Search

Although upper GI bleeding (UGIB) is a significant cause of inpatient admissions, no scoring method has proven to be accurate and simple as a standard for triage purposes. Therefore, we compared a previously described 3-variable score (1 point each for absence of daily proton pump inhibitor use in the week before the index presentation, shock index [heart rate/systolic blood pressure] ≥1, and blood urea nitrogen/creatinine ≥30 [urea/creatinine≥140]), the Horibe gAstRointestinal BleedING scoRe (HARBINGER), with the 8-variable Glasgow-Blatchford Score (GBS) and 5-variable AIMS65 to evaluate and validate the accuracy in predicting high-risk features that warrant admission and urgent endoscopy.Consecutive patients presenting with suspected UGIB between 2012 and 2015 were prospectively enrolled in 3 acute care Japanese hospitals. On presentation to the emergency setting, an endoscopy was performed in a timely fashion. The primary outcome was the prediction of high-risk endoscopic stigmata.Of 1486 enrolled patients, 637 (43%) harbored high-risk endoscopic stigmata according to international consensus statements. The area under the receiver operating characteristic curve (AUC) for the HARBINGER was .76 (95% confidence interval [CI], .72-.79), which was significantly superior to both the GBS (AUC, .68; 95% CI, .64-.71; P .001) and the AIMS65 (AUC, .54; 95% CI, .50-.58; P .001). When the HARBINGER cutoff value was set at 1 to rule out patients who needed admission and urgent endoscopy, its sensitivity and specificity was 98.8% (95% CI, 97.9-99.6) and 15.5% (95% CI, 13.1-18.0), respectively.The HARBINGER, a simple 3-variable score, provides a more accurate method for triage of patients with suspected UGIB than both the GBS and AIMS65.

Procedure of Endoscopic Hemostasis Using a Distal Attachment and Short Clips Against Upper Gastrointestinal Bleeding Naoki Hosoe, Hiroyuki Imaeda, Haruhiko Ogata, Hidekazu Suzuki, Tatsuhiro Masaoka, Manabu Nakashita, Tomoko Tsuda, Hiroyuki Kimura, Koichi Aiura, Yasushi Iwao, Hiroshi Nagata, Koichiro Kumai, Toshifumi Hibi, Hiromasa Ishii BACKGROUND and AIM: Endoscopic hemostasis using hemoclips is useful against upper gastrointestinal bleeding, however it is difficult to perform hemoclipping against tangent view-lesions. A hood which is attached to a distal tip of an endoscope makes hemoclipping more easier, however it is sometimes difficult to observe lesions because clot is put into the hood. The aim of this study is to assess endoscopic hemostasis using a distal attachment and short clips against upper gastrointestinal bleeding. SUBJECTS and METHODS: Subjects were 103 patients with 117 lesions of upper gastrointestinal bleeding. A distal attachment, which was a short hood (4 mm in length) was attached to the distal tip of an endoscope from start of endoscopic examination. If lesions of upper gastrointesitinal bleeding were detected, hemoclipping using short clips was performed against them. This procedure had been performed against 50 lesions, including 20 of gastric ulcer, 14 of duodenal ulcer, 12 of lesion resected endoscopically against early gastric cancer and gastric adenoma, 3 of Mallory-Weiss syndrome and 1 of stomal ulcer. Endoscopic hemostasis by this procedure was analyzed. RESULTS: It was easy to insert the endoscope with the distal attachment into the upper GI in all patients.Appropriate endoscopic view using the distal attachmentwas achieved without disturbance of clot in all lesions. For tangent view-lesions, this attachment made endoscopic observation and treatment more easier. In addition, the entire short clip was able to be stored into this attachment, and certainly clamped a protruded visible vessel even if the bottom of the lesion was hard. Forty-seven of 50 lesions (94.0%) had temporal hemostasis and forty-six lesions (90.0%) had permanent hemostasis. One patient with multiple gastric ulcers, advanced hepatocellular carcinoma and bleeding tendency associated with DIC had temporal hemostasis, however he died of rebleeding from the ulcers. Two patients with duodenal ulcer and one patient with gastric ulcer underwent interventional radiology because visible vessels were more than 3mm. One patient with duodenal Dieulafoy’’s lesion close to papilla of Vater underwent endoscopic hemoclipping using a lateral-viewing endoscope. CONCLUSION: Endoscopic hemostasis with a distal attachment and short clips is useful against upper gastrointestinal bleeding. *M1777 Sensitivity and Inter-Observer Variability for Capsule Endoscopy Image Analysis in a Cohort of Novice Readers Gary Chen, Simon Chan, Clifford Quan, Mary Kalpakian, Tam Tran, Pedram Enayati, Mary Lee Henderson, Gareth Dulai, Ian Arnott, Fumiaki Kitahara, Jeff Gornbein, Rome Jutabha Capsule endoscopy image analysis is a time consuming and tedious process. The process for selecting, training, and validating an individual’s (reader) ability to accurately detect abnormalities has not been well studied. Aim: To determine the sensitivity, specificity, and concordance rate of novice readers (4th year medical students withminimal endoscopic background) for detecting pre-specified capsule endoscopic findings. Hypotheses: 1) Medical students can reliably detect small bowel lesions deemed as definitive lesions for obscure bleeding with a high degree of sensitivity but low specificity, and 2) the concordance rates between the readers and gold standard will be high. Methods: Ten encoded test cases with definitive sites for bleeding were administered to the readers (n=9) in a pre-set order (bleeding: AVM-3, Meckel’s-1, ulcers/aphthous lesions-3, foreign body with ulceration-2). The gold standard findings were pre-defined by 2 expert capsule endoscopists (over 75 capsule cases each). Twelve gold standard targets were identified among the 10 test cases. The readers were given a 30 minute overview of the Rapid Reader (GIVEN Imaging) software and instructions for thumbnailing anatomic landmarks and any potential abnormality. A proprietary software program was developed to compile and analyze thumbnailed finding with respect to sensitivity and specificity as compared to a gold standard and overall concordance rates. Results: The average sensitivity among the 9 medical students was 67% with gradual improvement in sensitivity and specificity over time. All findings were detected by at least 2 to 9 of the students. An average of 19 false positive incidental finding’s per case were made by each student. The concordance (Kappa statistic) between readers was 22% for these 12 gold standard findings (intra-class correlation=0.22). Conclusions: A panel of medical students with minimal endoscopic experience can reliably detect high risk bleeding lesions on capsule endoscopy with moderate sensitivity, and can pre-screen recordings to thumbnail potential abnormalities. Due to low specificity, expert capsule endoscopists must review these thumbnails to determine the clinical relevance of each finding. Additional studies are ongoing to assess other reader cohorts (nurses, GI fellows, non-medical) abilities to detect various bleeding as well as nonbleeding abnormalities on capsule endoscopy.

Novel Procedure of Grasper-associated Endoscopic Mucosal Resection Against Early Gastric Cancer Hiroyuki Imaeda, Yasushi Iwao, Haruhiko Ogata, Hidekazu Suzuki, Naoki Hosoe, Tatsuhiro Masaoka, Manabu Nakashita, Tomoko Tsuda, Koichi Aiura, Hiroshi Nagata, Koichiro Kumai, Toshifumi Hibi, Hiromasa Ishii BACKGROUND and AIM: Endoscopic mucosal resection (EMR) by submucosal dissection against early gastric cancer has improved the rate of en blocresection, however more skillful technique should be necessary in comparisonwith strip biopsy technique, and it has been claimed that there are more complications. Recently many kinds of devices of instruments in EMR have been reported, however it is still difficult to dissect the submucosa of lesions especially located in gastric body. The aim of this study is to assess the usefulness of grasper-associated endoscopicmucosal resection which we have designed against early gastric cancer. SUBJECTS and METHODS: Subjects were 11 lesions of early gastric cancer which has been thought to be remained within the mucosa of gastric body. The average size of those lesions was 16.5mm (range 10-25mm). A short hood (4-mm in length) was attached to the distal tip of an endoscope, and a flexible overtube was inserted with the endoscope. After marking around lesions, 10% glycerin with indigocarmine and epinephrine was injected into the submucosa. After the circumferential mucosa around the lesion was cut by a needle knife, then the endoscope was pulled out once. Next, a grasper (inside grasper) was inserted through a working channel, and a distal portion of another grasper outside the endoscope (outside grasper) was grasped by inside grasper, thereafter the endoscope was inserted again with the outside grasper. After the lesion was grasped by the outside grasper under control of the inside grasper and the endoscope, the inside grasper was pulled out. Finally, while pulling up the lesion from anal towards oral side by the outside grasper, the submucosal dissection was performed from the anal side lesion. RESULTS: (1) Traction of those lesions towards the oral side by the outside grasper achieved more open and wide range of the vision of the submucosa. (2) All 11 lesions were able to be resected en block by this procedure. (3) Operating time was approximately 60minutes and shorter than conventional procedure. (4) During procedure, small amount of hemorrhage was the only complication, and there happened no perforation. Furthermore, combined this procedure with squirting water, detection of bleeding spot and instant hemostasis was able to be performed. CONCLUSION: This novel procedure of grasper-associated EMR could make submucosal dissection certainly, more easier and safer against early gastric cancer located in gastric body. **186 Testing of a New Device for Endoscopic Mucosal Resection for the Esophagus and Colon Paul Swain, Per-Ola Park, Maria Bergson, Annette Fritscher-Ravens, Keiichi Ikeda Background: Endoscopic mucosal resection of Barrett segments or other dysplastic abnormalites in the esophagus or colon is difficult to control precisely with injection and snare closure. A cylindrical end-cap on the tip of an endoscope limits the area of tissue that can be resected and takes circular disks of tissue from tubular structures. Large volume resections in esophagus and colon risk perforation. Aim: To test a device designed to resect large volumes of esophageal or colonic tissue in strips controlling the depth of resected tissue. Materials and methods: The resection device was formed using a flexible over-tube with an operating tip. The proximal end of the over-tube featured an airtight seal allowing passage of conventional endoscopes of 6-11mm, thumb control allowed the linear resection wire to be moved to cut a rectangular section of tissue at the tip, diathermy connector, adjustable needle injector and port for saline injection. At the distal end tissue could be sucked into a rectangular cavity of 13 3 cm.Depth of tissue cutwas controlled by a soft plastic sling under the cavity. The needle could be advanced into tissue in the centre of the proximal edge of the cavity. Saline could be injected with precision at any predetermined point in the proximal margin of the area to be resected. The overtube had a soft open conical endpiece, which could be occluded by pushing it against tissue and applying suction. The endoscope could be pushed through the cone to examine the stomach or colon. Results: The device was used in postmortem esophagus and colon, and the esophagus of the anesthetized pig. Intubation over an endoscope or catheter was atraumatic. The device was used with 11 mm (2.8 mm channel) and 5.9 mm (2.2 mm channel) endoscopes. Suction was effective in filling the rectangular resection cavity. The raised semitransparent bleb after injection filled half the cavity. The separation of the mucosa from the muscle could be clearly viewed as the linear diathermy wire cut through the tissue in the rectangular cavity. The resected specimens were a neat rectangle of mucosa and submucosal exposing the deep muscle. The tissue was retained in the tube for removal. Optimal diathermy settings were 40-60 W pure cut. No bleeding was observed. Perforation did not occur. Conclusion: A new device for resection of Barrett segments or other mucosal or submucosal abnormalities was described. It was easy to use, removedmucosal and submucosal strips of predetermined depth and appeared safe in bench and live animal testing.