RESUMO
OBJECTIVES: To investigate the validity, reliability, and time spent to perform a full orthodontic study model analysis (SMA) on cone-beam computed tomography (CBCT)-generated dental models (Anatomodels) compared with conventional plaster models and a subset of extracted premolars. SETTING AND SAMPLE POPULATION: A retrospective sample of 30 consecutive patient records with fully erupted permanent dentition, good-quality plaster study models, and CBCT scans. Twenty-two extracted premolars were available from eleven of these patients. MATERIALS AND METHODS: Five evaluators participated in the inter-rater reliability study and one evaluator for the intrarater reliability and validity studies. Agreement was assessed by ICC and cross-tabulations, while mean differences were investigated using paired-sample t-tests and repeated-measures anova. RESULTS: For all three modalities studied, intrarater reliability was excellent, inter-rater reliability was moderate to excellent, validity was poor to moderate, and performing SMA on Anatomodels took twice as long as on plaster. CONCLUSIONS: Study model analysis using CBCT-generated study models was reliable but not always valid and required more time to perform when compared with plaster models.