Using a mobile unit to deliver capsule sponge testing for Barrett's oesophagus in the community in England through GP invitation and self-referral: an implementation study.

Abstract

Objective: Offering individuals with gastro-oesophageal reflux a non-endoscopic capsule sponge test increased detection of Barrett's oesophagus (BO), a precursor to oesophageal adenocarcinoma. However, only 24% of individuals identified through general practitioner (GP) records participated. We therefore performed an implementation study to assess the feasibility of screening on a mobile unit and screening uptake among GP-invited and self-referred individuals.

Methods: In this implementation study, 3293 individuals aged ≥50 with a repeat proton pump inhibitor prescription were invited to participate by their GPs across three GP practices in the East of England between June 2021 and January 2023. Additionally, we advertised the study and allowed self-referrals. Eligibility was assessed based on age and reflux history. Eligible individuals were invited to an appointment. Acceptability was assessed after the appointment. Individuals with a positive capsule sponge test were referred for endoscopy.

Results: Half of the GP-invited individuals responded, of whom 59.6% (n=952/1598) were interested. Among interested individuals who could be contacted, 66.3% (n=574/866) were eligible, and 532 participated. Of these, 502 (94%) successfully swallowed a capsule sponge, representing 15% of those invited and 87.5% of eligible individuals. Among 468 individuals who self-referred and could be contacted, 48.5% (n=227) were eligible, 217 participated, and 208 (96%) successfully swallowed a capsule sponge (39.6% of self-referred, 91.6% of eligible individuals). 9.4% (n=47/502) of GP-invited and 12.0% (n=25/208) of self-referred individuals were referred for endoscopy, with 4.2% (n=21/502) and 3.4% (n=7/208), respectively, diagnosed with BO (positive predictive value of endoscopy referral=42.4%), including three with dysplasia. Over 99% of participants said the capsule sponge test was acceptable.

Conclusion: Screening on a mobile unit is feasible. Identifying individuals at increased risk of BO from GP records is feasible, and encouraging self-referrals increases uptake. BO prevalence was similar among eligible participants who self-referred and those identified through GP records.

Trial registration number: ISRCTN91655550.