ACCURACY IN DRY BONE DIAGNOSIS - A COMMENT ON PALEOPATHOLOGICAL METHODS

The objectives of palaeopathological research include: (i) diagnosis o f specific diseases in archaeological human remains; (ii) analysis of the impact of various diseases in human populations through time and s pace; and (iii) clarification of evolutionary interactions between hum ans and disease. Basic to all these objectives is the need for accurac y in diagnosing diseases in archaeological human remains. Tests of dia gnostic accuracy made during workshops in dry bone diagnosis held duri ng the annual meetings of the Paleopathology Association (PPA) in 1989 , 1990, 1991 and 1992 suggest important limitations that need to be co nsidered in the development of methods in palaeopathology. Although th e conditions do not approximate those found in most field or laborator y situations, these tests indicate an overall accuracy of 28.6 per cen t for specific disease recognition and 42.9 per cent accuracy for corr ect recognition of the more general categories of disease. The factors in diagnostic accuracy are complex and include the knowledge and expe rience of the observer. Another factor is the often substantial morpho logical overlap between different diseases. In the classificatory syst em used for the workshops there are seven general categories of diseas e, each of which has recognizable, although often non-specific, hallma rks in dry bone specimens. A differential diagnosis with these seven g eneral possibilities in mind rather than hundreds of specific diseases can be a powerful tool for the palaeopathologist in providing data th at are more comparable between observers. This improvement in classifi catory agreement has important methodological implications in the deve lopment of a data protocol in palaeopathological research. The identif ication of general disease category is recommended as part of descript ive and diagnostic reports on palaeopathological specimens.