Cone-beam computed tomography or conventional radiography for localising of maxillary impacted canines?

Data sourcesPubMed, Web of Science, CINHAL and the Cochrane Library were searched until May 2016. Unpublished data were searched in Pro-Quest Dissertation, Abstracts and Thesis and Google Scholar, supplemented with manual search of the included studies references. No language restriction was used.Study selectionAll types of study designs were included, except case reports, comparing CBCT data with conventional radiographs. The primary outcome was: diagnostic accuracy between modalities, agreement in position, treatment planning and outcome efficacy. The secondary outcome was intermodality agreement in lateral root resorption detection and intra and inter-observer agreement values.Data extraction and synthesisTwo reviewers independently selected the studies for inclusion, performed data extraction and evaluated risk of bias. Discrepancies were resolved by discussions and reaching consensus. The Newcastle-Ottawa Scale was used to assess the risk of bias for case-controlled and cohort studies and a modified version for cross-sectional studies. The Quality Assessment of Diagnostic Accuracy (QUADAS-2) tool was used to rate diagnostic accuracy studies.ResultsEight studies met the inclusion criteria, two on diagnostic accuracy. The remaining six included 292 impacted canines in 224 patients. Outcomes were presented as calculated level of agreement and statistical significance for each primary outcome reported. Based on the diagnostic accuracy of two in vitro studies, the CBCT accuracy ranged from 50% to 95% while for conventional radiographs it ranged from 39% to 85%.The other six studies reported inter-modalities agreement in localisation (six studies) and treatment planning agreement (three studies). The inter-modalities agreement varied from 0.20 to 0.82, with observed agreement of 64% to 84% in localisation of canine. The treatment planning agreement varied from 0.36 to 0.72.ConclusionsThe authors concluded that CBCT is more accurate than conventional radiographs in localising maxillary impacted canines and there is a broad range of inter-observer and modalities agreement for location and treatment planning. There is no robust evidence to support using CBCT as first line imaging method.