Ali patients presenting with acute upper gastrointestinal bleeding between November 1986 and April 1988 were admitted to a centralised joint medical/surgical unit, with a policy of early clinical and endoscopic assessment and rapid surgical intervention in those at high risk. Of the 430 patients admitted 69*5% were over the age of60 and 30% had significant additional medical conditions. 50 4% were bleeding from peptic ulcers and one third had been taking non-steroidal anti-inflammatory agents. Fifty five patients underwent surgery, which in two thirds was carried out within 24 hours of admission, usually for continued bleeding. In patients with peptic ulcer the operation rate was 21*6%. Overall mortality was 3*7%, and in those with bleeding gastric or duodenal ulcers 5-5%; surgical mortality in the later group was 15-2%. All patients who died had serious concomitant pathology and 87% were over 70 years of age. Adoption of a centralised approach to management ofhaematemasis and melaena is feasible in a District General Hospital and associated with an improved survival.

INTRODUCTION: “Push” or “pull” techniques with the use of snares, forceps, baskets, and grasping devices are conventionally used to manage esophageal food bolus impaction (FBI). A novel cap-assisted technique has recently been advocated to reduce time taken for food bolus (FB) removal. This study aimed to compare the effectiveness of the cap-assisted technique against conventional methods of esophageal FB removal in a randomized controlled trial. METHODS: Consecutive patients with esophageal FBI requiring endoscopic removal, from 3 Australian tertiary hospitals between 2017 and 2019, were randomized to either the cap-assisted technique or the conventional technique. Primary outcomes were technical success and FB retrieval time. Secondary outcomes were technical success rate, en bloc removal rate, procedure-related complication, length of hospital stay, and cost of consumables. RESULTS: Over 24 months, 342 patients with esophageal FBI were randomized to a cap-assisted (n = 171) or conventional (n = 171) technique. Compared with the conventional approach, the cap-assisted technique was associated with (i) shorter FB retrieval time (4.5 ± 0.5 minutes vs 21.7 ± 0.9 minutes, P < 0.001), (ii) shorter total procedure time (23.0 ± 0.6 minutes vs 47.0 ± 1.3 minutes, P < 0.0001), (iii) higher technical success rate (170/171 vs 160/171, P < 0.001), (iv) higher rate of en bloc removal (159/171 vs 48/171, P < 0.001), and (v) lower rate of procedure-related mucosal tear and bleeding (0/171 vs 13/171, P < 0.001). There were no major adverse events or deaths within 30 days in either group. The total cost of consumables was higher in the conventional group (A$19,644.90 vs A$6,239.90). DISCUSSION: This multicenter randomized controlled trial confirmed that the cap-assisted technique is more effective and less costly than the conventional approach and should be first-line treatment for esophageal FBI.