RESUMO
PURPOSE: To evaluate the feasibility of developing visual diagnostic skills for detecting and assessing dental caries using e-learning assisted practice in preclinical dental education. METHODS: A one-group before and after the study was conducted. After a theoretical lesson on cariology, 53 inexperienced second-year preclinical students assessed 78 clinical photographs using the Nyvad criteria; they received automated feedback upon completion of the test. After a week, all students reassessed the same set of photographs, which were randomly reordered. Differential diagnostic accuracy was analyzed category-pairwise, and overall accuracy measures were based on the receiver operator curve. Diagnostic accuracy in both attempts was evaluated and compared through estimation and pooling of individual student accuracies. RESULTS: Pooled category-pairwise accuracy was lower for discriminating Sound surface from Non-cavitated-active caries, and for discriminating inactive caries (surface discontinuity) from intact surface inactive caries and Cavitated-active caries. Pooled overall accuracy, after the theoretical lesson, was 0.79 (95% confidence interval [CI] 0.77-0.81), and it increased to 0.99 (95% CI 0.98-0.99) after feedback. Between-student variability in accuracy was reduced from I2 = 0.66 to 0.55. CONCLUSION: E-learning assisted practice is a feasible alternative to start developing visual diagnostic skills for detecting and assessing dental caries using the Nyvad criteria from preclinical dental education in cariology. However, further studies are required to evaluate its effectiveness in improving real-world practice knowledge and skills.