The effect of varying diagnostic thresholds upon clinical caries data for a low prevalence group.

This study investigated the impact of employing differing diagnostic thresholds on clinical caries data in studies of groups with low caries prevalence. Data from clinical examinations of 287 Hong Kong dental students were analyzed by means of the CARIES microcomputer software package. This software allows for re-calculation of raw data according to three different diagnostic thresholds (D3, D2, and D1). When "enamel" and "initial" lesions (as defined by WHO criteria) were included in the calculation of DMFT, its value increased from 3.0 (D3) to 5.9 (D1), while the percentage of individuals considered "caries-free" decreased from 28.2% to 7.0%. Little change was found in the magnitude of the intra-examiner reproducibility, when calculated at each threshold, for a random 10% of the subjects. It was unfortunately not possible to calculate inter-examiner reproducibility in this study. The use of criteria which might be misinterpreted as being similar, but which use differing effective diagnostic thresholds, can dramatically influence the reported level of dental caries. In view of these findings, it may be necessary for the question of diagnostic thresholds to be re-examined and to receive greater emphasis in future studies.