backgroundThis study aimed: (a) to determine the diagnostic performance of cone-beam computed tomography (CBCT) for detection of vertical root fractures (VRFs); (b) to evaluate the predictive value of diagnostic criteria regarding the definition of VRFs; and (c) to examine the robustness of the association of patient-, tooth-, and treatment-related variables with VRFs. MATERIAL AND

methods130 root-filled teeth with signs/symptoms of VRFs underwent clinical and CBCT assessments. Definite diagnosis of VRF was confirmed by endodontic microsurgical (EMS) exploration. Determination of diagnostic performance of CBCT was based on standard algorithms derived from two-way contingency table analysis. Predictive value of diagnostic criteria and the association between predictor variables with VRFs were analyzed using logistic regression models.

resultsVRFs were detected during EMS in 50% of the teeth. Based on the finding of fracture lines on CBCT scans, sensitivity, specificity, and accuracy were 86.2%, 13.8%, and 50%, respectively. Teeth having more than three diagnostic criteria present had significant higher odds for VRF diagnosis. After logistic regression analysis, parafunctional habits, one-canal roots, excessive root canal enlargement, and absence of intra-radicular posts remained as robust predictor variables of VRFs.

conclusionsAlthough the sensitivity of CBCT for VRFs detection is high, the risk of false-positive results related to its low specificity makes that all suspected cases must be confirmed by surgical exploration. VRFs cannot be reliably diagnosed by isolated clinical signs/symptoms; instead those teeth possessing more than three diagnostic criteria might be considered practically pathognomonic. The parafunctional habits, one-canal roots, excessive root canal enlargement, and the absence of intra-radicular posts may act strongly/independently for the occurrence of VRFs in endodontically treated teeth.