A randomised trial where the higher blood pressure target did worse

A thin day in the priority journals — the sweep returned two editorials and a correspondence letter. Both items below came from a wider search across all journals for on-scope trials and cohorts, which is worth doing when the named-journal sweep comes up empty.

1. Blood pressure targets during general anaesthesia — randomised feasibility trial

Journal · Published: Acta Anaesthesiologica Scandinavica, 1 September 2026 — 6 days old Pælestik, Risager, Mortensen, Schjørring, Bisgaard et al. (Denmark). NCT06047119.

Multicentre, randomised, factorial trial across eight Danish hospitals. 483 patients, ASA 3–5, scheduled for major surgery under general anaesthesia, randomised to a minimum MAP target of 60, 70 or 80 mmHg, or 90% of baseline systolic pressure. Primary outcomes were feasibility; clinical outcomes were secondary.

The authors are explicit about why feasibility was the question: “Previous trials on blood pressure targets have suffered from poor adherence and limited between-group separation.”

• 483 of 1,690 eligible patients randomised (29%) — below the 50% goal
• Separation in blood pressure was achieved, and pressure during vasoactive treatment stayed within predefined target ranges
• No difference in most postoperative outcomes
• But compared with MAP 60, the MAP 80 group had more acute kidney injury (OR 4.74, 95% CI 1.17–31.8) and more cardiac arrhythmias (OR 8.13, 95% CI 1.35–156)
• MAP 80 had fewer days alive and out of hospital at 30 days — median difference −1.88 days (95% CI −3.50 to −0.26)

Interpretation. Note first what this trial fixed. The BJA intraoperative hypotension meta-analysis found that trials in this field achieved MAP separation of usually under 10 mmHg, which meant most of them were not really testing their question. This one achieved separation and adherence — that is the point of the paper, and it is a genuine methodological advance.

Then note the direction. The higher-pressure arm did worse, on kidney injury, arrhythmia and days at home. That is the opposite of the prevailing assumption that intraoperative hypotension is the modifiable harm and a higher target is the safe choice.

Now be careful, because those confidence intervals are enormous — an odds ratio for arrhythmia running from 1.35 to 156 is a handful of events, not an effect estimate. This is a feasibility trial with 483 patients split four ways, and its clinical outcomes are secondary. It cannot establish that MAP 80 causes harm.

What it can do is make the mechanism plausible, and here it connects to two other papers in this archive. The off-pump tissue oxygenation study showed the MAP–cerebral oxygenation slope flattening across 70–100 mmHg — an autoregulatory plateau where extra pressure buys nothing. The pulse pressure and AKI cohort showed a flow surrogate predicting kidney injury at MAPs we call adequate. Chasing a MAP of 80 in an ASA 3–5 patient means vasoconstrictors, and vasoconstriction raises pressure by reducing flow. A signal of more AKI in the highest-target arm is exactly what that physiology predicts.

Worth watching for the full trial. If it replicates, a decade of “avoid hypotension” teaching acquires an upper bound.

Read the paper · PMID 42578528

2. Rewarming rate during cardiopulmonary bypass and postoperative delirium

Journal · Published: Perfusion, 4 September 2026 — 3 days old Xu, Xu, Zhu, Wang, Shi

Retrospective cohort, 548 adult cardiac surgery patients, January 2021 – December

1. Rewarming rate defined as the average temperature rise per minute from the lowest nasopharyngeal temperature to 36.5°C.

• Rewarming was faster in the delirium group: 0.095 ± 0.044 vs 0.072 ± 0.034 °C/min
• Predictive model AUC 0.775 (95% CI 0.736–0.814)
• Conclusion: rapid rewarming is associated with increased delirium risk and “may be a modifiable perioperative factor”

The deposited abstract has its p-values stripped by the markup, so exact significance values are not quoted here.

Interpretation. The hypothesis is old and biologically sound — rapid rewarming causes cerebral hyperthermia, and jugular bulb desaturation during rewarming has been documented since the 1990s. Slow, controlled rewarming with a limited arterial-to-venous gradient is already in most perfusion protocols for that reason.

The problem is the outcome definition, and it is serious enough to lead with. Delirium was identified by DSM-5 review or by documented symptoms combined with olanzapine use as a surrogate marker — and 300 of 548 patients (55%) were classified as delirious. A 55% delirium rate is far above what prospective CAM-ICU screening typically finds after cardiac surgery, which suggests the olanzapine surrogate is capturing sedation and agitation prescribing rather than delirium as such. If faster-rewarmed patients are sicker or longer on bypass, they are also more likely to be prescribed olanzapine for reasons other than delirium.

So: a plausible mechanism, a real association, and a measurement that cannot carry the conclusion. The authors ask for prospective studies with standardised screening, which is the right request. In the meantime, controlled rewarming remains good practice on the older physiological grounds rather than on this.

Read the paper · PMID 42696636

---

Dedupe note: the wider search also returned MERCURI-2 (JAMA, dapagliflozin and AKI after cardiac surgery) and ITACS (BMJ, IV iron before cardiac surgery) with September dates. Both are the issue versions of papers already in this archive — MERCURI-2 sent 24 August, ITACS 25 August. Neither is new; the archive’s items: dedupe caught both.