RESUMO
The purpose of this research was to test the reliability of a modified magnification method for determining the position of an impacted canine from a single panoramic radiograph. This retrospective study evaluated 114 panoramic radiographs showing 136 impacted maxillary canines. The widths of the impacted canines, contralateral erupted canines, and ipsilateral maxillary incisors were measured, and ratios for the canine-incisor index (CII) and canine-canine index (CCI) were calculated. The impacted canines were also classified according to their location in the vertical plane (apical, middle, or coronal zone) relative to the contralateral central incisor. Continuous data were analyzed for normal distribution, and logistic and multivariate logistic regression models were conducted. The Benjamini-Hochberg procedure with a false discovery rate of 0.05 was used to account for multiple testing. The intrarater reliability was excellent for impacted canine, central incisor, and contralateral canine measurements (intraclass correlation coefficient > 0.9). The CII and vertical zone were strong predictors of an impacted canine position with clinically useful sensitivity and specificity values (0.69 and 0.74, respectively, based on an area under the curve concordance statistic of 0.75). A predictive range was evident for the CII of palatally (1.10-1.39) and buccally (0.90-1.19) impacted canines in the middle and coronal zones, respectively. The occurrence of palatal or buccal positioning was not significantly associated with the CCI (P = 0.2). The CII and vertical zone identified from a single panoramic radiograph can be used to determine the buccopalatal position of an impacted canine, with more reliability if the impacted canine crown is in the middle or coronal zone of the contralateral central incisor.