Inter‐observer variation in coding osteoarthritis in human skeletal remains

Abstract

Abstract Thirty‐eight participants at the VIIIth European Meeting of the Paleopathology Association took part in a study of inter‐observer variation in scoring osteoarthritis in human skeletal remains. Ten specimens representing different joints were used and five criteria of osteoarthritis were scored. Eleven of the 38 participants ranked themselves as beginners, 13 as experienced and six as very experienced; the data were subsequently examined using the results from these 30, comparing beginners with experts. Agreement as to whether or not changes were present on the specimens and on the degree of change was seldom complete but was greater when scoring eburnation and the presence of new bone on the joint surface than for the three other criteria. There was little difference between beginners and experts. Although all the specimens were chosen to meet our published criteria for osteoarthritis, the experts were unanimous is agreeing the diagnosis in only three cases and the beginners in only one. These results suggest that more work needs to be done to develop operational definitions for the classification of disease in palaeopathology and that great care must be taken when comparing disease frequencies between studies.