Design: Methodological survey of guideline-cited randomised trials
Population: 161 superiority trials cited in perioperative guidelines, 2012–2022
Primary outcome: Fragility index of the binary outcome supporting each recommendation
Effect: Median fragility index 4 (IQR 2 to 8); paediatric median 1

This methodological survey, published 25 September in Anesthesiology, sampled randomised trials cited in North American and European perioperative guidelines from 2012 to 2022, and calculated the fragility index — the number of patients whose outcome would need to change to lose statistical significance — for 161 superiority trials with binary outcomes.

The median trial had 120 participants and a median fragility index of 4 (interquartile range 2 to 8). Cardiovascular trials were the most robust (median 6); paediatric trials the least (median 1). Single-centre trials were more fragile than multicentre ones (incidence rate ratio 0.52).

Many recommendations anaesthetists follow rest on results that a handful of different outcomes would overturn. That is a reason to read the underlying trial, especially for paediatric recommendations, not to ignore guidance.

Check the size and number of events in the trial behind a guideline recommendation before treating it as settled.
Be most cautious with recommendations from small single-centre or paediatric trials.
A fragility index of 4 means four patients with a different outcome would erase significance.
Use fragility alongside, not instead of, GRADE and risk-of-bias assessment.

Why it matters

Challenges the assumption that a guideline-backed practice is always backed by robust evidence.

The statistics, in plain English

A median fragility index of 4 in trials of about 120 people means the statistical conclusion often depends on a few events — and loss to follow-up can exceed that.