RESUMO
INTRODUCTION: Limited field of view cone-beam computed tomography (LFOV CBCT) is the primary imaging modality recommended for treatment planning before endodontic microsurgery (EMS). Persistent apical periodontitis, often treated with EMS, results in changes in the buccal cortical plate that may detrimentally impact prognosis. The accuracy of a preoperative LFOV CBCT to predict intraoperative findings is unclear. METHODS: Electronic health records (EHRs) of EMS performed at 2 endodontic offices between 2016 and 2018 were reviewed retrospectively. EHR data extracted were documented for surgical findings of intact buccal cortical plate, fenestration, dehiscence, and height of remaining buccal collar of bone. Two calibrated, independent reviewers evaluated presurgical LFOV CBCTs in the multiplanar paraxial and parasagittal planes at 2 different reconstructed viewing plane thicknesses. Reviewer findings were compared with EHR documentation. Data were analyzed by using χ2, logistic regression, and multivariable analysis. Significance was set at P < .05. RESULTS: Within the 125 EMS cases included in the study, the EHR prevalence of intact buccal cortical plate was 49%, dehiscence 7%, and fenestration 44%. The imaging predictive value, whether it was negative (NPV) or positive (PPV), was higher when predicting presence of buccal bone (PPV of intact buccal cortical plate = 86.5%; NPV of dehiscence = 96%; NPV of fenestration = 89%). Sensitivity and specificity ranged from 80%-90%. Accuracy in prediction was high for all variables, exceeding 80%. Accuracy was not significantly influenced by reconstructed viewing slice thickness, viewing plane, or reviewer. CONCLUSION: Preoperative LFOV CBCT was highly discriminatory and accurately predicted intraoperative buccal cortical bone status, especially intact buccal cortical plate and fenestration.
