Interobserver variation in diagnosing coeliac disease. A joint study by Danish and Swedish pathologists.

There is an almost 40-fold difference in incidence rates of symptomatic coeliac disease between Denmark and Sweden. In an attempt to explain this difference, the present study focused on the interobserver agreement when pathologists were assessing small intestinal biopsy specimens from children suspected of suffering from coeliac disease. The study was performed on 90 biopsy specimens from 73 children. Most of the biopsies came from children who turned out not to suffer from coeliac disease after a clinical evaluation including small intestinal biopsy. Using the kappa methodology, the interobserver agreement between two Danish pathologists and one Swedish pathologist, all of whom were experienced, was "moderate" to "substantial" or 0.57-0.75. Kappa indices when the pathologists evaluated selected histological elements were in the interval from 0.24 to 0.67. A comparison of a previous routine diagnostic assessment of the 90 biopsies (14 pathologists) with the results of the experienced pathologists in the present study gave kappa indices of from 0.53 to 0.57. The study could prove no major differences in the histopathological assessment of small intestinal biopsy specimens made by Danish and Swedish pathologists. The difference in clinical presentation of coeliac disease in Denmark and Sweden does not relate to differences in the histopathological assessment of small intestinal biopsies.