Observer variation in the radiographic assessment of the bone level on the buccal and lingual surfaces of mandibular molars.

OBJECTIVES To assess accuracy and precision of observers' diagnostic performance in locating the buccal/lingual bone level using conventional periapical radiographs. METHODS Standardized radiographs of molar regions of 10 dried mandibles were obtained. The bone crest on buccal and lingual surfaces of first and second molars was then marked with thin lead foil strips and fresh radiographs taken. A series of increasingly extensive furcation defects was created by sequential bone removal and radiographs with and without markers obtained after each step. The radiographs without indicators were presented to 10 observers on two separate occasions and the most apically located bone level marked on a transparent film. The distance between observers' markings and 'true' bone level on the radiographs was determined. RESULTS There was wide interobserver variation which increased as the bone loss increased. Lack of consistency was largely attributable to intra-observer variation except when a furcation defect was pre...