Interobserver agreement in endoscopic evaluation of reflux esophagitis using a modified Los Angeles classification incorporating grades N and M: a validation study in a cohort of Japanese endoscopists

Dis Esophagus. 2008;21(4):355-63. doi: 10.1111/j.1442-2050.2007.00788.x.


The Los Angeles classification system is the most widely employed criteria associated with the greatest interobserver agreement among endoscopists. In Japan, the Los Angeles classification system has been modified (modified LA system) to include minimal changes as a distinct grade of reflux esophagitis, rather than as auxiliary findings. This adds a further grading M defined as minimal changes to the mucosa, such as erythema and/or whitish turbidity. The modified LA system has come to be used widely in Japan. However, there have been few reports to date that have evaluated the interobserver agreement in diagnosis when using the modified LA classification system incorporating these minimal changes as an additional grade. A total of 100 endoscopists from university hospitals and community hospitals, as well as private practices in the Osaka-Kobe area participated in the study. A total of 30 video clips of 30-40 seconds duration, mostly showing the esophagocardiac junction, were created and shown to 100 endoscopists using a video projector. The participating endoscopists completed a questionnaire regarding their clinical experience and rated the reflux esophagitis as shown in the video clips using the modified LA classification system. Agreement was assessed employing kappa (kappa) statistics for multiple raters. The kappa-value for all 91 endoscopists was 0.094, with a standard error of 0.002, indicating poor interobserver agreement. The endoscopists showed the best agreement on diagnosing grade A esophagitis (0.167), and the poorest agreement when diagnosing grade M esophagitis (0.033). The kappa-values for the diagnoses of grades N, M, and A esophagitis on identical video pairs were 0.275-0.315, with a standard error of 0.083-0.091, indicating fair intraobserver reproducibility among the endoscopists. The study results consistently indicate poor agreement regarding diagnoses as well as fair reproducibility of these diagnoses by endoscopists using the modified LA classification system, regardless of age, type of practice, past endoscopic experience, or current workload. However, grade M reflux esophagitis may not necessarily be irrelevant, as it may suggest an early form of reflux disease or an entirely new form of reflux esophagitis. Further research is required to elucidate the pathophysiological basis of minimal change esophagitis.

Publication types

  • Validation Study

MeSH terms

  • Adult
  • Aged
  • Esophagitis, Peptic / classification*
  • Esophagitis, Peptic / diagnosis*
  • Esophagitis, Peptic / pathology
  • Esophagoscopy*
  • Female
  • Humans
  • Japan
  • Male
  • Middle Aged
  • Observer Variation*