Abstract
Objective
To determine if the Basic erosive tooth wear index (BEWE index) is able to assess and monitor ETW changes in two consecutive cast models, and detect methodological differences when using the corresponding 3D image replicas.
Methods
A total of 480 pre-treatment and 2-year post-treatment orthodontic models (n = 240 cast models and n = 240 3D image replicas) from 120 adolescents treated between 2002 and 2013 at the Gent Dental Clinic, Belgium, were scored using the BEWE index. For data analysis only posterior sextants were considered, and inter-method differences were evaluated using Wilcoxon Signed Rank test, Kappa values and Mc Nemar tests (p < 0.05). Correlations between methods were determined using Kendall tau correlation test.
Results
Significant changes of ETW were detected between two consecutive models when BEWE index was used to score cast models or their 3D image replicas (p < 0.001). A strong significant correlation (τb: 0.74; p < 0.001) was shown between both methods However, 3D image-BEWE index combination showed a higher probability for detecting initial surface changes, and scored significantly higher than casts (p < 0.001). Incidence and progression of ETW using 3D images was 13.3% (n = 16) and 60.9% (n = 56) respectively, with two subjects developing BEWE = 3 in at least one tooth surface.
Conclusions
BEWE index is a suitable tool for the scoring of ETW lesions in 3D images and cast. The combination of both digital 3D records and index, can be used for the monitoring of ETW in a longitudinal approach. The higher sensibility of BEWE index when scoring 3D images might improve the early diagnosis of ETW lesions.
Clinical significance
The BEWE index combined with digital 3D records of oral conditions might improve the practitioner performance with respect to early diagnosis, monitoring and managing ETW.
1
Introduction
Erosive tooth wear (ETW), which is the chemical-mechanical process of tooth surface loss caused principally by extrinsic and/or intrinsic acids, has become an important topic in dental research the last decade [ ]. Current dietary habits involving higher consumption of acidic food and beverages explain the increasing concerns with respect to the occurrence of ETW [ ], and this is reflected by the several prevalence studies of ETW that have been lately published [ ]. In 2015, an estimation of the worldwide prevalence indicated that 30% of children and adolescents were affected by some kind of ETW [ ], and in adults a multicenter study of 7 European countries indicated that 57.9% of them had at least one anterior tooth surface affected by ETW [ ]. The available data suggests that ETW is a common condition and that there is an apparent increase in the worldwide prevalence [ ].
On the other hand, assumptions with respect to the incidence and progression of ETW remains difficult: firstly, the available studies investigating this topic are scarce and secondly, their methodology to detect progression of ETW differs substantially [ ]. For instance, only five longitudinal studies assessed ETW progression in adolescents and all of them use different age groups, indexes and follow-up times, which complicate any comparison between results [ ]. For example, the first longitudinal study performed in 2001 used pre-orthodontic cast models found an incidence of 18% with 17.7% progression of ETW in a period of 5 years for 11 year-old adolescents [ ]. Those results are difficult to compare with the results of the latest prospective longitudinal study done in 2016, where progression of ETW was clinically evaluated using a different age group (13–14 year-old) and a longer follow-up time (4 years) [ ]. Nevertheless, despite these methodological differences present among the available longitudinal studies, all these results agree with the fact that ETW progresses in time, even after a 1.5 year period [ ]. This suggests that ETW should be detected and monitored from early ages, in order to prevent and diminish any progression of ETW.
The diagnosis of initial stages of ETW however, is difficult to perform [ ], and this limitation might be the first barrier for the correct monitoring of lesion progression. Therefore, there is need of chairside tools able to improve the diagnose, detection and record ETW lesions [ ]. Up to date, diagnostic tools such as the BEWE index have been recommended for the recording of ETW in cross-sectional studies and at the dental practice [ ]. This index has the additional advantage of providing guidance for the clinical management of patients according to their level of ETW risk, which can be low, medium or high depending on the severity reached. Despite these advantages, the criteria description that defines each level of ETW in this index provides general information, and therefore their creators do not recommend it for progression studies of ETW, where detailed information is needed [ ]. Nevertheless, the use of this index combined with additional records such as photographs, cast models or 3D images provided by intra-oral scanners could improve its performance for the monitoring of ETW between appointments [ ].
Previously, BEWE index demonstrated a good performance for the scoring of ETW in photographs [ ] and in 3D images [ ]. Recording 3D image data might improve significantly the detection and monitoring of ETW lesions based on the fact that digital images can be zoomed in, rotated and measured [ ]. In the past decade, the use of intra-oral scanners in the dental practice has increased [ ], and several brands are promoting these technologies [ ]. A recent study demonstrated that BEWE index was reliable for the score of 3D models, and suggested that this method of record could be useful for the monitoring of progression of ETW [ ]. However, up to date, their use in a longitudinal study combined with the scoring provided by BEWE index has not been tested.
The lack of longitudinal studies related with ETW and the need of improvements for the early diagnosis of ETW and between appointments monitoring have encouraged the present study. Therefore, the primary aim of this retrospective longitudinal study was to determine if BEWE index is able to assess and monitor ETW changes between two consecutive cast models, and detect methodological differences when using their 3D image replicas, and to determine the progression of ETW lesions during a two-year time period in a group of Belgian adolescents.
2
Materials and methods
2.1
Study population and ethical aspects
The target baseline cohort of this longitudinal retrospective study comprised all adolescents (age range 11–13 year olds) who were treated between 2002 and 2013 in the Orthodontic Department of the Gent University Hospital, Gent, Belgium, and who had pre- and post-orthodontic cast models available in their records. Only models from patients aged 11, 12 and 13 at baseline (pre-orthodontic model) and who had a follow-up model after a minimum time of 1.8 years and maximum time of 2.8 years, were included in this study.
Exclusion criteria for models were any absence of first permanent molars, bad quality models having presence of casting pearls or voids, broken teeth, presence of orthodontic appliances, such as brackets or orthodontic bands that could compromise the final score of the model.
This study obtained approval of the local ethical committee of the Gent University Hospital, which follows the “ICH Good Clinical practice” of the declaration of Helsinki (BC2016/0615 & 2016/0616).
2.2
Examination and score of the models
The presence and severity of ETW lesions was recorded using two indices: the BEWE index and the Erosive index (EI). EI index was previously used in a longitudinal study for the assessment of cast models [ ], and for this study it was exclusively chosen in order to compare the BEWE index performance to score cast models in a longitudinal basis. BEWE index was used as suggested by Bartlett et al. [ ]. For BEWE index, all permanent tooth surfaces of posterior teeth (4 sextants) were evaluated using a 4-step categorical scale (0–3). The 4 score criteria used during examination were: ‘0’ an indication for the absence of ETW, ‘1’ an indication for initial loss of surface texture (visually detectable), ‘2’ an indication for distinct defect, hard tissue loss less than 50% of the surface area, and ‘3’ an indication for hard tissue loss equal or more than 50% of the surface area (scores 2 and 3 can also involve dentin). The score of the most affected tooth per sextant represented the final sextant score. BEWE sum was calculated by adding the cumulative sextant scores of four posterior sextants.
2.3
Examiners and calibration
Two examiners (FM & LD) were educated and instructed by an experienced senior (LM) over a two-month period in relation to ETW, using photographs of diverse clinical cases of tooth wear and ETW displayed in a Power Point presentation. For calibration, each examiner was challenged to score a total of 36 cast models with different grades of ETW (not included in the study) using both indices (EI and BEWE) and in a second instance they had to score a set of 3D images of other 36 clinical cases using BEWE index. The 3D images and the cast models were scored at tooth level, by sextant or by the final BEWE sum. To determine intra-examiner agreement, every examiner had to re-score after 1 month the same casts and 3D images. The examinations of models started only after the inter-examiner agreement reached a kappa value equivalent to the cut off 0.61–0.8 or higher indicated by the scale of Landis & Koch (substantial level or higher).
2.4
Cast examination
The scoring of the cast models was made under standard illumination (LED light source) in the same room. Only the visible surfaces of the permanent dentition of the posterior sextants, with exception of third molars and partially erupted teeth were evaluated. Anterior sextants were excluded due to aesthetic modifications that some orthodontic patients received at the end of the treatment. In addition, the presence of different types of tooth wear such as wedge-shaped defects, attrition (matching facets, flat, and sharp bordered) and abrasion were not considered as affected (score 0 in both indices), when the role of this type of wear was major and no typical features of acidic interaction were present.
2.5
3D image examination
3D image replicas from the cast models at baseline and follow-up were obtained from a confocal intra-oral scanner TRIOS ™ (3Shape) and were transferred to the software Preview 8.1 (OSX10.11.4), which allowed the zooming in, rotation and inclination of the images during examination of the models ( Fig. 1 ). The same examiners and parameters described for the cast models were used to score 3D images. Score procedure was performed 1 month after the cast examination and the examiners did not have access to the previous scores given for the cast models.
2.6
Data management and statistical analysis
The presence of ETW was determined using as cut-off BEWE sum >0 and EI >0. For inter-method agreement, BEWE sum and the scores obtained by each posterior sextant were analysed separately. Additionally, changes according to the level of risk were determined by the cut-off criteria originally proposed by Bartlett et al. [ ], which uses BEWE sum >2–8, BEWE sum >8–13 and BEWE sum >13–18 as indicators for low, medium and high risk respectively [ ]. Incidence, distribution and progression of the lesions between baseline and follow-up were analysed using the aforementioned cut-offs, the mean BEWE sum and using the total number of affected teeth for each BEWE criterium (BEWE = 0, = 1, = 2 and = 3).
Data were analysed using IBM SPSS v. 22.0 and v. 24.0 (SPSS Inc., Armonk, NY.). Intra and inter- examiner reliability was calculated using kappa coefficients. Overall, descriptive statistics were performed and to compare methods of scoring (EI, BEWE in cast and BEWE in 3D models) Wilcoxon Signed Rank test, Paired sample test and Mc Nemar test at significance level of p < 0.05 were used. Correlations between methods (significance level p < 0.01) were determined using Kendall tau correlation test.
2
Materials and methods
2.1
Study population and ethical aspects
The target baseline cohort of this longitudinal retrospective study comprised all adolescents (age range 11–13 year olds) who were treated between 2002 and 2013 in the Orthodontic Department of the Gent University Hospital, Gent, Belgium, and who had pre- and post-orthodontic cast models available in their records. Only models from patients aged 11, 12 and 13 at baseline (pre-orthodontic model) and who had a follow-up model after a minimum time of 1.8 years and maximum time of 2.8 years, were included in this study.
Exclusion criteria for models were any absence of first permanent molars, bad quality models having presence of casting pearls or voids, broken teeth, presence of orthodontic appliances, such as brackets or orthodontic bands that could compromise the final score of the model.
This study obtained approval of the local ethical committee of the Gent University Hospital, which follows the “ICH Good Clinical practice” of the declaration of Helsinki (BC2016/0615 & 2016/0616).
2.2
Examination and score of the models
The presence and severity of ETW lesions was recorded using two indices: the BEWE index and the Erosive index (EI). EI index was previously used in a longitudinal study for the assessment of cast models [ ], and for this study it was exclusively chosen in order to compare the BEWE index performance to score cast models in a longitudinal basis. BEWE index was used as suggested by Bartlett et al. [ ]. For BEWE index, all permanent tooth surfaces of posterior teeth (4 sextants) were evaluated using a 4-step categorical scale (0–3). The 4 score criteria used during examination were: ‘0’ an indication for the absence of ETW, ‘1’ an indication for initial loss of surface texture (visually detectable), ‘2’ an indication for distinct defect, hard tissue loss less than 50% of the surface area, and ‘3’ an indication for hard tissue loss equal or more than 50% of the surface area (scores 2 and 3 can also involve dentin). The score of the most affected tooth per sextant represented the final sextant score. BEWE sum was calculated by adding the cumulative sextant scores of four posterior sextants.
2.3
Examiners and calibration
Two examiners (FM & LD) were educated and instructed by an experienced senior (LM) over a two-month period in relation to ETW, using photographs of diverse clinical cases of tooth wear and ETW displayed in a Power Point presentation. For calibration, each examiner was challenged to score a total of 36 cast models with different grades of ETW (not included in the study) using both indices (EI and BEWE) and in a second instance they had to score a set of 3D images of other 36 clinical cases using BEWE index. The 3D images and the cast models were scored at tooth level, by sextant or by the final BEWE sum. To determine intra-examiner agreement, every examiner had to re-score after 1 month the same casts and 3D images. The examinations of models started only after the inter-examiner agreement reached a kappa value equivalent to the cut off 0.61–0.8 or higher indicated by the scale of Landis & Koch (substantial level or higher).
2.4
Cast examination
The scoring of the cast models was made under standard illumination (LED light source) in the same room. Only the visible surfaces of the permanent dentition of the posterior sextants, with exception of third molars and partially erupted teeth were evaluated. Anterior sextants were excluded due to aesthetic modifications that some orthodontic patients received at the end of the treatment. In addition, the presence of different types of tooth wear such as wedge-shaped defects, attrition (matching facets, flat, and sharp bordered) and abrasion were not considered as affected (score 0 in both indices), when the role of this type of wear was major and no typical features of acidic interaction were present.
2.5
3D image examination
3D image replicas from the cast models at baseline and follow-up were obtained from a confocal intra-oral scanner TRIOS ™ (3Shape) and were transferred to the software Preview 8.1 (OSX10.11.4), which allowed the zooming in, rotation and inclination of the images during examination of the models ( Fig. 1 ). The same examiners and parameters described for the cast models were used to score 3D images. Score procedure was performed 1 month after the cast examination and the examiners did not have access to the previous scores given for the cast models.
