Effect of diagnostic threshold on the validity and reliability of epidemiological caries diagnosis using the Dundee Selectable Threshold Method for caries diagnosis (DSTM)

Objectives: To develop a method for recording dental caries at the D-1 (ena mel and dentine) diagnostic threshold (without loss of D-3 information) and assess its reliability, 'benchmark' validity and potential effects on repo rted caries prevalence and needs assessment. Methods: Multi-examiner traini ng, calibration and validation trial. Two groups of 10 dental examiners wer e trained to diagnose dental caries at the D-1 (enamel and dentine) diagnos tic threshold under the conditions of a caries prevalence survey, prior to a calibration trial being undertaken. Results: Where 'experienced examiners ' were trained to examine at the D-1 (enamel and dentine) diagnostic thresh old, under the conditions of a cross-sectional epidemiological survey, ther e was no significant deterioration in inter-examiner agreement on the asses sment of teeth and a significant difference in one of two comparisons on th e assessment of surfaces using the kappa statistic. Assessed against a benc hmark examiner, there was no significant loss of sensitivity at the D-1 dia gnostic threshold compared with the D-3 threshold and, although there was a significant loss of specificity at the D-1 threshold, all specificity valu es could be considered to be high. Conclusions: Modifying the diagnostic cr iteria typically used in surveys of caries prevalence (to allow assessment of the levels of enamel caries which could benefit from preventive care as well as dentinal caries requiring restorative care) in adolescents does not adversely affect the reliability or benchmark validity of experienced exam iners to a significant degree.