Convergent Validity of BEWE in Clinical Examination and Digital Models

Objectives: To determine the convergent validity of using the Basic Erosive Tooth Wear Examination (BEWE) to assess Erosive Tooth Wear (ETW) using digital models obtained with two intra-oral scanners compared to conventional clinical examination.
Methods: Eighty subjects were enrolled after signing the written informed consent. They were allocated into two groups (n=40): with high risk for ETW (determined by a questionnaire based on the presence of at least one etiological factor of ETW) and low risk. They were then subjected to a clinical examination by a trained operator to screen for ETW using the BEWE. Afterward, subjects had their arches scanned with two intra-oral scanners (T5: TRIOS5 – 3shape and VV: Virtuo Vivo – Straumann), and the BEWE was determined on the digital models. Data were analyzed with Friedman and Mann-Whitney tests. The relationship of BEWE in the scans and clinical examination was evaluated by Bland-Altman plots and intraclass correlation coefficient (95%CI). Weighted kappa values (95%CI) using the quadratic approach were used to compare the results considering the BEWE categorized according to the risk. The significance level was set at 5%.
Results: For all subjects, the scores of T5 models (median/IQ=6.00/5.00-8.00) were higher (p<0.001) than the ones of VV (5.00/4.00-7.50), which in turn were higher (p<0.001) than the clinical examination (5.00/4.00-7.00). For all methods, the scores of the high-risk group were higher than the low risk (p<0.001). Overall, the scores of the T5 models had lower agreement values with the clinical examination than the VV. VV showed higher agreement with clinical examination (weighted kappa/95%CI= 0.897/0.757-1.000) than T5 (0.797/0.626-0.968) in the risk categorization based on BEWE scores.
Conclusions: BEWE scores are higher in digital models than in clinical examinations. The scores also vary according to the scanner used, as BEWE scores in T5 models were higher than in VV.