Introduction Barrett's columnar lined oesophagus (CLO) is the pre-malignant lesion to oesophageal adenocarcinoma (OAC). The presence of dysplasia, when diagnosed in surveillance programmes, is an important marker of risk of progression and an indication for endoscopic therapy. Barrett's surveillance should be undertaken with a minimum of four quadrant biopsies every 2 cm, with documentation of length of BE segment by Prague C/M criteria, yet adherence to this is variable. We aimed to assess the quality of Barrett's surveillance and effect on dysplasia detection.

Methods Prospective database of patients undergoing Barrett's surveillance over 3 year period at tertiary referral upper GI centre. Patients with a previous diagnosis of dysplasia/OAC were excluded. Endoscopists were separated into two groups; Group A—endoscopist has newly diagnosed patient with dysplasia over study period; Group B—endoscopist has not diagnosed patient with dysplasia at any time point. Analysis was by independent t-tests for continuous variables and chi-squared tests for categorical variables.

Results 395 patients with Barrett's CLO underwent endoscopy between 2007 and 2010. Of these 23/395 were diagnosed with dysplasia/OAC by group A endoscopists (n=14) vs none in Group B (n=32). Results are shown in Abstract PTU-207 table 1. There was no significant difference in patient's age, sex or length of Barrett's mucosa between the two groups. There was a significant difference in adherence to Seattle protocol four quadrant biopsies every 2 cm between the two groups. The detection rate of Specialised Intestinal Metaplasia (SIM) and documentation of length of Barrett's mucosa by Prague criteria were significantly higher in group A than group B. The use of High Resolution Endoscopy was similar between both groups.

Conclusion This study demonstrates that endoscopists who detect dysplasia arising in Barrett's CLO are more likely to undertake high quality Barrett's surveillance as evidenced by higher rates of SIM, adherence to Seattle protocol biopsy and documentation of length of Barrett's mucosa by Prague classification. This lends support to the argument that Barrett's surveillance should be centralised and undertaken on dedicated lists by trained endoscopists with a specialist interest, in order to maximise dysplasia detection rates. A prospective study is warranted.

Competing interests None declared.