The prediction of impacted versus spontaneously erupted mandibular third molars

To evaluate the predictive capacity of orthodontists and oral maxillofacial surgeons (OMFSs) in anticipating the process of impaction or eruption of lower third molars (L3Ms) through the examination of serial panoramic radiographs.

Sixty-eight lower third molars (L3Ms) were analyzed in 34 orthodontically treated patients without extraction. Twenty-seven OMFSs and 27 orthodontists were randomized in order to analyze the radiographs. Initially, the evaluators issued the prognosis for the L3Ms in XR1, a posterior for the XR1 + XR2. Concordance of the diagnosis was examined using Kappa statistics, and the differences between the groups of evaluators were examined using the chi-square test at p<0.05.

When examining XR1 in cases where the teeth erupted spontaneously, the prognostic accuracy rate for OMFSs and orthodontists was similar, 63 and 65.7%, respectively (p=0.19). When evaluating XR1 + XR2, the accuracy among orthodontists (60%) was similar to that reported for XR1 (p=0.19), while OMFSs presented a reduction in the accuracy (55.3%, p<0.0001). When the L3Ms remained impacted, accuracy in XR1 was lower than in spontaneously erupting L3Ms, although similar between OMFSs (50.1%) and orthodontists (49.1%). Furthermore, for impacted L3Ms, when examining XR1 + XR2, the OMFSs presented a significant higher accuracy (71.8%, p <0.0001).

Orthodontists and OMFSs seem unable to predict spontaneous eruption or impaction of the lower third molars from single or longitudinal x-rays. When adding a second longitudinal x-ray, orthodontists and more significantly OMFSs tend to indicate more extractions.