A quiet Saturday worked from the backlog: minimally invasive extracorporeal circulation in 41 trials, DNR status and withheld rib fracture blocks, and a dexmedetomidine preprint

A note on today’s search, because the zero is real but does not mean what it looks like

The tracked-journal sweep for 11–12 September returned zero hits. Following the rule in this project’s source notes — never conclude “quiet day” from an unverified zero — I ran controls:

• Same journal list, 10–11 September: 12 hits. The query works.
• All of Europe PMC, 11 September alone: 1,035 records. The index has that day.
• All of Europe PMC, 12 September alone: 0 records. Nothing at all has been indexed for today, in any journal, on any subject.
• Tracked journals, 11 September alone: 0 hits — a real negative for those journals on a Friday.

So today’s emptiness is an indexing lag, not a quiet day in the literature. Papers published today exist; Europe PMC has not caught up. Saturday-dated records typically appear within a day or two, and the next run should re-sweep 11–13 September rather than only the new date. That is now written into the source notes.

What follows is therefore drawn from a topic-wide sweep of 11 September and from the outstanding list.

1. Minimally invasive versus conventional extracorporeal circulation — 41 randomised trials

Journal · Published: Perfusion, 11 September 2026 — published yesterday Motawea, Ahmed, Khalil, Ibrahim, Abdelkader, Pelletier, El Diasty, Elgudin, Sabik, Abu-Omar

Meta-analysis of randomised trials comparing minimally invasive extracorporeal circulation (MiECC) with conventional extracorporeal circulation (CECC) in adult cardiac surgery. Four databases to July 2026.

41 RCTs, 5,215 patients — 2,652 MiECC, 2,563 CECC.

MiECC was associated with reductions in:

• Myocardial infarction — RR 0.50 (0.34–0.73), P = 0.0004
• Stroke — RR 0.50 (0.27–0.93), P = 0.03
• Delirium — RR 0.42 (0.24–0.73), P = 0.002
• Atrial fibrillation — RR 0.82 (0.72–0.93), P = 0.002
• Red cell units transfused — MD −0.67 (−0.84 to −0.49), P < 0.00001
• Chest tube drainage — −123 mL (−165 to −81), P < 0.00001
• Ventilation duration — −2.71 h (−3.72 to −1.69), ICU stay −0.99 days, hospital stay −0.61 days

No significant difference in: mortality, transient ischaemic attack, acute kidney injury, reoperation for bleeding, or platelet transfusion.

The authors’ conclusion: MiECC “may provide myocardial and neurological protection, better hemostasis, and faster early postoperative recovery… supporting its adoption as a preferred perfusion strategy.”

Interpretation. Take the effect sizes seriously and the conclusion cautiously.

A halving of myocardial infarction and stroke, and a 58% reduction in delirium, are very large effects for a perfusion circuit change. Large enough to be suspicious. MiECC trials are mostly small, single-centre, unblinded to the surgical team, and conducted by groups enthusiastic about the technique — exactly the conditions that inflate effect estimates. 41 trials averaging 127 patients each is a different object from one trial of 5,215, and the abstract reports no heterogeneity statistics, no funnel plot, and no risk-of-bias summary. Those are the three things I would want before believing RR 0.50 for stroke.

The outcomes I find most credible are the mechanical ones, because they are least susceptible to ascertainment bias and most plausibly mechanistic: less chest tube drainage, fewer red cells, shorter ventilation. MiECC means a smaller circuit, less haemodilution, reduced air–blood contact and closed-circuit suction — those directly explain less bleeding and less transfusion, and less transfusion plausibly explains part of the rest.

Mortality was null, which is the honest anchor. So was acute kidney injury — and that is worth noting given yesterday’s ADQI XXXVI / ELSO consensus, which named non-pulsatile flow and haemolysis among the mechanistic drivers of AKI during extracorporeal support. If circuit design mattered as much for the kidney as this analysis suggests it does for the brain and heart, you might expect an AKI signal. There isn’t one. That cuts slightly against the circuit-centred account of perfusion-related renal injury, or else says the relevant exposure is duration and flow character rather than circuit size.

It also sits beside the Annals of Thoracic Surgery critical perspective on perfusion modalities published 9 September — “Comparison of Two Forms of Pulsatility” (Ündar, DOI 10.1016/j.athoracsur.2026.08.020, PMID 42716270). That one has no abstract deposited and the full text was not reachable, so I cannot tell you what it argues — it stays on the outstanding list rather than being characterised from its title.

Practically: MiECC looks like a reasonable default on bleeding and recovery grounds, and the neurological and myocardial claims should be held loosely pending a look at the heterogeneity and bias assessments in the full paper.

Read the meta-analysis · PMID 42723534

2. DNR status and regional analgesia for fractures — 761,139 hospitalisations, and the least frail patients lost the most

Journal · Published: Regional Anesthesia and Pain Medicine, 4 September 2026 — eight days old, and the paper flagged yesterday as the missing half of a pair Siddiqui, Rodriguez, Hu, Choi, Jano, Manohara, Marshall, Douglas, Xu

Cross-sectional study using the US National Inpatient Sample, adult hospitalisations for rib and hip fractures, 2016–2022. Exposure: do-not-resuscitate status. Outcome: receipt of regional analgesia. The premise, in the authors’ words: “Do-not-resuscitate orders are intended to prioritize patient comfort, but do-not-resuscitate patients may receive fewer procedural interventions.”

298,343 rib fracture and 462,796 hip fracture hospitalisations.

• Regional analgesia was received by 8,568 (2.9%) of rib fracture and 13,730 (3.0%) of hip fracture admissions — low in absolute terms for everybody
• DNR status present in 31,882 (10.7%) of rib and 77,255 (16.7%) of hip fracture admissions
• Rib fractures: DNR associated with lower odds of regional analgesia — OR 0.76 (0.69–0.83)
• Hip fractures: no association — OR 1.00 (0.94–1.06)
• Frailty-stratified: the association was strongest among the LEAST frail rib fracture hospitalisations — OR 0.34 (0.25–0.48)

Interpretation. This is the paper that completes yesterday’s picture, and the frailty stratification is what makes it more than a disparity audit.

Read it against the Annals of Surgery code-status cohort sent yesterday. There, continuing a DNR order into theatre was associated with higher early mortality — and the authors showed it reflected dying trajectories, not withheld rescue, with 81% of deaths following a transition to comfort care. The reassuring conclusion was that code status was tracking prognosis rather than driving treatment.

This paper is the counterexample, and it is specific. If DNR patients received less regional analgesia because they were sicker, frailer, nearer death — plausible, and the kind of confounding that explains most such findings — then the effect should be largest in the frailest. It is the opposite. OR 0.34 in the least frail rib fracture patients means a robust patient with a DNR order had roughly a third the odds of getting a block. Frailty cannot explain that. What remains is that the DNR order itself changed the decision, in patients whose physiology gave no reason for it to.

And the intervention being withheld is analgesia — the one thing a DNR order is explicitly meant to prioritise. Hence the authors’ title. A rib fracture block is a comfort measure with no resuscitative content whatsoever; withholding it from a DNR patient is not conservative care, it is the precise inversion of what the order asks for.

Why rib and not hip? The abstract does not say, and I will not invent a mechanism. Two possibilities worth holding: hip fracture has well-embedded pathways and national audit pressure that may standardise analgesia irrespective of code status, whereas rib fracture blocks are more discretionary, more often requested ad hoc, and more often declined. Or the populations differ in ways the adjustment did not capture.

Limits, stated plainly: administrative claims data, so regional analgesia is captured by procedure coding and will be undercounted; DNR status is a coded flag with no information on when it was placed or what conversation produced it; cross-sectional, so no causal claim — which the authors concede, calling for prospective work. And 2.9–3.0% overall utilisation means the entire analysis concerns a rarely-used intervention.

Practically, this is the actionable pair of the week: a DNR order is not a reason to withhold a fascia iliaca or serratus plane block, and this data suggests it is functioning as one. Worth checking in your own institution, because it is checkable.

Read the paper · PMID 42697603

3. Dexmedetomidine versus propofol after CABG — a 300-patient null, in preprint

PREPRINT — not peer reviewed. Research Square, 11 September 2026 · DOI 10.21203/rs.3.rs-10756507/v1 · NCT05849597 (retrospectively registered, 9 May 2023) Preveden, Todić, Vučković, Pantić, Perić, Bakić, Marković, Jerković, Mihajlović, Crnomarković, Preveden, Maletin, Ubović, Andrić, Tatić, Redžek, Zdravković (Novi Sad)

Flagged as a preprint because it is one. It has not been peer reviewed, the numbers may change, and it should not be cited as established. It is included because the question is squarely primary scope and the result is a useful null.

Prospective, randomised, controlled, single-blind trial. 300 patients undergoing elective CABG with cardiopulmonary bypass, randomised 1:1 to postoperative sedation with dexmedetomidine 0.1–0.7 µg/kg/h or propofol 1.5–4 mg/kg/h. Delirium assessed by CAM-ICU on the day of surgery and the first postoperative day.

• Postoperative delirium in 53 patients (17.7%) overall — 23 (15.3%) dexmedetomidine vs 30 (20.0%) propofol, P = 0.364
• OR 0.724 (0.398–1.317), P = 0.290 — not significant
• Median ventilation 7.0 h vs 7.5 h, P = 0.159 on univariable comparison
• In multivariable analysis, dexmedetomidine was independently associated with shorter ventilation (B = −1.217, β = −0.147, P = 0.010)
• No difference in ICU or hospital stay, postoperative atrial fibrillation, or transfusion

Interpretation. A clean null on the primary outcome, and worth having for that reason.

The ASA’s 2025 practice advisory on older adults, covered here on 7 September, recommends considering dexmedetomidine for delirium prophylaxis while balancing cardiovascular risk — and noted that the cardiac surgical population is where both the benefit and the bradycardia/hypotension are largest. This trial, in exactly that population, did not find the benefit. 300 patients is reasonable for a single-centre trial and underpowered for a 5-percentage-point difference; the confidence interval on the odds ratio (0.40–1.32) comfortably includes both a meaningful benefit and a modest harm. So this does not refute the advisory — it declines to support it, which is a different and weaker thing.

Two methodological points worth flagging. Delirium was assessed only on the day of surgery and postoperative day 1 — a short window that will miss delirium emerging on day 2 or 3, when much of it appears after cardiac surgery. And the ventilation finding is significant only in multivariable analysis (P = 0.010) after a null univariable comparison (7.0 vs 7.5 hours, P = 0.159). A half-hour median difference that becomes significant on adjustment is the kind of result that deserves to survive peer review before being believed.

Read the preprint

---

Notes for the next run

• Re-sweep 11–13 September, not just the new date. Europe PMC had zero records indexed for 12 September at the time of this run, verified against a 1,035-record count for 11 September and a working control query. Today’s papers will appear retrospectively.
• Still blocked, third attempt: Annals of Thoracic Surgery, 9 September, “Comparison of Two Forms of Pulsatility: A Critical Perspective on Perfusion Modalities” (Ündar, DOI 10.1016/j.athoracsur.2026.08.020, PMID 42716270) — no abstract deposited. Relevant to both the MiECC analysis above and the ADQI non-pulsatile flow statement. Needs a PDF.
• Cleared from the outstanding list today: the RAPM do-not-resuscitate and regional analgesia paper, flagged yesterday and now sent.
• Also seen on 11 September and not pursued: a Perfusion-adjacent nationwide survey of methylene blue for refractory paediatric shock in Türkiye, a MIMIC-IV analysis of β-blockers in elderly sepsis with atrial fibrillation, two Chinese esketamine trials in abdominal and colorectal surgery, and a World Journal of Pediatric Congenital Heart Surgery argument for a regional cardiac surgery centre in the South Pacific.