Variability in measurement of radiomorphometric indices by general dental practitioners

To assess the variability of general dental practitioners (GDPs) in measurement of radiomorphometric indices on panoramic radiographs following basic instruction and to examine whether the variability could be reduced by more individualised instruction.

Nine GDPs measured Gonion Index (GI), Antegonion Index (AI), Mental Index (MI) and Mandibular Cortical Index (MCI) on copies of 10 panoramic radiographs following a lecture on osteoporosis and the use of radiomorphometric indices. Their measurements were related to expert-derived measurements of the same copy radiographs. Mean differences and limits of agreement (2x standard deviation of differences) were calculated for quantitative indices (GI, AI, MI) and agreement of GDPs with expert-derived MCI assessments was determined using weighted kappa. Following individualised feedback to GDPs, all measurements were repeated after 2 weeks and the statistical analysis repeated.

There was extensive variation amongst GDPs in measurement of GI, AI and MI and in assessment of MCI. There was a general tendency of GDPs to record thicker mandibular cortices than did the experts. Limits of agreement were wide relative to the mean values of each quantitative index at both readings. Agreement of the GDPs with experts in assessment of MCI was moderate at both readings, but with a wide range in assessment.

Variability in measuring radiomorphometric indices amongst the GDPs was high and was not predictably improved by individualised instruction. This study casts considerable doubt on the potential value of radiomorphometric indices given their lack of precision.