Design: Open-label, stepped-wedge cluster randomised trial in 18 hospitals
Population: 2267 people under Barrett's oesophagus surveillance without known dysplasia
Primary outcome: Neoplasia detection rate in any biopsy
Effect: 7.4% vs 7.1%; OR 0.97 (95% CI 0.65 to 1.44)

The ACID trial, a stepped-wedge cluster randomised trial in 18 Dutch community hospitals published in the BMJ, enrolled 2267 people under Barrett's oesophagus surveillance without known dysplasia. Hospitals switched in random order to acetic acid chromoendoscopy after training; random four-quadrant biopsies continued throughout, and an experienced gastrointestinal pathologist confirmed all neoplasia.

Neoplasia was found in 7.4% with acetic acid and 7.1% without (OR 0.97, 95% CI 0.65 to 1.44). Acetic acid more than doubled the share of patients with a suspected lesion (14.0% vs 6.4%), but the yield of neoplasia from targeted biopsies did not rise (2.1% vs 1.8%).

For the pathologist, more suspicious lesions mean more targeted biopsy jars without more diagnoses. Random four-quadrant biopsies remain the backbone of detection.

Expect more targeted biopsy specimens, but not more dysplasia, where acetic acid chromoendoscopy is used.
Keep random four-quadrant biopsy protocols in place; they carried most of the diagnostic yield.
Seek expert gastrointestinal pathology review for every diagnosis of dysplasia in Barrett's oesophagus.
Record specimen sites separately so targeted and random yields can be audited.

Why it matters

A technique that finds more suspicious areas without finding more neoplasia adds workload, not diagnoses.

The statistics, in plain English

An odds ratio of 0.97 with an interval from 0.65 to 1.44 means no detectable difference. In a stepped-wedge design every hospital eventually uses the intervention, so the comparison is adjusted for calendar time.