Variability of study cast assessment among orthodontists

This study examined the accuracy and reliability with standardized definitions among 30 orthodontists in assessing different aspects of malocclusion from the study casts of 10 patients. Previous studies have indicated a lack of consistency among orthodontists in describing malocclusions. The diagnostic subcategories measured in this study included maxillary and mandibular crowding, overbite, overjet, and a modified molar, premolar, and canine Angle classification. Intraclass correlations indicated that the reliability of all parameters was good (intraclass correlation > 0.80), with overbite being the most reliable (0.98) and maxillary crowding and canine classification somewhat less reliable (0.89 and 0.87, respectively). Among Angle classifications, molar classification was the highest (0.95), followed by premolar (0.92) and canine (0.87). This study also measured the accuracy of the orthodontists' measurements compared with established true values (validity). T tests indicated that mandibular arch crowding and overbite were not statistically different from the true values. Overjet and molar classification were assessed very close to the true values. Although premolar and canine classifications were statistically different from the true values, they were within a one-eighth cusp of the true value (a clinically insignificant difference). The orthodontists consistently overestimated the amount of maxillary arch crowding or spacing. The results suggest that high reliability and reasonable accuracy can be expected for assessing study casts when using standardized definitions for the diagnostic subcategories.