The increase in adult orthodontic patients in the last 20 years makes it more likely that a clinician will have adult cases to present to the American Board of Orthodontics (ABO) for board certification. The applicability of the Objective Grading System (OGS) to adult dentitions, however, has not been reported. Therefore, the purpose of this study was to determine if well-treated adult cases, as designated by orthodontists, could pass the ABO clinical examination using the OGS.
Posttreatment casts of 35 adult patients from Temple University, Department of Orthodontics were ranked by 4 examiners based on visual examination of the quality of case finish using the Q-sort method. Data on each examiner’s personal selection criteria to classify a well-treated case were gathered. Discrepancy index (DI) and OGS scores were measured, and the number of missing teeth was recorded.
There was a moderate correlation between the rankings of the 4 examiners and their rankings to the OGS score. A 63% overall ABO pass rate was found. The pass rate for the well-treated cases and failure rate in the lower ranked group were both 100%. The DI scores of 97% of the cases were high enough to qualify as a board case, and 77% had 1 or more missing teeth after treatment. Examiners largely used the factors of the OGS in deciding if a case was well treated, with intercuspation being the most important factor.
This study demonstrates that well-treated adult cases can pass the ABO clinical examination using the OGS. In addition, it illustrates that missing teeth automatically inflate the DI score and improve the OGS score.
Orthodontic treatment in adults has become much more common in the last 20 years and is currently the fastest growing subspecialty in orthodontics. The reasons for this increase include improvements in orthodontic appliances, a heightened awareness of the availability of treatment, an increased understanding of how orthodontic treatment can positively aid in other dental treatments, and an increased social acceptance of adult orthodontic treatment. Psychological motives also play a role in adults who seek orthodontic treatment, with the most motivating factor being a desire to improve dental appearance. Other factors include functional improvement, dental/periodontal health, and general health and speech.
With this need for adult orthodontic treatment, there is an increased likelihood that clinicians will have adult cases to present to the American Board of Orthodontics (ABO) for certification. Currently, the ABO offers a written and clinical examination for orthodontists who wish to become board certified. Although board certification is not required for orthodontists, it is encouraged. In striving to make the examination as fair as possible to all examinees, in 1999 the Objective Grading System (OGS) for case evaluation was implemented. As with any index to assess treatment outcomes, it is important to examine the OGS for validity and reliability.
The OGS is composed of the following 8 criteria: alignment, marginal ridges, buccolingual inclination, occlusal relationships, occlusal contacts, overjet, interproximal contacts, and root angulation. Many of these criteria use descriptors such as incisal edges, central grooves, marginal ridges, and cusps to describe the ideal final position of the teeth. In an intact young and healthy dentition, these anatomic landmarks are easy to find. In a worn or restored dentition, however, these landmarks may not be as easy to discern or may not be present at all. Adult occlusions typically fit into the latter group. With the increasing number of adult orthodontic patients, it is of value to determine if the OGS is applicable to adult dentitions. In general, when validating occlusal indices, the expert opinion of orthodontists has been considered the “gold standard.” The purpose of this study, therefore, was to determine if well-treated adult cases, as judged by expert orthodontists, can pass the ABO clinical examination using the OGS.
Material and methods
Subjects were selected from the existing records of consecutive patients over the past 11 years at Temple University, Department of Orthodontics. Prior to any data collection, institutional review board (IRB) approval was obtained. To be included in the study, the patient must have been more than 30 years of age at the onset of treatment, have had completed treatment, have had a pretreatment cephalometric radiograph, and have had pre- and posttreatment plaster models available. Eighty-three patients were identified, and of these 83, 35 subjects had complete records that met the inclusion criteria.
Qualifying patients’ posttreatment plaster models were randomly numbered. Four expert clinician evaluators were chosen from the faculty in the Department of Orthodontics at Temple University. The 35 posttreatment models were randomly laid out on a table. Examiners examined each model visually and ranked them according to the excellence of treatment finish, using the following Q-sort method: examiners were asked to select the best-treated and the worst-treated cases. These models were removed from the table. The examiners were then instructed to select the next 2 best- and worst-treated cases, and these models were removed from the table. They then selected the 3 best- and worst-treated cases, and these models were removed from the table. They were then instructed to select the next 4 best- and worst-treated cases, and these models were removed from the table. They were then instructed to select the next 5 best- and worst-treated cases. These models were removed from the table, after which 5 models remained. The cases were arranged into a normal-shaped distribution based on this sorting, with the best- and worst-treated cases being on either end of the distribution ( Table I ).
After all posttreatment models were ranked, examiners were given a questionnaire to collect information on their personal selection criteria used to classify a well-treated case. The examiner was not aware of the questionnaire before the sorting occurred.
Based on the ranking by the examiners, the models were grouped into higher-ranked (well-treated) and lower-ranked groups. Posttreatment models for all casts were then measured using the OGS. Root angulation was not measured or included in the OGS score in this study. The corresponding pretreatment models were also measured using the DI to determine the complexity of each case. Measurements were done by 1 investigator to ensure consistency. The investigator was calibrated to the OGS with the aid of the ABO OGS calibration kit. Information regarding the number of missing teeth after treatment was recorded from the casts.
Intraclass correlation coefficients (2-way mixed model, consistency type, with single measure reliability) were used to evaluate the interrater reliability for the 4 examiners’ rankings and to compare the 4 examiners’ rankings to the OGS score measured by the investigator. Calculations were done using SPSS for Windows software, version 16.0 (SPSS, Chicago, Ill). Descriptive statistics were used to evaluate the factors the examiners considered during their ranking process. A Pearson correlation was used to determine intrarater error for the DI and OGS measurements.
The case rankings (Q-sort results) for the 4 examiners (A, B, C, and D) are shown in Table I . The cutoff for the higher-ranked cases was chosen to be between columns 7 and 8 and for the lower-ranked cases between columns 4 and 5. This was done to allow for grouping of the cases with agreement among the examiners. Case numbers within shaded squares indicate cases that were placed in the higher-ranked case group (well-treated group) or the lower-ranked group by all 4 examiners. The rankings given by all 4 examiners correlate with each other at a moderate level ( r = 0.584, P <0.0001).
The OGS scores for the 35 cases are shown in Table II . Taking into account that root angulation was not included in the OGS score in this study, an OGS score of 27 points or less was considered passing. Twenty-two of the 35 cases had an OGS score low enough to pass the ABO clinical examination. There was a weak, yet statistically significant, to moderate correlation between ranks given by examiners B and C and the OGS scores ( r = 0.324, P <0.027 and r = 0.401, P <0.008, respectively) and a weak correlation between the ranks of examiners A and D and the OGS scores ( r = 0.135, P <0.216 and r = 0.205, P <0.115, respectively). The OGS score distributions as viewed by the examiner rankings are shown in Table III . The numbers marked with an asterisk have OGS scores from 0 to 27 and the numbers marked with a dagger have OGS scores from 28 up. A few of the nonpassing OGS scores appear in the higher-ranked group and vice versa; however, the general trend is for the cases with passing OGS scores to be ranked toward the higher-ranked end and the cases with the nonpassing OGS scores to be ranked toward the lower-ranked end, although this is not statistically significant.
|Case Number||OGS Score||DI Score||Missing Teeth||Case Number||OGS Score||DI Score||Missing Teeth|
The DI scores for the 35 cases are shown in Table II . All but one of the cases had a DI score high enough to qualify as a board case. Thirty-four of the 35 cases had a DI of 10 or higher and of these 17 had a DI of 20 or greater.
The missing teeth for each case are listed in Table II . Twenty-six of the 35 cases involved 1 or more missing teeth. Teeth were missing because of previous loss, congenital absence, or extractions for orthodontic treatment. A total of 93 teeth were missing in this sample after treatment, with 84 of them being posterior teeth.
Cases 2, 10, 15, and 21 were identified by all examiners as being higher ranked, or well treated ( Table I ). All 4 of these cases would pass the clinical portion of the ABO examination with OGS scores of 14, 20, 18, and 23, respectively ( Table II ).
In addition, the cases that were lower ranked by all 4 examiners were cases 12, 17, and 19 ( Table I ). None of these 3 cases would have passed the ABO clinical examination, having OGS scores of 50, 30, and 41, respectively ( Table II ).
The results of the questionnaires that were given to the examiners after they ranked the cases can be found in Table IV . The degree to which each factor played a role in the examiner’s decision was marked on a visual analog scale with a total length of 100 mm.