Introduction Barrett’s Oesophagus (BO) is the only recognised precursor to oesophageal adenocarcinoma. The second generation EG Scan™ system is portable with a disposable, ultrathin, probe-based imaging capsule; therefore it may be used in the community setting closer to the patient’s home. We aimed to compare the performance of unsedated transnasal endoscopy (TNE) using the EG Scan™ with conventional oesophagogastroduodenoscopy (C-OGD) for the detection of BO.

Method This was a prospective, two-centre, diagnostic study with tandem design. Consecutive adult patients with histologically confirmed BO and those referred for assessment of reflux or dyspepsia were invited to participate. We excluded patients with recurrent epistaxis; nasal obstruction; and disease of the nasal cavity. All subjects underwent unsedated TNE (index test) followed by C-OGD (reference standard) on the same day, by two different operators blinded to the findings of each other.

The primary outcome measure was the diagnostic accuracy for BO. In addition, procedure preference; tolerability (10-point visual analogue scale (VAS), 0 = worst and 10 = best); and adverse events questionnaires were administered on day 0 and day 14 after the procedures.

Results 100 patients agreed to participate out of 439 subjects who were invited (22.8%). The mean age was 59.5 years (+/-13.7) and 63% were males. Prevalence of BO was 50%.

89 patients (89%) completed both procedures (11% failed TNE due to the inability to intubate the nasopharynx). Sensitivity, specificity and area under the receiver operating characteristic of TNE for the diagnosis of BO were 0.96 (95%confidence interval (CI) 0.85–0.99), 0.91 (95% CI 0.78–0.97), and 0.93 (95% CI 0.88–0.99), respectively.

Patients reported higher preference for TNE compared to C-OGD on both day 0 (73.0% vs. 18.0%%, p < 0.001) and day 14 (64.2% vs. 16.0%, p < 0.001). Preference for TNE remained significantly higher in patients who had C-OGD under sedation (57.1% vs. 31.4%, p < 0.001). Subjects also reported significantly better experience (mean VAS) with TNE compared to C-OGD on day 0 (7.5 vs. 5.6, p < 0.001) and day 14 (7.3 vs. 5.1, p < 0.001). Two patients (2.2%) experienced transient pre-syncopal episodes. No serious adverse events occurred.

Conclusion TNE using the EG Scan™ device was highly accurate for the detection of BO. More importantly, a significant majority of patients preferred TNE over C-OGD regardless of sedation use. The low test uptake could be due to the requirement for patients to undergo 2 procedures on the same visit. The EG Scan™ could potentially provide a true community-based screening tool for BO and oesophageal adenocarcinoma.

Disclosure of interest None Declared.