Abstract

Observer variation in calculating various indices for estimating the severity and activity of Crohn's disease is reported. Seven prospective users of Crohn's disease activity indices were presented with 10 'cases histories' compiled from relevant patient data and asked to calculate (independently) various indices of severity and activity from this information. The results showed a disquieting degree of observer variation in all indices studied. Similar results were obtained when 15 members of the International Organisation for the Study of Inflammatory Bowel Disease (IOIBD) reviewed one case history and also when members independently reviewed a series of real life cases. It is suggested that each index so far proposed is open to considerable observer variation, which casts some doubt upon the validity of studies so far carried out. Clarification of indices and the use of ranking methods can, however, do much to overcome this discrepancy.