Endoscopy 2008; 40(10): 799-805
DOI: 10.1055/s-2008-1077596
Original article

© Georg Thieme Verlag KG Stuttgart · New York

Mucosal morphology in Barrett’s esophagus: interobserver agreement and role of narrow band imaging

W.  L.  Curvers1 , C.  J.  Bohmer2 , R.  C.  Mallant-Hent3 , A.  H.  Naber4 , C.  I.  J.  Ponsioen1 , K.  Ragunath5 , R.  Singh5 , M.  B.  Wallace6 , H.  C.  Wolfsen6 , L.-M.  Wong Kee Song7 , R.  Lindeboom8 , P.  Fockens1 , J.  J.  Bergman1
  • 1Department of Gastroenterology and Hepatology, Academic Medical Center, Amsterdam, the Netherlands
  • 2Department of Gastroenterology and Hepatology, Spaarne Hospital, Hoofddorp, the Netherlands
  • 3Department of Internal Medicine, Flevohospital, Almere, the Netherlands
  • 4Department of Internal Medicine, Hospital Hilversum, Hilversum, the Netherlands
  • 5Wolfson Digestive Disease Centre, Queen’s Medical Centre, Nottingham, UK
  • 6Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, Florida, USA
  • 7Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota, USA
  • 8Department of Epidemiology and Biostatistics, Academic Medical Center, Amsterdam, the Netherlands
Further Information

Publication History

submitted 5 May 2008

accepted after revision 27 June 2008

Publication Date:
30 September 2008 (online)

Zoom Image

Background and study aims: We have recently proposed a classification of mucosal morphology in Barrett’s esophagus based on three criteria: regularity of mucosal pattern, regularity of vascular pattern, and presence of abnormal blood vessels. We aimed to evaluate the interobserver agreement with the proposed mucosal morphology classification and to assess the additional value of narrow band imaging (NBI) over high resolution white light endoscopy (HR-WLE).

Patients and methods: Five international experts in the field of Barrett’s imaging and seven community endoscopists with no expertise in this field independently evaluated magnified still images from 50 areas, obtained with HR-WLE and NBI, in Barrett’s esophagus patients. Visual analogue scales (VAS) were used for scoring imaging quality. Interobserver agreement for mucosal morphology and yield for identifying early neoplasia were assessed.

Results: Imaging qualities of NBI were rated more highly than HR-WLE, when evaluated separately as well as in a side-by-side comparison. The interobserver agreement ranged from 0.40 to 0.56 and did not significantly differ between expert and non-expert endoscopists. The overall yield for correctly identifying images of early neoplasia was 81 % for HR-WLE, 72 % for NBI and 83 % for HR-WLE + NBI, with no significant difference between experts and non-experts.

Conclusion: Interobserver agreement for the classification of mucosal morphology was moderate. Although NBI was rated more highly than HR-WLE for imaging quality, this did not result in improved interobserver agreement or increased yield for identifying early neoplasia in Barrett’s esophagus. This applied to non-expert as well as expert endoscopists.