Amnon Sonnenberg1. 1. The Portland VA Medical Center, P3-GI, 3710 SW US Veterans Hospital Road, Portland, OR 97239, USA. sonnenbe@ohsu.edu
Abstract
BACKGROUND: In search for a source of gastrointestinal bleeding, endoscopy frequently reveals mucosal lesions of questionable dignity. AIM: To investigate the probability of ascertaining conclusive evidence for gastrointestinal bleeding from a suspicious mucosal lesion through a single or multiple consecutive endoscopies. METHODS: A mathematical model is developed to estimate the probability of successful diagnosis of a bleeding gastrointestinal lesion associated with single or multiple endoscopies. RESULTS: The probability of a successful confirmation through endoscopy depends on the length of time that signs of recent bleed persist at the site of the mucosal lesion and on the number of repeat endoscopies that one is willing to invest in confirmation. Assuming persistence of endoscopic evidence for 6-12 hours after the initial bleeding, a single endoscopy is associated with a 22 %-38 % chance of observing a suspicious site with clear evidence of bleeding. Using potentially up to 2 additional repeat endoscopies can raise such chances to 52 %-76 %. CONCLUSION: The rates of success may provide useful guidance in scheduling endoscopies for the work-up of gastrointestinal bleeding and decision making about the utility of repeat endoscopy in instances of suspicious but inconclusive mucosal lesions.
BACKGROUND: In search for a source of gastrointestinal bleeding, endoscopy frequently reveals mucosal lesions of questionable dignity. AIM: To investigate the probability of ascertaining conclusive evidence for gastrointestinal bleeding from a suspicious mucosal lesion through a single or multiple consecutive endoscopies. METHODS: A mathematical model is developed to estimate the probability of successful diagnosis of a bleeding gastrointestinal lesion associated with single or multiple endoscopies. RESULTS: The probability of a successful confirmation through endoscopy depends on the length of time that signs of recent bleed persist at the site of the mucosal lesion and on the number of repeat endoscopies that one is willing to invest in confirmation. Assuming persistence of endoscopic evidence for 6-12 hours after the initial bleeding, a single endoscopy is associated with a 22 %-38 % chance of observing a suspicious site with clear evidence of bleeding. Using potentially up to 2 additional repeat endoscopies can raise such chances to 52 %-76 %. CONCLUSION: The rates of success may provide useful guidance in scheduling endoscopies for the work-up of gastrointestinal bleeding and decision making about the utility of repeat endoscopy in instances of suspicious but inconclusive mucosal lesions.