Can preoperative imaging help to predict postoperative outcome after wisdom tooth removal? A randomized controlled trial using panoramic radiography versus cone-beam CT

Objectives: The primary objective of the study was to compare the postoperative complications following surgical removal of impacted third molars using panoramic radiography (PAN) images- and cone-beam computed tomography (CBCT)-based surgeries for "moderate-risk" cases of impacted third mandibular molars. The secondary objective was to compare the reliability of CBCT with that of PAN in preoperative radiographic determination of the position of the third molar, number of roots, and apical divergence. Materials and methods: A randomized controlled multicenter trial was conducted to compare the surgical complications of PAN- and CBCT-based surgeries of impacted third molars. The sample consisted of impacted third molars from 256 patients with a close relation to the inferior alveolar nerve (IAN). Exclusion criteria were "no risk" and "high risk" of damage to the IAN based on the assessment of the panoramic radiograph. Patients were divided into two groups: the CBCT group (n = 126) and the PAN group (n = 130). The incidences of IAN sensory disturbance and other postoperative complications were recorded for each group at 7 days after surgery. Statistical analysis (kappa values) was used to compare the diagnoses of five trained dentomaxillofacial radiologists and to relate radiologic diagnoses to perioperative findings. Logistic regression was used to determine whether the imaging modality influenced occurrence of postoperative complications. Results: Two extractions (1.5 %) in the CBCT group and five (3.8 %) in the PAN group resulted in IAN sensory disturbance (p = 0.45). Logistic regression models did not show that CBCT modality decreased postoperative complications following surgical removal of impacted third molars. Yet, CBCT revealed the number of roots and apical divergence of the roots more reliably than panoramic radiographs. Conclusions: CBCT was not better than panoramic radiography in predicting postoperative complications for moderate-risk cases of impacted third mandibular molars. Nonetheless, a CBCT buccolingual view can accurately confirm the number of roots and root morphology of the third molar better than PAN. Clinical relevance: For management of postoperative complications for moderate-risk cases of impacted mandibular third molars, careful preoperative radiographic planning followed by an atraumatic surgical approach seems to be valuable, irrespective of the 2D or 3D nature of the preoperative images. © 2013 Springer-Verlag Berlin Heidelberg.