To evaluate the different endoscopic ways of diagnosing Barrett's esophagus (BE) and assess the quality of BE endoscopic reports using a rating grid.

Three-hundred and forty-six reports concerning BE (1997-2000) were divided into two groups, circular and non-circular BE, and rated on a scale (0 to 20 points) by two readers using a grid to evaluate, in particular, the description of BE, hiatal hernia and esophagitis and biopsy practice.

The score was > 10 points for 129 reports (37.3%). Points were deducted because of inadequate description of BE and hiatal hernia and particularly for a lack of biopsies (54% of reports). The indication of endoscopy (BE monitoring) and the form of medical arrangements (out- or in patients) had a significant effect on the mean score of the reports. Circular BE (46.5% of all reports) was described with respect to the height of the squamo-columnar junction (Z line) and the length of BE, and non-circular BE (40.4% of all reports) with respect to the number and height of the strips. Hiatal hernia, if detected, was described (67.2%) by its length, and esophagitis by standard classifications (70.1%). Finally, biopsy practice differed significantly as a function of the indication, the experience and status of the operator, or the medical management of the patient.

This study showed that the endoscopic reports of BE are not sufficiently rigorous and that the absence of biopsies is frequently not justified. In the future, standardized reports could be proposed by working groups of experts and then prospectively tested and approved.