ABSTRACT
This study tested the reliability and subtraction frequency of the study model-scoring system of the American Board of Orthodontists (ABO). We used a sample of 36 posttreatment study models that were selected randomly from six different orthodontic offices. Intrajudge and interjudge reliability was calculated using nonparametric statistics (Spearman rank coefficient, Wilcoxon, Kruskal-Wallis, and Mann-Whitney tests). We found differences ranging from 3 to 6 subtraction points (total score) for intrajudge scoring between two sessions. For overall total ABO score, the average correlation was .77. Intrajudge correlation was greatest for occlusal relationships and least for interproximal contacts. Interjudge correlation for ABO score averaged r = .85. Correlation was greatest for buccolingual inclination and least for overjet. The data show that some judges, on average, were much more lenient than others and that this resulted in a range of total scores between 19.7 and 27.5. Most of the deductions were found in the buccal segments and most were related to the second molars. We present these findings in the context of clinicians preparing for the ABO phase III examination and for orthodontists in their ongoing evaluation of clinical results.
Subject(s)
Educational Measurement/statistics & numerical data , Models, Dental/statistics & numerical data , Orthodontics/statistics & numerical data , Specialty Boards/statistics & numerical data , Adolescent , Adult , Female , Humans , Judgment , Male , Malocclusion/therapy , Middle Aged , Observer Variation , Orthodontics/education , Reproducibility of Results , Statistics, Nonparametric , Treatment Outcome , United StatesABSTRACT
This study examined differences in pretreatment severity and treatment outcome among orthodontic patients treated in 3 different practice-management modes. Samples of pretreatment (T1) and end of treatment (T2) study casts were selected from traditional private practices (TPP, 3 offices, 81 cases), a dental corporation (COMP, 2 offices, 53 cases), and a dental management service organization (DMSO, 1 office, 36 cases). Orthodontic specialists had treated all patients. Cases were initially selected on a consecutive start basis. From each practice, the first 30 cases satisfying the study criteria were included in the sample. The T1 and T2 study casts were evaluated with the PAR and HLD indexes. The PAR and HLD indexes showed a high level of agreement on T1 cast scores but not on the T2 casts. Mean T1 scores were highest in the COMP cases, followed by the DMSO and the TPP cases. T2 scores were lowest in the TPP cases, followed by the DMSO and the COMP cases. The percentage of PAR score reduction showed that, in all 3 modes, patients were treated to a high standard.
