Reliability of cephalometric superimposition for the assessment of craniofacial changes: a systematic review

Superimposition of serial cephalometric radiographs enables the assessment of craniofacial changes over time, and therefore, several methods have been suggested in the literature. The aim of the present study is to summarize and critically evaluate the available evidence on the reliability of methods used to superimpose serial cephalometric radiographs. Electronic searches were performed in MEDLINE, EMBASE, Google Scholar, and Cochrane Databases, without time limit (last update: 1 November 2020). Unpublished literature was searched on the Open Grey and Grey Literature Report databases. Studies that tested the accuracy, precision, or agreement between different cephalometric superimposition techniques, used to evaluate the craniofacial changes due treatment or growth. Reference lists of relevant articles were screened and authors were contacted, if needed. All study selection steps, data extraction, and risk of bias (QUADAS-2 tool) assessments were performed independently by two authors on predefined forms. There were 27 eligible studies. From these, 17 tested superimpositions methods on the anterior cranial base, 10 on the maxilla and 12 on the mandible. There were three studies that compared superimpositions on the cranial base with those on the maxilla and one that compared the cranial base with the mandibular superimposition. There was high heterogeneity among studies in terms of sample size, growth, radiographic machines, selection criteria, superimposition methods, references, and outcomes measured. Furthermore, almost all studies presented important methodological limitations, with only two studies having unclear risk of bias and the rest 25 presenting high risk. Currently, there is no cephalometric superimposition method that has been proved to deliver accurate results. There is an urgent need for further research in this topic, since this is a primary assessment method to assess craniofacial changes over time for several relevant disciplines. PROSPERO (CRD42020200349).