathway that the intervention can modify. What the evidence shows: heterogeneity is real but modest. If the peeler problem were the whole story, splitting patients by biology should reveal large, consistent opposite effects. Four recent papers test that.

Sources: primary papers as listed. Venkatesan: PMC holds an author manuscript whose body could not be retrieved, so only the abstract was used.

Our view: read together, these papers support the blog’s diagnosis more than its prognosis. Biology does modify some treatment effects, and averaging over a mixture can hide that. But the credible signals are few, small and post hoc, the hyperinflammatory label showed no interaction for cisatracurium, rosuvastatin or IVIG (and a separate four-cluster classification none for convalescent plasma), and even well-defined subphenotypes contain wide differences in risk. Splitting the syndrome is a hypothesis to test, not yet a better way to treat.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 19 of 28

The counter-examples. Several cause-agnostic syndrome trials have found benefit, which is enough to refute “never”; ADRENAL is shown for contrast, as its primary outcome was null. Effect sizes below are from each trial’s PubMed abstract.

Absolute differences are crude percentage-point differences from the abstract rates, calculated by us without CIs.

These trials do not settle the argument entirely. Lynn himself cites lung-protective ventilation (ARMA) as an example of his “physiologic funnel” exception, where different causes share one vulnerability; PROSEVA’s physiological gate (severe ARDS only) arguably fits it too, though that is our reading. ARMA was also stopped early, which tends to overstate benefit. The steroid trials agree that shock resolves faster but disagree on mortality in different populations and regimens, which is the kind of non-transportable average the blog describes. Our view: syndrome trials can find real benefit when the intervention acts on a shared physiological problem, and are more likely to fail when it targets one pathway among many.

Bottom line (our view): be sceptical of both the trial average and the post hoc subgroup. Low tidal volume ventilation in all ARDS and prone positioning in moderate or severe ARDS remain established, and the FICM/ICS ARDS guideline still recommends both. Watch for prospective subphenotype-stratified trials such as PANTHER (ISRCTN81435672, recruiting in England since November 2025) before changing practice for any subphenotype.

How to read the next subphenotype paper.

ICEMAN: Instrument to assess the Credibility of Effect Modification Analyses (Schandelmaier et al, CMAJ 2020), as used by Van Nynatten et al.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 20 of 28

INFORMING PRACTICE

SR/MA

FINAL FRCA

Treatment Effects by Inflammatory Endotype in ARDS, Systematic Review (11 Secondary Analyses of 10 RCTs, n=5,514): 4 Showed Effect Modification, Only 1 (Liberal Fluids in Hyperinflammatory ARDS) Rated High Credibility

Van Nynatten LR, McChesney C, Basharat M, et al. Heterogeneity of treatment effect by molecular endotype in acute respiratory distress syndrome: a systematic review. Ann Am Thorac Soc. 2026 (advance article, published online 12 September 2026). doi:10.1093/annalsats/aaoag287. PMID: 42730647. PROSPERO CRD420261308321. Source route: accepted manuscript, open access (OUP, CC BY-NC). Appraised from full text

Systematic review of secondary and post hoc analyses of adult ARDS RCTs that tested whether a molecular endotype modified the treatment effect (MEDLINE, Embase and CENTRAL, January 2000 to 8 January 2026), with the credibility of each interaction rated by two reviewers using ICEMAN. Eleven analyses from ten trials were included; nine used the hyperinflammatory/hypoinflammatory classification from small plasma protein panels, in which the hyperinflammatory group was 26–40% of patients and carried higher mortality (29–62% vs 7.5–25%). By the review’s count, four analyses found a significant interaction, each driven by the hyperinflammatory group: higher PEEP, liberal fluids, vv-ECCO2R and simvastatin (a HARP-2 reanalysis, McKelvey 2022, p=0.01); three were suggestive only (open lung, CC16, and the original HARP-2 simvastatin analysis, p=0.14), and four found none (cisatracurium, rosuvastatin, convalescent plasma, IVIG). Only the FACTT fluid analysis was rated high credibility; results were not pooled.

• 11 / 5,514 SECONDARY ANALYSES / PATIENTS
• 4 of 11 SIGNIFICANT INTERACTION (REVIEW’S COUNT)
• 1 of 11 HIGH ICEMAN CREDIBILITY (FACTT FLUIDS)
• 0 ENDOTYPE-GUIDED STRATEGIES VALIDATED PROSPECTIVELY

Source: Van Nynatten et al, Table 2 and Results. Only 73 of 412 REST patients were endotyped (Table 1). CRITICAL APPRAISAL Question and protocol registered

Yes PROSPERO CRD420261308321, registered before abstract screening; molecular endotype required

Design: SR/MA | Certainty: low (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing yet: do not choose fluids, PEEP or statins by inflammatory subphenotype outside a trial | Confidence in that: high Main concerns: every analysis was secondary or post hoc and most were not adjusted for multiple testing, and null subgroup analyses may go unpublished, so the review likely overstates real effect modification; small subgroups (73 endotyped in REST) give large, imprecise effects such as 20.0% vs 62.5%, prone to inflation; endotype was measured once at enrolment although the hyperinflammatory state falls from about 26% to under 6% by day 3–4, so the direction of any misclassification bias is unclear Strengths: registered protocol with formal credibility grading of every interaction rather than counting p-values Disagreement: FICM/ICS suggest conservative fluids for all ARDS, whereas the FACTT reanalysis found lower mortality with liberal fluids in hyperinflammatory ARDS; untested prospectively, so the guideline stands.

UK practice: The FICM/ICS Guidelines on the management of ARDS (2018; Griffiths et al, BMJ Open Respir Res 2019), still the current ICS version, recommend lower tidal volumes in all ARDS and prone positioning in moderate or severe ARDS, suggest conservative fluids, and high PEEP in moderate or severe ARDS, and make ECCO2R a research recommendation; subphenotypes are not mentioned.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 21 of 28

INFORMING PRACTICE

OBSERVATIONAL

UK

FINAL FRCA

Bedside ARDS Subphenotyping in 30 UK and Irish ICUs (PHIND, n=490 Subphenotyped): a 1-Hour Near-Patient Assay Classed 18% as Hyperinflammatory, With 60-Day Mortality of 51% vs 28%

Reddy K, Sinha P, Antcliffe DB, et al, for the PHIND Investigators. Bedside identification of subphenotypes in acute respiratory failure (PHIND): a multicentre, observational cohort study. Lancet Respir Med. 2026;14(7):589–600. doi:10.1016/S2213-2600(26)00040-8. PMID: 41887245. ClinicalTrials.gov NCT04009330. Source route: open-access full text (publisher version of record, CC BY 4.0, via the Queen’s University Belfast repository). Appraised from full text

Prospective observational cohort in 30 NHS and Irish ICUs (November 2019 to September 2023) of adults within 72 hours of ARDS onset receiving invasive ventilation, NIV or high-flow nasal oxygen. Plasma IL-6 and soluble TNFR1 were measured at the bedside on a Randox multiSTAT analyser in about 1 hour and combined with arterial bicarbonate in a published logistic model (probability 0.5 or more = hyperinflammatory); site investigators were masked to the label. Of 1,853 screened, 525 were recruited; after 13 withdrew all consent and 22 assays failed, 490 were subphenotyped, of whom 89 (18%) were hyperinflammatory, and 486 had the primary outcome. The primary outcome, 60-day mortality, was 45/88 (51%) vs 111/398 (28%) (RR 1.8, 95% CI 1.4 to 2.4; unadjusted difference 23.2 points, 11.9 to 34.6; adjusted OR 2.7, 1.6 to 4.4), and bedside probabilities agreed closely with laboratory ELISA (intraclass correlation 0.94, 95% CI 0.92 to 0.95).

• 18% HYPERINFLAMMATORY BY NEAR-PATIENT ASSAY (89/490)
• 51% vs 28% 60-DAY MORTALITY, HYPER VS HYPO (45/88 VS 111/398)
• aOR 2.7 60-DAY MORTALITY (95% CI 1.6 TO 4.4); 2.33 (1.36 TO 3.97) WITH APACHE II AND P/F ADDED, POST HOC
• ICC 0.94 BEDSIDE VS ELISA PROBABILITY (0.92 TO 0.95)

Source: Reddy et al, Table 2 and Results. Differences are unadjusted. CRITICAL APPRAISAL Exposure (subphenotype) measured reliably

Partly ICC 0.94 vs ELISA, but the bedside assay under-identified hyperinflammatory patients and failed in 22/525 (4%)

Design: Observational | Certainty: moderate (GRADE-style, our judgement) | What it changes for UK anaesthesia and ICU: Nothing at the bedside yet: the label predicts outcome but has not been shown to change treatment; consider PANTHER recruitment | Confidence in that: high Main concerns: prevalence was 18% against an expected 25%, with sampling up to 72 h and an assay that under-calls vs ELISA, so some high-risk patients are labelled hypoinflammatory, likely understating how many could be enriched; the 0.5 cut-off is arbitrary (a Youden threshold of 0.274 exists), so a binary label hides risk near the threshold; funding from Randox, the assay maker, and Innovate UK may bias reporting towards assay performance, although the paper states the funders had no role in design, analysis or reporting Strengths: large prospective UK and Irish cohort with real-time classification, masked labels and complete primary outcome

UK practice: Neither the FICM/ICS Guidelines on the management of ARDS (2018; Griffiths et al, BMJ Open Respir Res 2019) nor any NICE guidance we found (site search, 2 Nov 2026) uses ARDS subphenotypes. The PANTHER platform trial (ISRCTN81435672; NIHR-funded, sponsor Imperial College London) randomises patients within each subphenotype to simvastatin 80 mg, baricitinib 4 mg or usual care, and began enrolling in England on 20 November 2025.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 22 of 28

6 — QUICK HITS

ICU: SEDATION, DELIRIUM AND RECOVERY ICU sedation and recalled experience (abstract only; planned secondary analysis of the A2B RCT, 289 of 1,009 90-day survivors): dexmedetomidine was associated with higher recalled awareness of surroundings than propofol (adjusted mean difference 3.6, 95% CI 1.0 to 6.2), clonidine was not (2.4, −0.2 to 5.0), and no other ICE-Q domain differed. Only 28.6% of survivors completed the questionnaire, and the authors say it does not support choosing dexmedetomidine over propofol. Masterson JA, et al. Anaesthesia 2026. PMID: 42838510 Music therapy during weaning (abstract only; RCT in a Spanish general ICU, n=100): daily individualised music therapy of about 25 minutes did not change weaning duration (p=0.97), the primary outcome. Agitation-sedation scores were better on days 1 and 3, fewer patients needed propofol (p=0.028) and tracheostomy was less common (reported as 14%, p=0.025; comparator rate and CIs not reported), but these are secondary outcomes in a small trial. Pereiro S, et al. Crit Care Med 2026. PMID: 42904879 Open ICU visiting and delirium (abstract only; SR/MA of 13 RCTs, n=3,861): open or flexible family presence was associated with less delirium than restricted visiting (RR 0.48, 95% CI 0.29 to 0.81) and slightly shorter ventilation (−0.36 days, −0.46 to −0.26), with no difference in hospital-acquired infection (RR 1.00, 0.81 to 1.23) or ICU mortality (RR 0.88, 0.72 to 1.08). Certainty was low or very low and heterogeneity substantial. Sella N, et al. Crit Care Med 2026. PMID: 42878614 Post-intensive care syndrome (consensus guideline, AAST): Patients in critical care for more than 48 h, or judged high risk, should be considered for mental health screening. Up to 69% of family members have clinically significant psychological symptoms in the months afterwards. Nohra EA, et al. Trauma Surg Acute Care Open 2026. PMID: 42921153

ICU: SHOCK, MONITORING AND ORGAN SUPPORT Arterial line and cuff disagree (review): A mean arterial pressure gap of more than 10 mmHg between the two should prompt structured checks of levelling, zeroing and damping before treating either number; a 10 cm levelling error alone shifts the reading by 7.5 mmHg. On vasopressors a radial line can read falsely low, so the authors suggest considering a femoral line if there is hypoperfusion. Zamith N, et al. J Clin Med 2025. PMID: 41464519 Brachial-to-radial pressure gradient on vasoactive drugs (prospective observational, n=303, 13 Polish ICUs) (abstract only): 18.4% had a non-invasive brachial-to-invasive radial gradient greater than 11 mmHg (median 4.5 mmHg). Vasoactive dose was the only independent association (OR 1.68, 95% CI 1.12 to 2.52), and ICU mortality was 48.0% vs 30.1%. Putowski Z, et al. Crit Care Med 2026. PMID: 42919451 POCUS in shock (review): A large UK study found that only 7% of ICU patients with shock had an echocardiogram within 1 h, and 43% within 24 h. The review sets out how POCUS helps in each major category of shock, and notes that GPICS states that all ICUs must have the equipment to provide it. Flower L, et al. BJA Educ 2026. PMID: 42924199 Lung ultrasound recommendations explained (commentary): The update separates diagnostic scanning from quantitative scoring and sets out three diagnostic levels. “Fluid tolerance” is framed as a safety margin, not a prompt to give fluid, and no validated quantitative thresholds exist yet for bedside use. Volpicelli G, et al. Intensive Care Med 2026. PMID: 42895537 Fluids in sepsis (review, Belfast): Restrictive fluid strategies have not changed mortality: CLASSIC (n=1,554) reported 90-day mortality of 42.3% vs 42.1%, and in CLOVERS (n=1,563) death before discharge home by 90 days was 14.0% vs 14.9%. The authors cover deresuscitation once shock resolves. McMullan RR, et al. BJA Educ 2026. PMID: 42453803 BICARICU-2 editorial (editorial): The trial was powered for an “implausibly large” 10% absolute mortality reduction, and 90-day mortality was 62.1% vs 61.7% (difference 0.4 percentage points, 95% CI −7.2 to 8.0). Kidney replacement therapy, a secondary outcome, was lower with bicarbonate (35% vs 50%), but this comes from an open-label design, which the editorialists note carries a risk of clinician bias. Kusirisin P, et al. JAMA 2025. PMID: 41159825 Haemoadsorption in cardiac surgery (SR/MA, 12 RCTs, n=713): No mortality difference against control (OR 1.08, 95% CI 0.59 to 2.01). Only one trial was at low risk of bias, and 10 used the CytoSorb device. Pittaway HCA, et al. Br J Anaesth 2026. PMID: 42481274 Polymyxin B haemoadsorption in endotoxic shock (consensus, US Delphi): The panel recommends considering it at an endotoxin activity of 0.60 or more and below 0.90, starting as early as possible, with 36 h or less from septic shock onset as a practical benchmark. The Surviving Sepsis Campaign suggests against polymyxin B haemoperfusion, and the consensus meeting was supported by Vantive. Yessayan L, et al. Crit Care 2026. PMID: 42928302

NEUROCRITICAL CARE, TRAUMA AND TOXICOLOGY Traumatic brain injury on ICU (review): The guideline intracranial pressure threshold has moved from above 20 to above 22 mmHg, and tiered treatment targets an ICP of 22 mmHg or less. Brain tissue oxygen-guided treatment is still being tested in phase III trials. Plummer MP, et al. BJA Educ 2026. PMID: 42764890 What’s new in trauma critical care (commentary, ESICM): Two prehospital trials of whole or component blood showed no mortality benefit, and the latest Cochrane review found no blood product strategy superior to another. The authors also note that thromboprophylaxis is often delayed despite evidence of safety once patients are stabilised, including after traumatic brain injury. Gauss T, et al. Intensive Care Med 2026. PMID: 42911398 High-dose insulin for beta-blocker and calcium channel blocker overdose (review, Royal Cornwall): Their critical care protocol gives a 1 unit kg−1 bolus of Actrapid, then 1 unit kg−1 h−1, uptitrated to a maximum of 10 units kg−1 h−1. Keep potassium above 3.5 mmol L−1 and glucose at 6–10 mmol L−1, with arterial sampling. Garland A, et al. BJA Educ 2026. PMID: 42453804

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 23 of 28

THEATRE: AIRWAY AND REGIONAL Capnography to check nasogastric tube position (prospective multicentre diagnostic study, 21 Hong Kong hospitals, n=508; abstract only): radiography found only 1 airway placement (0.2%), so the study cannot estimate how sensitive capnography is for the event that matters; the authors still call it a first-line tool alongside pH testing, which the 2016 NHS patient safety alert on nasogastric tube checks (pH, then x-ray) does not include. Chau JPC, et al. Crit Care 2026 (PMID not yet assigned). doi:10.1186/s13054-026-06246-9 PEEP set at BMI/3 during general anaesthesia (single-centre German RCT, n=60, non-laparoscopic non-cardiothoracic surgery): median driving pressure was 7.9 vs 8.9 cmH2O with PEEP 5 cmH2O (difference −1.0, 95% CI −1.9 to −0.16) and new atelectasis on lung ultrasound 0 vs 7 patients (ARD 0.23, 95% CI 0.043 to 0.43), but postoperative pulmonary complications did not differ (7 vs 11; RR 1.6, 95% CI 0.73 to 3.5, PEEP 5 vs BMI/3); outcomes are surrogates and no patient had a BMI above 40 kg/m2. Selpien H, et al. Anaesthesia 2025. PMID: 40551551 PENG vs supra-inguinal fascia iliaca block before spinal for hip fracture (abstract only; single-centre Chinese RCT, n=64, aged 65 or over): block success was 90.6% with PENG vs 68.8% (CI not reported; success not defined in the abstract), with lower pain scores and faster block and spinal performance; morphine use over 2 days and adverse effects did not differ, and the trial was observer-blinded only. Liao H, et al. J Clin Anesth 2025. PMID: 41330095 Erector spinae plane block (review): a BJA Education article on sonoanatomy, technique (typically 15–30 mL of long-acting local anaesthetic in adults), mechanisms, indications and complications. Its most useful point is that spread to the paravertebral space is inconsistent; the best evidence is for major spinal surgery, and paravertebral, interpectoral or TAP blocks are likely to be better for thoracic, breast, abdominal and caesarean surgery. De Cassai A, et al. BJA Educ 2026. PMID: 42630662 Regional anaesthesia and chronic postsurgical pain (SR and network meta-analysis, 158 RCTs, n=18,794; UK NIHR-funded): regional anaesthesia reduced chronic pain at 3 months or more vs no block (RR 0.73, 95% CI 0.67 to 0.80; I² 59.1%), with reductions after mastectomy, thoracotomy, VATS and knee arthroplasty, and neuraxial blocks ranked above peripheral blocks after thoracotomy and VATS; long-term opioid use did not differ (RR 0.88, 95% CI 0.61 to 1.28), certainty was low for chronic pain and very low for opioid use, and the funnel plot suggested small-study effects. Taylor-Rowan M, et al. Br J Anaesth 2026. PMID: 42425794

PERIOPERATIVE MEDICINE AND TRANSFUSION Cold-stored versus room-temperature platelets in cardiac surgery (CHIPS, phase 3 adaptive non-inferiority RCT, 27 US and Australian sites, n=1,000 children and adults; abstract only): platelets cold-stored for up to 21 days were non-inferior to room-temperature platelets on a 1 to 5 bleeding score (pooled mean difference 0.09, 95% credible interval −0.06 to 0.23; margin 1 point), with no difference in 24-hour chest drain output; the abstract reports more re-exploration with cold-stored platelets, and the 1-point margin is generous on a 5-point score. Spinella PC, et al. JAMA 2026. PMID: 42606886 Fluid-responsiveness-guided preinduction volume optimisation (single-centre RCT, Shanghai, n=100 adults aged 65 or over with frailty risk, elective GI surgery): passive leg raise with LVOT VTI, and 8 mL/kg gelatin boluses for responders (46 of 50), cut any MAP below 65 mmHg in the 15 minutes after induction from 72% to 38% (risk difference −34.0 points, 95% CI −50.1 to −14.5; NNT 3, 95% CI 2 to 7); the minimum MAP differed by only 4.3 mmHg and no postoperative benefit was shown; the colloid used was succinylated gelatin. Wang M, et al. J Clin Anesth 2026. PMID: 42854351 Fibrinogen concentrate and blood product use (SR/MA of 58 RCTs and observational studies, n=13,530, PROSPERO CRD42024602804): in adults, fibrinogen concentrate was associated with fewer red cell (IRR 0.72, 95% CI 0.59 to 0.89), plasma (IRR 0.69, 0.54 to 0.89) and platelet units (IRR 0.56, 0.33 to 0.95), largely driven by cardiac surgery studies, with no difference in mortality or thromboembolic events; comparators mixed placebo, no fibrinogen, FFP and cryoprecipitate, heterogeneity was high, and the review was funded by CSL Behring, a fibrinogen concentrate manufacturer, with an employee as author. Rahe-Meyer N, et al. Br J Anaesth 2026. PMID: 42864190 Stopping SGLT2 inhibitors before noncardiac surgery (secondary analysis of two Basel prospective cohorts, n=451 chronic users): 90-day acute heart failure admission or cardiovascular death occurred in 1/58 (1.7%) who continued, 10/177 (5.7%) stopped for 1 day, 13/155 (8.4%) for 2 days and 7/61 (11.5%) for 3 days or more (adjusted OR 1.58 per day stopped, 95% CI 1.08 to 2.30), with 1 euglycaemic DKA; observational, with only 58 continuers and likely confounding by who was told to stop, so it is a reason for a trial, not for abandoning current guideline advice to omit them. Haziri F, et al. Br J Anaesth 2026. PMID: 42270529 Cancellation and level-of-care changes in high- and moderate-risk surgery (UK multicentre retrospective cohort, REPACC secondary analysis, n=5,990 at 65 sites): 539 operations were cancelled (9.0% by our calculation; the paper prints 9.6%), and 275 of them (51.0%) for lack of resources, mostly beds; of 5,451 patients who had surgery, 3.9% were escalated and 14.8% de-escalated, mostly for clinical reasons, with bed shortage behind only 49 of 809 de-escalations. Reasons came from free-text records, and 263 of 809 de-escalations (32.5%) had no documented reason. Wong JKL, et al. J Intensive Care Soc 2026. PMID: 42871370 IV iron adverse events (NATA consensus recommendations with a structured literature review, 7 authors): it recommends pausing and then restarting more slowly for mild infusion reactions, no rechallenge after a suspected IgE-mediated reaction, saline flushes before and after infusion with follow-up for up to 8 weeks after extravasation, and weighing formulation-dependent hypophosphataemia (pooled 47%, 95% CI 36 to 58%, after ferric carboxymaltose). It suggests that routine 30-minute post-infusion observation is not required, which conflicts with the MHRA Drug Safety Update (December 2014) advice to observe patients for at least 30 minutes after every IV iron dose; the search was PubMed only, there was no formal Delphi, and several authors report payments from iron manufacturers. Saal-Bauernschubert L, et al. Br J Anaesth 2026;137:708–720. PMID: 42379899

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 24 of 28

CORE REVISION — ANAESTHETICS & ICU ISSUE 10 — NOVEMBER 2026

Core Revision — Malignant Hyperthermia: Recognition and Dantrolene

Pathophysiology, triggers, signs, crisis management, both dantrolene products, referral to Leeds — Final FRCA

Exam goal: Explain the cellular basis of malignant hyperthermia (MH), recognise a crisis from its pattern of signs, run the treatment with weight-based dantrolene doses for whichever product your theatre stocks, and describe the referral and testing pathway after the event. The table compares the Association of Anaesthetists (2020) and European Malignant Hyperthermia Group (EMHG, 2024) guidelines.

1. PATHOPHYSIOLOGY AND HOW COMMON IT IS

MH susceptibility is classically described as autosomal dominant (each first-degree relative has a 50% chance of carrying the variant), although a 2026 UK analysis found that the clinical risk of reacting fits a threshold model better. The genes implicated (EMHG) are RYR1 (the ryanodine receptor, the calcium release channel of the sarcoplasmic reticulum), CACNA1S (the dihydropyridine receptor, DHPR) and STAC3 (which links RYR1 and the DHPR). In a susceptible patient, a volatile agent or suxamethonium releases excess calcium into the muscle cell. Energy-dependent ATPases work to clear it, driving up oxygen consumption, CO2 production, heat and heart rate. When calcium homeostasis fails, the muscle contracts, leaks potassium and then myoglobin and creatine kinase (CK), and the acidosis becomes mixed. Dantrolene binds RYR1 and suppresses calcium release from the sarcoplasmic reticulum. The Association of Anaesthetists 2020 guideline puts UK incidence at about 1 in 100,000 general anaesthetics (about 20 confirmed cases a year against about 3 million anaesthetics from NAP5 and NAP6), or about 1 in 250,000 if one anaesthetic per 10 people a year is used as the denominator; UK mortality is about 4%. The 2026 UK modelling paper in this issue estimates 7.8 reactions per million general anaesthetics (1 in 128,000) for 1988–93, and a prevalence of clinical MH risk of 1 in 38,800 (95% credibility interval 1 in 34,600 to 1 in 43,700).

2. TRIGGERS, SAFE AGENTS AND PATIENTS AT RISK

Triggers are all volatile anaesthetic agents and suxamethonium. Safe options are regional anaesthesia or total intravenous anaesthesia (TIVA). Prepare the machine by flushing with 100% oxygen to below 5 ppm of volatile, or fit activated charcoal filters to both limbs; every UK hospital giving general anaesthesia should have these filters. Treat as at increased risk (Association of Anaesthetists, Box 5): confirmed MH susceptibility (a pathogenic variant or a positive in vitro contracture test, IVCT); blood relatives of a confirmed case; a suggestive personal or family anaesthetic history; a myopathy caused by a gene implicated in MH (RYR1, CACNA1S, STAC3, as named by EMHG); a variant of unknown significance in those genes; unexplained or recurrent rhabdomyolysis; idiopathic raised CK; and unexplained exertional heat illness. EMHG 2026 adds that patients with an RYR1-related myopathy, a STAC3 myopathy or a CACNA1S/CACNB1 myopathy should not receive volatiles unless their own IVCT is MH-negative. Suxamethonium is contraindicated in congenital myopathies; use rocuronium for rapid sequence induction and keep sugammadex available. Prophylactic dantrolene is not recommended, because it worsens weakness. In muscular dystrophies the concern is perioperative rhabdomyolysis: volatiles may be used at the lowest dose for the shortest time, but TIVA is preferred.

3. RECOGNITION: EARLY AND LATE SIGNS

No single sign is diagnostic; the pattern matters. The cardinal triad (Association of Anaesthetists, Box 1) is an unexplained, unexpected rise in end-tidal CO2, heart rate and temperature. Early signs (EMHG 2024) are raised CO2 production or tachypnoea, increased O2 consumption, mixed acidosis, sweating and mottling; tachycardia, ventricular ectopics or bigeminy and unstable blood pressure; and masseter spasm after suxamethonium or generalised rigidity. Late signs are hyperkalaemia, a rapid rise in core temperature, grossly raised CK and myoglobin, dark urine, severe arrhythmias or arrest, and disseminated intravascular coagulation (DIC). Timing: onset can be within 10 min of exposure or hours later, especially with desflurane. A temperature rise that starts before any rise in CO2 production is not MH. Differentials (EMHG 2024) include light anaesthesia, sepsis, inadequate ventilation, machine fault, anaphylaxis, phaeochromocytoma, thyroid crisis, laparoscopic CO2, ecstasy, neuroleptic malignant syndrome and serotonin syndrome (new in 2024). In an undiagnosed myopathy, a rapidly rising end-tidal CO2 and mixed acidosis point to MH. Sudden arrest or cola-coloured urine points to rhabdomyolysis, where the priority is cardiac output while lowering potassium.

4. MANAGING THE CRISIS

Declare an emergency and call for help. Three things happen together: eliminate the trigger, give IV dantrolene and start active cooling. EMHG advises asking the surgeon to stop or postpone surgery. The UK guideline allows surgery to be completed under IV anaesthesia once the reaction is controlled and its consequences need no further management.

In a UK theatre (Association of Anaesthetists 2020): give dantrolene 2–3 mg kg−1 on actual body weight now, giving each syringe as soon as it is made up; then 1 mg kg−1 every 5 min until end-tidal CO2 is below 6 kPa with normal minute ventilation and core temperature is below 38.5 °C; formally reappraise the diagnosis at a cumulative 10 mg kg−1, but continue if MH is still likely. The guideline sets no per-dose cap.

Worked examples (calculated, at 2.5 mg kg−1, which is inside the UK range and is the SmPC first dose for both products; volumes are reconstituted solution: Dantrium 20 mg in 60 mL, Agilus 120 mg in 22.6 mL). 70 kg: first dose 175 mg = 8.75 Dantrium vials (525 mL; make up 9) or 33.0 mL of Agilus (make up 2 vials). Each 1 mg kg−1 repeat is 70 mg = 3.5 Dantrium vials (210 mL) or 13.2 mL of Agilus. 10 mg kg−1 is 700 mg = 35 Dantrium vials or 5.8 Agilus vials.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 25 of 28

100 kg: first dose 250 mg = 12.5 Dantrium vials (750 mL; make up 13) or 47.1 mL of Agilus (make up 3 vials). Each repeat is 100 mg = 5 Dantrium vials (300 mL) or 18.8 mL of Agilus. 10 mg kg−1 is 1,000 mg = 50 Dantrium vials or 8.3 Agilus vials. The 300 mg cap per dose comes from EMHG 2024 and the Agilus SmPC, not the UK guideline. At 2.5 mg kg−1 it limits only patients above 120 kg (the SmPC table caps 120 kg and 144 kg alike at 300 mg, 56.5 mL); at the UK upper dose of 3 mg kg−1, a 100 kg

patient’s 300 mg sits exactly at it (calculated).

Other consequences (Association of Anaesthetists): treat arrhythmias with amiodarone, a short-acting beta-blocker or magnesium. Aim for a urine output above 2 mL kg−1 h−1 and give bicarbonate for myoglobinuria. Treat DIC with platelets, FFP and cryoprecipitate (not

tranexamic acid). CK may not peak for 24 h; watch clinically for compartment syndrome (fasciotomy). The main disagreement: the UK guideline reserves calcium for extremis, as extra calcium may add to the overload in the muscle cell; EMHG lists calcium chloride as an option. The two formulations licensed in Great Britain (eMC SmPCs; SmPC regimens for comparison). Dantrium IV 20 mg (PL 20011/0034): dissolve each vial in 60 mL of water and draw it up through the filter supplied. Each vial contains 3 g of mannitol. SmPC dose: at least 2.5 mg kg−1 by rapid injection; 10 mg kg−1 per 24 h is usually enough. Agilus 120 mg (GB licence PLGB 20011/0075, first authorised 8 July 2024): 20 mL of water gives 22.6 mL at 5.3 mg mL−1. SmPC dose: 2.5 mg kg−1 rapidly every 10 min while signs persist, no single dose above 300 mg (2.5 vials). In the bioavailability study a vial reconstituted in a mean of 50 s, against 90 s for a 20 mg vial. It contains hydroxypropylbetadex (3530 mg per vial), with a potential risk of hearing impairment. No clinical efficacy studies of Agilus were done (that study had 21 volunteers). The 2020 guideline predates Agilus (it calls Ryanodex unlicensed in the UK), so know which product your theatre stocks.

5. AFTER THE CRISIS

Before discharge, give the patient written information and ask them to warn all blood relatives. Point them to the UK MH Registry (ukmhr.ac.uk). The consultant anaesthetist in charge must refer directly to the MH Investigation Unit in Leeds, which serves the UK and Ireland, sending the perioperative records and a narrative account. Testing is nationally commissioned in England; elsewhere it needs funding approval. Testing: DNA screening on a blood sample taken by the GP usually comes first, but it detects only about 50% of susceptibility. If no variant is found, patients aged over 10 years have an IVCT on freshly excised muscle at Leeds. For an index case under 10, both parents are offered biopsy. Biopsy is not done within 4 months of a reaction, and the result is given on the day. Those at risk are advised to wear a MedicAlert or similar tag and given warning cards. The EMHG 2025 investigation guideline adds a new “MHG” (MH genotype) designation alongside “MHS” (MH susceptible). It also defines clinical MH as limited to the classic triggers.

FINAL FRCA FOCUS Genes and dantrolene at RYR1; the triad of rising end-tidal CO2, heart rate and temperature; remove trigger,

dantrolene and cool together; first dose 2–3 mg kg−1 (UK) or 2–2.5 mg kg−1 up to 300 mg (EMHG); calcium only in extremis (UK); refer to Leeds, DNA then IVCT, warn the family. Sources: [S1] Snoeck MMJ et al, EMHG muscle disorders consensus, Br J Anaesth 2026, PMID 42532738; [S2] Hopkins PM et al, Association of Anaesthetists MH guideline 2020, Anaesthesia 2021;76:655–664, PMID 33399225; [S3] Glahn KPE et al, EMHG crisis guideline 2024, Br J Anaesth 2025, PMID 39482150; [S4, S5] Agilus 120 mg and Dantrium IV 20 mg SmPCs, eMC; [S6] Aboelsaod EM et al, Br J Anaesth 2026, PMID 42850084; [S7, S8] Rüffert H et al, EMHG investigation guideline 2025, Br J Anaesth 2026, PMID 41478797, and authors’ correspondence.

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 26 of 28

7 — ACTION POINTS

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 27 of 28

TRIALS TO WATCH

REPORTING SOON

SINFONIA UK trial of sugammadex vs neostigmine reversal to prevent pulmonary complications after major surgery (ISRCTN15109717; sponsor Belfast HSC Trust, NIHR HTA funded). Target 2,500 patients; recruitment closed 1 November 2025 and the study ended 1 November 2026. The primary outcome is days alive and out of hospital at 30 days. Only protocol papers were in PubMed on 2 November 2026.

WITHIN THE NEXT YEAR

LONGER TERM

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

Anaesthetics & ICU Evidence Rundown, Issue 10, November 2026. Page 28 of 28