This Week in Critical Care — Sep 8, 2026

Welcome to This Week in Critical Care. This week we're covering 10 notable papers spanning sedation and liberation from mechanical ventilation, bedside physiology for shock and kidney injury, resuscitation rhythms and phenomena, and the shifting epidemiology of severe viral illness. Let's dive in.

We start with the ventilated patient, where three papers ask how we sedate, how we get people off the ventilator, and how we predict who will fail non-invasive support. In Critical Care Medicine, Casamento and colleagues report a double-blind, placebo-controlled trial of low-dose ketamine as an adjunct for analgosedation at two university ICUs in Melbourne. One hundred and twenty ventilated adults already on an opioid infusion were randomised to ketamine at 0.15 milligrams per kilogram per hour or placebo. Hourly opioid dose in fentanyl equivalents was lower with ketamine, a median difference of about 13 micrograms per hour, and the Bayesian analysis put the probability of some opioid-sparing benefit at roughly 95 percent, though the credible interval still crossed no effect. Importantly, delirium, ventilator-free days, ICU and hospital-free days, and serious adverse events did not differ. So this is a safety-and-feasibility signal for opioid sparing, not evidence of a patient-centred benefit, and the authors say plainly that larger trials are needed before ketamine becomes routine.

Staying with liberation from the ventilator, Critical Care published the X-COPD trial from Karagiannidis and colleagues, testing whether extracorporeal carbon dioxide removal could facilitate early extubation in severe COPD exacerbations requiring invasive ventilation. This trial deserves attention mainly as a cautionary tale: the sponsor terminated it for financial reasons after just 18 of a planned 192 patients had been randomised. The primary composite of death or severe disability at day 60 occurred in none of the eight extracorporeal patients and in three of nine controls, a difference that was not statistically significant. Ventilation duration was shorter with the device, a mean difference of about 17 days, and there were no cases of ventilator-associated pneumonia in the extracorporeal arm versus three in the control arm, but with eight patients per group these are hypothesis-generating observations only, and one extracorporeal patient had severe bleeding. Nothing here should change practice; it should, however, sustain the case for an adequately powered multicentre trial.

The third ventilation paper, in the Journal of Critical Care, looks upstream at high-flow nasal cannula failure. Yang and colleagues performed serial right-sided diaphragm ultrasound at high-flow initiation and at one, three, six, nine and twelve hours in 270 adults with acute hypoxaemic respiratory failure, of whom just under a quarter were intubated within 48 hours. Diaphragmatic thickening fraction tracked risk: each standard-deviation rise raised the odds of subsequent intubation by about half, after adjustment for age and severity of illness. Diaphragmatic excursion, the measure many of us find easier to obtain, carried no independent signal. The ROX index remained the stronger bedside marker, and adding thickening fraction to ROX improved discrimination only modestly. The practical message is that if you are going to put a probe on the diaphragm during high-flow therapy, measure thickening fraction rather than excursion, and treat it as complementary to ROX and clinical judgement rather than a replacement.

Our second theme is bedside physiology for shock and kidney injury, and here three papers push toward serial, dynamic measurement rather than single snapshots. In Critical Care Medicine, Dieiev and colleagues report from the multicentre resuscitative transoesophageal echocardiography registry across 23 hospitals, examining what happens when transoesophageal lung ultrasound is added to resuscitative echocardiography in ventilated adults with shock. Of 379 qualifying examinations, about a quarter included lung imaging, and operators reported identifying a shock aetiology more often when the lungs were interrogated, roughly 87 percent versus 75 percent, which corresponds to nearly double the odds. Reported changes in management were more frequent but not statistically significantly so, and the addition cost only about two and a half minutes of procedure time with no excess complications. This is operator-reported and unadjudicated, so treat it as feasibility data supporting a technique rather than proof of benefit.

Turning to the kidney, Shock published a prospective multicentre cohort from Fu and colleagues across six Chinese ICUs, enrolling 456 critically ill adults with or at risk of acute kidney injury and performing Doppler-derived semi-quantitative renal perfusion and resistive index at admission, 24 and 48 hours. Trajectory modelling split patients almost evenly into sustained-high and sustained-low perfusion phenotypes. Sustained-high perfusion was associated with roughly 44 percent lower odds of persistent kidney injury, with less non-recovery at seven days and less renal replacement therapy on days three and seven, although the primary association was attenuated in expanded sensitivity models. Notably, resistive index trajectories did not discriminate renal outcomes at all, which is a useful negative for a measure many units still favour. Alongside this, Shock also published a multicentre cohort from Han and colleagues of over 3,500 patients with community-acquired sepsis, examining the change in arterial pH between the emergency department and the first 24 hours in the ICU. Falling pH beyond baseline-specific thresholds was associated with roughly a tripling of 28-day mortality, and the signal was strongest, more than a four-fold hazard, in patients whose emergency department pH had been normal. The lesson for the receiving intensivist is that the trend matters more than the admission value, and that a seemingly stable patient with a normal presenting pH who then drifts downward is the one to worry about.

Our third theme is resuscitation, with two papers in Resuscitation that both challenge conventional teaching. Wong and colleagues interrogated the American Heart Association Get With The Guidelines-Resuscitation registry across more than two decades, analysing over 9,400 children with in-hospital cardiac arrest whose initial rhythm was pulseless electrical activity or asystole. Pulseless electrical activity was the more common rhythm, and asystole carried roughly 30 percent lower odds of return of spontaneous circulation and of restoration of circulation. But the apparent survival advantage of pulseless electrical activity essentially disappeared once neonatal intensive care events were excluded, and age-stratified analysis showed the difference was driven by neonates. So lumping these rhythms together as simply non-shockable obscures real differences in short-term resuscitation physiology, while for survival to discharge in older children the two rhythms look alike. In the same journal, Howard and colleagues examined cardiopulmonary resuscitation-induced consciousness in more than 43,000 out-of-hospital arrests in Victoria over 16 years. Just over three percent of patients showed it, and the incidence rose by about eight percent per year, plausibly reflecting better bystander response and mechanical compression use. After adjustment, consciousness during compressions was associated with nearly triple the odds of survival to hospital discharge, and that held for both interfering and non-interfering presentations. This is a marker of favourable physiology and effective perfusion, not a nuisance to be sedated away reflexively, though the observational design cannot tell us how best to manage it.

Finally, two papers on the changing face of severe viral illness. In Critical Care Medicine, Beck-Friis and colleagues used Swedish nationwide registry data covering more than 92,000 COVID-19 hospitalisations across five pandemic periods and over 21,000 influenza hospitalisations. ICU admission rates were lowest in the Omicron period, and that period came to resemble seasonal influenza in patient age, length of stay, ICU use and in-hospital mortality, with about half of Omicron patients aged 75 or older. Advanced age was the dominant predictor of severe outcomes in both infections, and for COVID-19 being unvaccinated remained independently associated with both ICU admission and death. Meanwhile Intensive Care Medicine published a national Chilean cohort from Meza-Fuentes and colleagues on Andes virus hantavirus cardiopulmonary syndrome, spanning 31 hospitals and 215 patients. Of the 148 who needed respiratory support, roughly half required extracorporeal membrane oxygenation, overall in-hospital mortality was about 21 percent, and case fatality was around 30 percent for those on invasive ventilation and a third for those on extracorporeal support. Strikingly, admission during the COVID-19 pandemic period was independently associated with nearly triple the odds of dying, a reminder of what happens to a resource-intensive, low-incidence disease when critical care capacity is consumed elsewhere.

If you only have time for one paper this week, make it the ketamine analgosedation trial in Critical Care Medicine [1]. It is the week's only adequately blinded randomised trial addressing a decision most of us make every day, and its honest, modest result — opioid sparing without demonstrated patient-centred benefit — is exactly the nuance to carry to the bedside.

Here are the key takeaways from this week in Critical Care. First, low-dose ketamine appears safe and modestly opioid-sparing in ventilated adults, but delirium and ventilator-free days were unchanged, so do not expect downstream gains. Second, when you use ultrasound to predict high-flow failure, diaphragmatic thickening fraction carries information and excursion does not, but ROX remains the better bedside marker. Third, in sepsis the trajectory of arterial pH from emergency department to ICU predicts death better than the presenting value, and the highest-risk patient may be the one who arrived with a normal pH. Fourth, in resuscitation, pulseless electrical activity and asystole in children are not physiologically interchangeable, and consciousness during compressions in out-of-hospital arrest marks a patient likely to survive. And fifth, the extracorporeal carbon dioxide removal trial in COPD was stopped after 18 patients and answers nothing definitively — a shorter ventilation duration in eight patients is not a reason to change practice.

That's your roundup for This Week in Critical Care. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And if you have a paper you have been meaning to read, upload the PDF, or paste any link, at audioscholar dot C C. We will turn it into audio like this one, in any of thirty-one languages.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.