PHYSICIANS ABDOMINAL AUSCULTATION - A MULTI-RATER AGREEMENT STUDY

Background: Abdominal auscultation has an important position in the ph ysical examination of the abdomen. Little is known about rater agreeme nt. The aim of this study was to describe rater agreement and thus, in directly, the value of the examination. Methods: In a semi-virtual set up 12 recordings of the intestinal sounds from 8 patients with acute a bdominal pain and 4 healthy volunteers were presented to 100 physician s. The physicians were asked to characterize the intestinal sounds as normal or pathologic. Fisher's exact test was used for comparison betw een groups of physicians. Results: Overall, 72% of the answers with re gard to healthy volunteers concluded that the sounds were normal (equa lling agreement), whereas 64% of answers with regard to intestinal obs truction concluded that the sounds were pathologic (but agreement was higher due to agreement on wrong diagnosis in one case). Bowel sounds from colonic obstruction were diagnosed as pathologic in 94 of 100 ans wers. In peritonitis disagreement dominated. Specialists in gastrointe stinal medicine diagnosed bowel obstruction significantly more poorly than nonspecialists in gastrointestinal surgery (P < 0.05). Conclusion : Rater agreement in normal subjects and in patients with intestinal o bstruction was acceptable for a clinical examination. Abdominal auscul tation is a helpful clinical examination in patients with acute abdomi nal pain.