This Week in Anesthesiology — Sep 28, 2026

This AudioScholar briefing is generated by artificial intelligence for healthcare professionals and trainees. It is not medical advice.

Welcome to This Week in Anesthesiology. This week we're covering 10 notable papers spanning perioperative haemodynamics and oxygen delivery, the limits of our favourite pharmacological adjuncts, airway practice in the smallest patients, and the systems side of anaesthesia — from greenhouse gas emissions to clinical decision support. Let's dive in.

We start with blood pressure and oxygen, two of the levers we pull most often, and two papers that pull in different directions. In Anesthesiology, Ripollés-Melchor and colleagues report a Bayesian random-effects meta-analysis of fifteen randomised trials of protocolised intraoperative arterial pressure targeting in major noncardiac surgery [1]. For all-cause mortality across thirteen trials, the estimate sat almost exactly on no effect, with prediction intervals wide enough to accommodate both meaningful benefit and meaningful harm. Myocardial injury after noncardiac surgery, across eight trials, was likewise compatible with no treatment effect. Acute kidney injury gave the most favourable signal, with a point estimate suggesting roughly a tenth lower odds, but the authors are explicit that this remained compatible with no clinically important benefit, that heterogeneity was moderate, and that dropping any single trial moved the result. Their conclusion is that protocolised pressure targeting has not demonstrated consistent incremental benefit over contemporary usual care, and that any renal benefit may depend on the wider haemodynamic context in which a target is applied — a reminder that a number on a monitor is not the same as perfusion. That framing sits interestingly beside a retrospective cohort of 4,358 adult cardiac surgical patients, also in Anesthesiology, from Robitaille and colleagues, looking at indexed oxygen delivery during cardiopulmonary bypass [4]. Here, where flow, haemoglobin and oxygenation are combined into a single delivery variable, the association with outcome was strong. Each 10 millilitre per minute per square metre fall in time-weighted average oxygen delivery was linked to about a sixteen percent rise in the odds of 30-day mortality, and adjusted mortality was roughly twice as high at the lower quartile of delivery as at the upper. Lower delivery also tracked with acute kidney injury and longer intensive care and hospital stays, though not with stroke. Seventy-five patients died within thirty days, so these are modest absolute numbers in a single-centre observational dataset, and the authors position their work as a rationale for mechanistic and interventional studies rather than a mandate. Taken together, the week's evidence suggests that pressure alone may be the wrong target, while composite oxygen delivery remains a hypothesis worth testing prospectively.

The second theme is pharmacology that underperforms its reputation. In Anesthesia and Analgesia, Desverreaux and colleagues used a large United States electronic health record network to propensity-match patients undergoing lumbar spine surgery — nearly fifty-four thousand opioid-naive patients and just over two thousand with opioid use disorder — and asked whether intraoperative ketamine reduced chronic opioid dispensing [7]. It did not. In opioid-naive patients, ketamine exposure was actually associated with somewhat more outpatient opioid prescribing at forty-eight hours and at six weeks, with no meaningful differences at three, six or twelve months; in the opioid use disorder group the excess was confined to the first forty-eight hours. Opioid-related adverse events and new diagnoses of opioid use disorder were not reduced and in places were more frequent. The authors are careful: this is observational, dose and protocol are invisible in the database, and residual confounding by indication is the obvious explanation for early prescribing differences. What it does show is the absence of any population-level opioid-sparing signal out to a year. A companion note in the British Journal of Anaesthesia from Ahrens and Schaefer makes a related point about dexamethasone [3]. Commenting on a secondary analysis of the individualised perioperative open-lung ventilation trials, they highlight that antiemetic doses of four to eight milligrams did not reduce postoperative pulmonary complications — and argue this is pathophysiologically unsurprising, because the dominant driver of those complications is atelectasis determined by surgery, induction and ventilatory management, not systemic inflammation. The editorialists reaffirm the established safety of standard-dose dexamethasone while cautioning against expecting an anti-inflammatory drug to fix a mechanical problem. In the same journal, Pitesa and colleagues argue the converse case for minimal access abdominal surgery: as the somatic wound shrinks, peritoneal inflammation and visceral pain dominate recovery, and they position non-steroidal anti-inflammatory drugs and corticosteroids as proven foundations with intraperitoneal local anaesthetic instillation as a complementary adjunct reported to reduce visceral pain scores, opioid consumption and delays in gastrointestinal recovery [10]. That is a narrative argument for a paradigm shift, not a trial, and the authors themselves flag delivery method and patient-reported outcomes as open research priorities.

Transfusion and the airway supply the next pairing, and both deliver results that cut against expectation. Also in the British Journal of Anaesthesia, Butt and colleagues review two pragmatic randomised trials of prehospital low-titre group O whole blood in traumatic haemorrhage — TOWAR, which cluster-randomised forty-four North American air medical bases, and SWiFT, which randomised individual patients across ten English air ambulance services [6]. Both returned the same null: at a dose of two units, whole blood was not superior to conventional components, and neither trial signalled meaningful harm. The commentators note that TOWAR's cluster design left the whole blood arm sicker at baseline and could bias its estimate, but cannot explain why a cleanly randomised independent trial landed in the same place — hence their argument that the burden of proof has shifted. They also point out that neither trial compared whole blood against goal-directed factor concentrate resuscitation. On the airway, Ukiya and colleagues analysed 3,250 intubation encounters in neonates and infants from the Japanese paediatric difficult airway registry, with propensity-weighted adjustment [5]. Videolaryngoscopy produced substantially better glottic views, yet was associated with about eleven percent lower relative probability of first-attempt success than direct laryngoscopy, and with roughly double the risk of desaturation; respiratory adverse events were similar. The authors call this an efficacy–effectiveness gap in non-standardised real-world practice, and suggest structured implementation and targeted training may be needed before better visualisation translates into successful first attempts. This is registry data with the confounding that implies, and it should not be read as overturning trial evidence, but it is a caution about assuming device performance transfers automatically to everyday practice.

Finally, three papers on systems. A Welsh national retrospective cohort in the British Journal of Anaesthesia from Evans and colleagues covered 46,614 adults undergoing major elective surgery [2]. Just under a third of patients were anaemic, and about half of those anaemic patients had iron deficiency. After adjustment, preoperative anaemia was independently associated with longer stay, roughly a fifth higher odds of thirty-day readmission, and about seventy percent higher odds of death, with mortality elevated at every time point out to five years; iron deficiency and anaemia with inflammation marked the worst outcomes. This is association, not causation, but the authors argue it carries implications for routine preoperative screening. In Anesthesia and Analgesia, Sutin and colleagues report an interrupted time-series across eleven New York City public hospitals and nearly 149,000 anaesthetics [8]. The striking finding is that at one hospital, about ninety-five percent of centrally supplied nitrous oxide was lost before reaching any patient, generating an estimated 2,136 metric tonnes of carbon dioxide equivalent a year — more than twice the combined emissions from every clinically administered anaesthetic gas across all eleven sites. Deactivating central pipelines produced the dominant reduction in nitrous-attributable emissions, although the association with total emissions in the primary model was not statistically significant. A clinical decision support advisory produced a modest but significant fall in fresh gas flow. And in Anesthesiology, Murphy and colleagues tested in-line decision support flagging fifty-one conditions warranting an ASA physical status of at least three, across more than half a million anaesthetics at two academic centres [9]. At the intervention site the proportion classified three or higher rose roughly nine percentage points, to just over sixty percent, against almost no change at the control site — and among patients with the targeted conditions, classification rose to ninety-four percent. Whether that represents accuracy or simply upward drift in documentation is the question the design cannot fully settle.

If you only have time for one paper this week, make it the Bayesian meta-analysis of intraoperative blood pressure targeting in Anesthesiology [1]. It reopens one of the most confidently held assumptions in perioperative medicine — that protocolised pressure targets improve hard outcomes — and shows the randomised evidence is compatible with no incremental benefit over usual care.

Here is what this week's evidence adds up to in Anesthesiology. First, pooled randomised data do not support a consistent outcome benefit from protocolised intraoperative pressure targeting, while observational data on bypass suggest composite oxygen delivery may be the more meaningful variable — a hypothesis, not yet a target. Second, two of our habitual adjuncts underperformed: intraoperative ketamine showed no population-level reduction in chronic opioid use after lumbar spine surgery in a large database, and antiemetic-dose dexamethasone did not reduce pulmonary complications, plausibly because those complications are mechanical rather than inflammatory. Third, two independent randomised trials of prehospital whole blood at a two-unit dose agree on a null result, and registry data raise the possibility that videolaryngoscopy in neonates and infants underperforms direct laryngoscopy when implementation is unstandardised. And fourth, on the systems side, preoperative anaemia remains common and strongly associated with worse outcomes at every time point, while nitrous oxide pipeline losses appear to dwarf clinically administered gas emissions — findings the authors link to existing patient blood management and American Society of Anesthesiologists sustainability recommendations.

That's your roundup for This Week in Anesthesiology. 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 it saves you time, a quick rating in your podcast app helps other physicians find the show.

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