Development of an assessment strategy in preclinical fixed prosthodontics course using virtual assessment software—Part 1

Abstract The purpose of this study was to develop an assessment strategy for preclinical foxed prosthodontics using a virtual assessment software. This descriptive study examined 80 collected ivory teeth from previous classes prepared during preclinical fixed prosthodontics course. Ivory teeth prepared for the complete cast (tooth no. 46), metal ceramic (teeth nos 24 and 46), and all ceramic (tooth no. 21) crowns were scanned and superimposed with their respective standard preparations. Differences between these teeth and standard teeth prepared by faculty were computed using Compare software to generate comparison percentages. In addition, average finish line width, average total occlusal convergence, average axial wall height, and undercut presence/absence were quantified for student preparations using the software. Software‐generated values were then descriptively compared with faculty assessments. Comparison percentages aligned with faculty assessments for the amount of occlusal/incisal reduction and finish line location. The average axial wall height and finish line width calculated by Compare software were grouped based on their respective faculty assessment criteria. Software‐generated comparison percentages may be used to assess the amount of occlusal/incisal reduction and finish line location in student preparations. Additionally, averages extracted from Compare software could be used to assess student performance for axial wall height and finish line width.

advancing to clinical practice. Lack of consistency in assessment may lead to confusion in students' understanding of the principles of tooth preparation, as well as lack of improvement in the psychomotor skills required for tooth preparation.
In 1982, Mackenzie, Antonson, Weldy, and Simpson (1982) described 16 areas that contribute to inconsistent assessment, including checkpoint ambiguity, faculty memory, incomplete coverage of dimensions, unspecified exceptions, untrained estimation of size, undersized aids to judgment, unspecified methods of observing, incomplete operational definitions, unsystematic inspections, discrepancies in visual acuity, degrees of leniency, inadequacy of verbal definitions, inadequate communication with nonverbal examples, definition ambiguities, differences in background, and differences in mental processing. To address the many factors that contribute to lack of faculty agreement about student work, including grading scales, rater calibration, training, and subjective influences, Knight (1997) suggested implementation of faculty calibration and well-defined grading criteria in order to overcome problems of inconsistent faculty assessment (Feil & Gatti, 1993). However, even after calibrating faculty evaluation of student work as acceptable or unacceptable, faculty members still frequently mark unacceptable student work as acceptable in practice (Haj-Ali & Feil, 2006). This issue becomes even more difficult when calibrating faculty for the use of larger grading scales (Fuller, 1972;Haj-Ali & Feil, 2006;Lilley, Bruggen Cate, Holloway, Holt, & Start, 1968;Salvendy, Hinton, Ferguson, & Cunningham, 1973;Sharaf, AbdelAziz, & El Meligy, 2007). Faculty assessment is also complicated by significant levels of disagreement that are often observed between graders when evaluating student performance (Fuller, 1972;Lilley et al., 1968;Salvendy et al., 1973). Moreover, when a grader evaluates the same work on separate occasions, discordance in grading has been observed (Fuller, 1972;Lilley et al., 1968;Salvendy et al., 1973). Despite all of the variables involved in assessment of student performance, the Commission on Dental Accreditation mandates use of assessment forms and faculty calibration for U.
In an attempt to eliminate human bias from assessment of dental student work, E4D Technologies developed a virtual assessment tool.
In 2013, Renne et al. suggested that technology could provide an alternative to faculty grading of dental student performance. This study reported that software could be used to consistently and reliably compare student tooth preparations to standard preparations.
They concluded that the numerical comparison generated by the software is more precise than faculty assessments. However, in 2015, Callan, Haywood, Cooper, Furness, and Looney found no correlation between faculty assessments and the percentage comparison values computed by the software. This discrepancy may arise from the fact that the percentage comparison feature does not assess numerous criteria considered fundamental to tooth preparation that are assessed by faculty. Therefore, the purpose of this descriptive study was to develop a quantitative assessment strategy to accompany the use of Compare software by considering the fundamental of tooth preparation for evaluating amount of occlusal/incisal reduction (O/IR), finish line location (FLL) and finish line width (FLW), axial wall height (AWH), total occlusal convergence (TOC), and presence or absence of undercut. The use of Compare software may provide an objective measurement when evaluating preparation for complete coverage restoration in preclinical setting.

| METHODS
A preexisting assessment form for fixed prosthodontics from the University at Buffalo School of Dental Medicine (UB SDM) was used as a template for this study. Traditionally, the following criteria have been used to evaluate students' performance for preparation of complete The standard preparation is defined as teeth prepared by the faculty following the design and the amount of reduction presented to students. However, to determine comparison%, the software calculates the discrepancy in reduction (over-reduction or under-reduction) between student preparations and standard preparations. The software calculates the percentage of the surface area of the student preparation that fell within the tolerable range of discrepancy from the standard preparation. The area within tolerances is defined as the comparison%.
Teeth have a complicated morphology; as a result, it would be mathematically impossible to calculate the actual surface area for a tooth prepared for a complete coverage restoration. However, when the complex anatomy of a prepared tooth is simplified, it resembles a frustum of a cone ( Figure 1). Then, the surface area of a tooth prepared for a complete coverage restoration can be estimated using the following formula: where "h" is the height of the tooth from finish line to occlusal table, "b" is the diameter of the tooth at the finish line, and "a" is the radius of the tooth on the occlusal table. The comparison% calculates the percentage of the surface area matching the standard preparation.
On the basis of the formula above, the comparison% should be influenced by the amount of O/IR and the FLL. Figure 2 shows a two-dimensional schematic of a frustum of a cone. In dentistry, with a constant height "h," the radius "a" decreases to "a 1 " for the same tooth preparation with increased TOC. It is therefore possible to calculate this change and evaluate the influence of increased TOC on the surface area of the frustum of a cone (tooth preparation). Figure 2c shows a superimposed image of Figure 2a where α is the angulation between the long axis of the tooth and the axial wall. The "c" measurement shows the reduction from radius "b" needed to yield radius "a 1 ." The following formulas can be used to calculate radius "a 1 ."

| Experiment
For documenting the student progress, ivory teeth prepared by students along with their faculty-evaluated assessment form during preclinical fixed prosthodontics are kept at UB SDM. The preparations used in this study are selected from the previous fixed prosthodontics  Eighty teeth (Kilgore International, Inc., Coldwater, MI, USA) prepared by dental students were selected for development of this assessment strategy. Preparations were selected based on facultyevaluated assessment forms in order to include a range of prepared teeth with "excellent," "standard," and "standard not met" scores for each criterion. Teeth were selected from preparations for complete cast crown of tooth no. 46, metal ceramic crown of teeth nos 24 and 46, and all ceramic crown of tooth no. 21, each 20. Prior to the fixed prosthodontics course, the course director prepared a tooth for student observation for each restoration following the same criteria.
These preparations were defined as the standard preparations. The design and the amount of reduction for each tooth are presented in

| RESULTS
Video S1 depicts a Microsoft Excel file containing the formulas shown in Section 2. When the value of "h" increases or decreases, the amount of surface area changes considerably. Changes in "h" are influenced by the magnitude of the O/IR and the FLL. However, if radius "a" or "b" increases or decreases, the amount of surface area does not change significantly. Changes in radii "a" and "b" are influenced mostly by changes in TOC and FLW, respectively. The mathematical calculation demonstrates the minimal influence an increase in TOC has on the total surface area of the frustum of a cone. members was used to place tooth preparations into assessment categories (excellent, standard, or standard not met). Figure 3 shows the

| DISCUSSION
Simplified mathematical calculations describing tooth preparation demonstrate that for the same tooth, when the AWH "h" changes, the surface area of the preparation changes considerably. Video S1 shows that a 0.5-mm increase in "h" has a substantial influence on the surface area of a prepared tooth. In dentistry, changes in AWH occur as a result of changes in the amount of O/IR and in FLL. In addition, when "h" is constant, the radius "a" for the same tooth preparation decreases to "a 1 " with an increase in TOC. Video S1 also shows that when "α" (angulation between the long axis of the tooth and the axial wall) increases from 3°to 10°and 20°, only minor changes occur in the total surface area.
Descriptive results from this study show that the comparison% can serve as a surrogate for the amount of O/IR and the FLL for student tooth preparations. Analysis of data from 80 prepared teeth revealed that comparison% ≥85% were associated with excellent faculty evaluations for the amount of O/IR and the FLL. However, 85 > comparison% ≥ 70 and comparison% < 70 were accounted for standard and standard not met in faculty evaluations, respectively.
Results for average FLW from Compare software were grouped by FLW from faculty assessment in Table 3. Table 4 shows the average  Table 4 are accurate if each finish line design is considered to cover 180°of the tooth. Results of this descriptive ranking for the average FLW (Table 3) align with ideal average FLL (Table 4).
Our descriptive study did not address the quality of the finish line, a crucial factor in tooth preparation for complete coverage restorations. It is important for students to be attentive to the continuity and evenness of the finish line and the presence or absence of unsupported enamel, as well as to choose an appropriate finish line design for preparations with two finish line designs (as occurs in preparation for metal ceramic crowns). These qualitative criteria are subjective and cannot be easily measured or detected using available virtual evaluation software. Therefore, when using Compare software for assessment, in addition to average FLW, it is also important to implement FIGURE 7 Comparison percentage of student complete cast crown preparations (beige color) superimposed on the standard complete cast crown preparation (gray color). (a) Student preparation with an "excellent" finish line location and occlusal reduction leading to a comparison percentage of 87%. (b) Student preparation with an "excellent" finish line location and a "standard" occlusal reduction leading to a comparison percentage of 83%. (c) Student preparation with an "excellent" finish line location and a "standard not met" occlusal reduction leading to a comparison percentage of 48% Suggested average AWH for molars and premolars and suggested mid-lingual AWH for anterior teeth are summarized in  Figure 9). Pairing the Loma Linda TOC guide with Compare software could provide a less subjective means of TOC assessment.
The Compare software also failed to recognize the presence of clinically relevant undercut in the student preparations. These data suggest that the software measures detailed information in detecting undercut, but this detailed information does not appear to be clinically relevant.
Following this study and after the curriculum committee approval, syllabus was modified to incorporate the use of Compare software at   MCC 24 X ≥ 3 3 > X ≥ 2.5 X < 2.5 ACC 21 X ≥ 2 2 > X ≥ 1.5 X < 1.5 Note. E = excellent; S = standard met; N = standard not met; CCC = complete cast crown; MCC = metal ceramic crown; ACC = all ceramic crown.

FIGURE 9
Loma Linda total occlusal convergence guide superimposed on a prepared premolar. The proposed method to measure total occlusal convergence of a tooth sliced mesio-distally at the central groove UB SDM preclinical curriculum fixed prosthodontics. In this course, Tables 3, 4, and 5 were used to evaluate the amount of O/IR and FLL, average of FLW, and average of AWH, respectively. In addition, the faculty members assessed TOC, finish of the preparation, quality of the finish line, and adjacent teeth.

| CONCLUSIONS
Compare software may be used to evaluate a complete coverage preparation by objective measurement. Within the limitations of this descriptive study, the following conclusions can be drawn: 1. Compare-generated comparison% in 350-μm tolerance can be used as quantitative measurement of the sum of the amount of O/IR and FLL.
2. Average FLW calculated using Compare software could serve as a quantitative measurement of FLW using the suggested ranges.
3. Ranges shown for the average AWH of posterior teeth and midlingual AWH of anterior teeth can be used as criteria for quantification of AWH.
4. Average TOC and measurements of the presence or absence of undercut by Compare software should not be used to assess student preparations.