The American Journal of Gastroenterology has published findings from a study conducted at University of Maryland, Baltimore, which indicate that use of capsule endoscopy may yield disappointing rates of lesion detection, regardless of experience.
The article abstract can be found here. It details results from 17 endoscopists, with experience ranging from 23 to over a thousand capsule endoscopies, who were faced with a sample of 24 clips, eighteen of which contained abnormalities. Different reading modes were adopted and Single View (as opposed to quad view) 25 frames per second was in all cases a significantly less successful modality, identifying only 26% of all lesions and only 17% of angioectasias, ulcers/erosions, masses/polyps, and blood(This compared with, for example SingleView 15fps rate of 69%).
The item concludes that “Quality control measures to compare and improve lesion detection rates need further study.”
Tough to argue on these figures.
Source: Nature.com, American Journal of Gastroenterology