Incomplete and Low Quality Inspection of the Stomach During Upper Endoscopy

Abstract

Purpose: Recent reports have shown a surprisingly low protective effect of colonoscopy against colorectal cancer. Upper endoscopy (UE) is the preferred technique to diagnose gastric cancers, but as for colonoscopy, UE is associated with a substantial miss rate for gastric cancers. Aim: To determine completeness and quality of inspection of the stomach during routine UE by measurement of the time of endoscope retroflexion in the stomach and quality of inspection of the peri-esophageal fundus. Methods: All endoscopies performed in 8 endoscopy rooms at Mayo Clinic Rochester were automatically captured as part of a prospective, periodic assessment of quality and stored as de-identified, digital video files during a random week in the fall of 2009. Endoscopists were aware of the existence of the capturing system. UE video files were selected and analyzed for (1) time spent on retroflexion of the endoscope in the stomach and (2) the quality of inspection of the peri-esophageal fundus. Retroflexion time was measured in seconds by recording the first and last video frames showing retroflexion and dividing the difference by the frame rate. Absence/presence of retroflexion was scored as 0 or 1. Retroflexed fundus and esophago-gastric junction inspection was scored for absence/presence of (1) clearness of image, (2) 360 degree view around endoscope shaft and (3) adequate close-up views; each was scored as 0 or 1. For overall quality all scores were combined and categorized as follows: 0=Not Done; 1=Very Poor, 2=Poor, 3=Fair, 4=Good. Results: 96 video files represented UEs; all were of excellent video quality. All files were completely de-identified; therefore the indication for UE was unknown, a potential weakness of the study. Retroflexion was not performed in 4 UEs (4.2%). When performed, retroflexion time ranged from 1 to 127 seconds; excluding 4 outliers with retroflexion times greater than 30 seconds, the mean retroflexion time was 11.7+6.9 (M+SD) seconds. Overall retroflexion quality was Very Poor for 28 (29.2%), Poor for 1 (1%), Fair for 18 (18.8%) and Good for 45 (46.9%). Linear regression of overall retroflexion quality and retroflexion time showed an association that was highly significant (p<0.0001), even after removal of the 4 outlier UEs with retroflexion times greater than 30 seconds. Conclusion: 1. There is a large variation in time and effort spent on retroflexed inspection of the peri-esophageal fundus of the stomach during UE. 2. Incomplete or low quality inspection of the stomach occurs in about a third of UEs. 3. Longer inspection times are associated with higher quality UEs. 4. New methods need to be implemented that guarantee a high quality UE at all times for all patients. Disclosure: Piet C. de Groen, Wallapak Tavanapong, JungHwan Oh, Johnny Wong - Stockholder/Ownership interest: EndoMetric.