P172 Systematic review of potential quality indicators for dysplasia and cancer detection at upper gastrointestinal endoscopy

Introduction There are no established quality indicators for upper gastrointestinal (UGI) endoscopy. We conducted a systematic review of potential quality indicators for UGI endoscopy that are associated with the diagnosis of pre-neoplastic or neoplastic UGI conditions. Methods Bibliographic databases were searched until June 2021. Studies reporting diagnostic accuracy, detection rates or the association of endoscopy or endoscopist related factors and a diagnosis of UGI pre-neoplastic or neoplastic conditions were included. Modified CASP and JBI checklists were used to assess the quality of systematic reviews and primary studies respectively. Selected results were presented in forest plot, but pooling was not feasible due to substantial heterogeneity. A protocol was prospectively registered with PROSPERO (#225339) Results Seven systematic reviews and 9 primary studies were included. Systematic reviews and primary studies were mostly of good methodological quality but included study populations were not always representative of the global population in terms of underlying risk (of cancer/dysplasia). Use of advanced imaging techniques, especially narrow band imaging (NBI), had high diagnostic accuracy to detect dysplasia and early gastric cancers (pooled sensitivity-0.87 (95% CI 0.84-0.89) and specificity-0.97 (0.97-0.98)). Higher detection rates with longer observation times during endoscopy were reported by two studies, but no difference was seen in two other studies (Figure 1). Endoscopist biopsy rate was strongly associated with the detection of gastric cancer and premalignant lesions (odds ratio for higher biopsy rate 2.5 (95% CI 2.1-2.9)). Detection rates for both early oesophageal cancers (2.62 vs 2.17 per 10,000 person years) and early gastric cancers (2.39 vs 1.91 per 10,000 person years) were higher in endoscopies performed under sedation. For Barrett’s oesophagus, NBI targeted biopsies and use of acetic acid offered high diagnostic accuracy to detect high grade dysplasia and early oesophageal cancer. Detection rates were higher if Barrett’s examinations were performed by a trained endoscopist on a dedicated list (8% vs 3%, p<0.001) and if inspection time of Barrett’s mucosa was more than 1minute/cm (54.3% vs 13.3%, p 0.04). Neoplasia detection rate on index endoscopy for patients undergoing screening for Barrett’s oesophagus was 7 (4-10)%. Conclusion Use of advanced imaging techniques, longer inspection times, endoscopist biopsy rate and use of sedation are potential quality indicators for UGI endoscopy. Neoplasia detection rate at index endoscopy is an additional potential quality indicator for Barrett’s oesophagus examination.