Examiner's experience and the outcome interpretation of ICDAS and Nyvad's system - a prospective in vitro study.

Objective: To evaluate, prospectively, the influence of examiner's experience in interpreting and applying the caries detection systems ICDAS (IC) and Nyvad (NY).

Material And Methods: Twelve second-year undergraduate dental students (UG) and 12 postgraduates (PG) MSc level analysed and codified 77 clinical dental caries photographs at three different moments: initially, without any training; after one week of receiving training through a theoretical class; and after two years. Reproducibility and correlation was evaluated; sensitivity, specificity and area under ROC curve (AUC) were dichotomized according to the presence of cavitation (IC) and in relation to disease activity (NY).

Results: IC presented good kappa values for the first two evaluations. Both criteria resulted in good Spearman's correlation after two years (IC = UG: 0.89; PG: 0.93/NY = UG: 0.81; PG: 0.82). Sensitivity, specificity and AUC were statistically higher in the third evaluation by UG for Nyvad.

Conclusions: ICDAS criteria seem to be instinctively understood by students without clinical experience. Nyvad's concepts performed better after two years where the students deepened their theoretical knowledge and experienced clinical practice, collaborating with the identification of activity signs.