Adding regional to general anaesthesia tracked with slightly better function and lower mortality in 680,000 orthopaedic patients

Design

target trial emulation with stabilised inverse probability of treatment weighting, national claims database

Population

244,212 patients having general plus regional anaesthesia and 435,462 having general alone for major orthopaedic surgery, Japan 2016-2023

Primary outcome

change in Barthel Index between admission and discharge

Effect

mean difference 0.73 (95% CI 0.57-0.88); length of stay -1.40 days (-1.45 to -1.35); in-hospital mortality HR 0.85 (0.76-0.96)

A target trial emulation used the Japanese Diagnosis Procedure Combination database from 2016 to 2023 to compare major orthopaedic surgery under general anaesthesia combined with regional anaesthesia - 244,212 patients, of whom 183,564 had a peripheral nerve block and 60,648 an epidural - against general anaesthesia alone in 435,462. The primary outcome was the change in Barthel Index between admission and discharge, with stabilised inverse probability of treatment weighting on preoperative covariates.

The combined group did better on all three outcomes, by small margins. Barthel Index change was higher by a mean 0.73 points (95% CI 0.57-0.88), postoperative stay was 1.40 days shorter (-1.45 to -1.35) and in-hospital mortality was lower (hazard ratio 0.85, 95% CI 0.76-0.96). Sensitivity analyses ran in the same direction.

The honest reading turns on what 0.73 Barthel points means, which is essentially nothing for an individual patient - the scale runs to 100 and moves in increments of 5. A 1.4-day shorter stay is worth having and a 15% relative reduction in in-hospital mortality would be worth a great deal if it were causal. It may not be: the authors themselves name residual confounding, and who receives a block is decided by the same clinicians assessing how well the patient is likely to do. Read it as consistent with regional anaesthesia being beneficial, and as nowhere near sufficient to establish it.

• Continue offering regional anaesthesia alongside general for major orthopaedic surgery where it is appropriate and feasible
• Do not quote a mortality benefit to a patient on the strength of this - it is observational
• Note that peripheral nerve blocks made up three-quarters of the combined group; the results do not separate block types
• Where blocks are not being done for capacity reasons, the length-of-stay finding is the more usable argument
• Watch for the prospective confirmation the authors call for before changing a departmental standard

Why it matters

It is the largest comparison available and it still cannot answer whether regional anaesthesia changes survival.

Don't overread it

An administrative-database emulation - the mortality and functional differences are associations, and confounding by indication is the obvious alternative explanation.

The statistics, in plain English

With 680,000 patients, almost any difference reaches statistical significance - which is why the Barthel Index result (mean difference 0.73, interval 0.57-0.88) is a textbook case of a finding that is statistically certain and clinically negligible. The mortality hazard ratio of 0.85 has an interval (0.76-0.96) that stays below 1.0, but propensity weighting can only balance the variables that were recorded in a claims database, and frailty, cognition and the reason a block was or was not attempted are not among them.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.