Interobserver and intra-observer consistency in the endoscopic assessment of colonic pit patterns

Gastrointest Endosc. 2004 Oct;60(4):520-6. doi: 10.1016/s0016-5107(04)01880-2.

Abstract

Background: The colonic pit pattern is recognized as an aid to the differential diagnosis between hyperplastic lesions, adenoma, and carcinoma, and is a focus for observation by magnification chromoendoscopy, especially in Japan. This study evaluated intra- and interobserver agreement of experienced endoscopists in the assessment of colonic pit patterns when using the Kudo classification.

Methods: A total of 220 magnification chromoendoscopic pictures of colonic lesions were selected, of which 215 were collected from a consecutive series of patients. The pictures were randomly displayed twice to 6 experienced endoscopists at an interval of 1 week. Each picture was assessed for predominant pit pattern by using the classification of Kudo. Histopathologic diagnosis also was predicted based on the pit pattern diagnosis. Kappa statistics were used to estimate intra- and interobserver variation.

Results: The mean (standard deviation) inter- and intra-observer kappa values for experienced endoscopists were 0.716 (0.031) and 0.810 (0.084), respectively. For prediction of histopathology according to the pit pattern diagnosis, the mean (standard deviation) inter- and intra-observer kappa values were 0.776 (0.032) (p = 0.001) and 0.862 (0.069) (p = 0.028), respectively.

Conclusions: For experienced endoscopists, the inter- and intra-observer reproducibility of the classification of colonic pit pattern is good.

Publication types

  • Clinical Trial
  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Colon / pathology*
  • Colonic Neoplasms / pathology*
  • Colonic Polyps / pathology
  • Colonoscopy*
  • Color
  • Female
  • Humans
  • Hyperplasia
  • Intestinal Mucosa / pathology
  • Male
  • Middle Aged
  • Observer Variation
  • Reproducibility of Results