Article Text

Download PDFPDF

PTU-173 Gord Symptoms and Demographic Factors as a Pre-Screening tool for Barrett’S Oesophagus
  1. X Liu1,2,
  2. A Wong2,
  3. S R Kadri2,
  4. M O’Donovan3,
  5. P Lao-Sirieix2,
  6. R Burnham4,
  7. R Fitzgerald2
  1. 1Oncology, University of Cambridge
  2. 2MRC Cancer Cell Unit
  3. 3Department of Histopathology, University of Cambridge, Cambridge
  4. 4Queen’s Hospital, Essex, UK


Introduction Barrett’s oesophagus (BO) occurs as consequence of reflux and is a risk factor for oesophageal adenocarcinoma (OAC). The current “gold-standard” for diagnosing BO is endoscopy which remains prohibitively expensive and impractical as a population screening tool. Therefore, we aimed to investigate the epidemiological factors and symptoms associated with BO in order to develop a pre-screening tool to aid decision making for diagnostic referrals.

Methods A prospective (training) cohort of 1603 patients attending for gastroscopy was used for identification of risk factors to develop a risk prediction model. Factors significantly associated with BO in the univariate analysis were selected to develop a prediction model that was validated in an independent, external cohort of 504 patients in primary care. We used two definitions of BO in the current study: 1) columnar lined epithelium of oesophagus (CLE) of any length reported in the endoscopy report with columnar epithelium on biopsy 2) an intestinal metaplasia confirmed on histopathological assessment with endoscopic length of BO ≥2 cm (IM ≥ 2 cm).

Results An eight-factor panel, including age, sex, smoking status, heartburn, acid reflux, chest pain, abdominal pain, anti-reflux medication, was identified from the training cohort with an area under the ROC curve (AUC) of 0.74 (95%CI: 0.70–0.77) for CLE, and 0.80 (95% CI: 0.75–0.84) for IM ≥ 2cm. This panel was significantly associated with BO in the external cohort, and the odds ratios for each factor increase were 1.43 (95%CI: 1.02–2.01) and 1.30 (95%CI: 1.04–1.62) for CLE and IM ≥ 2cm, respectively. The AUCs of the 8-factor panel for CLE and IM ≥ 2cm in the external cohorts were 0.85 (95%CI: 0.77–0.92) and 0.78 (95%CI: 0.71–0.84), respectively. After controlling for over-fitting, the AUCs dropped to 0.65 (95% CI: 0.54–0.77) and 0.67 (95%CI: 0.60–0.74), respectively.

Conclusion An eight-factor panel can help predict the presence of BO and has the potential to be applied in the general population with GORD symptoms as a pre-screening tool to help select patients for further investigation.

Disclosure of Interest None Declared.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.