ACCEPT TRIAL

Short summary Removal of adenomas during colonoscopy reduces colorectal cancer incidence and mortality. We and others have recently developed artificial intelligence (AI) systems which aim to optimise colonoscopy quality by increasing the adenoma detection rate (ADR). Meta-analyses of randomized trials at dedicated centres suggests that use of AI in colonoscopy increases ADR of from 20% to 30. These trials, however, enrolled mainly patients with clinical symptoms. Thus, the unsolved key question is whether the reported increase in adenoma detection is reproducible in population-based cancer screening programs. The ACCEPT project aims at addressing this important knowledge gap. ACCEPT is an international, population-based, randomised health service implementation project. The project will be embedded in the participating national cancer screening programs. A minimum of 9,252 participants who undergo colonoscopy in Norway, Poland, and Spain will be included. Participants will be assigned to undergo colonoscopy with or without the use of AI for polyp detection based on cluster randomisation. Clusters are defined by days of colonoscopy. We utilize the databases of the screening programmes and population registries to identify participants, obtain data on performance and endpoints during follow up. The primary endpoints are the detection rates of adenomas and advanced adenomas.