RESUMO
BACKGROUND/OBJECTIVE: Several studies have shown that NBI can predict the colorectal polyp histology with moderate to high accuracy. Most of them were conducted by highly experienced endoscopists at academic centers by using high magnification colonoscopes. We evaluated the accuracy of standard definition low magnification NBI in predicting the real time histology of colorectal polyps when used by inexperienced endoscopists and whether it can meet the thresholds recently defined for discard, resect and discard strategies of diminutive polyps by ASGE. METHODS: The study was a prospective observational trial and involved 430 patients who underwent colonoscopy. A total of 214 polyps were detected in 317 patients. Colonoscopies were performed by a gastroenterology attending physician and two fellows who had no experience in NBI. Standard definition low magnification colonoscopes were used. The endoscopists recorded the size, Paris classification, and location of polyps under white light and consecutively described the superficial mucosal architecture and vascular pattern based on NBI International Colorectal Endoscopic criteria. RESULTS: The overall accuracy of NBI in predicting adenomatous histology was 93.4%. The negative predictive value for diagnosing adenomatous histology in diminutive rectosigmoid polyps (93.3%) and the accuracy in predicting the colonoscopy surveillance interval (92.3%) were above the threshold levels defined by ASGE. CONCLUSIONS: Low magnification NBI by inexperienced endoscopists has a high accuracy in predicting in vivo colorectal polyp histology encouraging its use in routine clinical practice.
