Orthognathic surgery is considered nowadays as a revolutionary treatment option for treating skeletal discrepancies and severe malocclusions in the sagittal, vertical and transverse dimensions. This surgery allows both the restoration of facial harmony and the achievement of satisfactory dental occlusion. The technology of computer-assisted surgeries including virtual surgical simulation programs and planning software greatly contributes to providing a three-dimensional simulation and precise mobilization of the maxilla and/or mandible, thus allowing the prediction of the final outcome in soft tissues. This study aims to systematically review the available scientific literature about the accuracy of the hard and soft tissue predictions delivered by the many promoted three-dimensional simulation software.

An electronic search was conducted on various databases: Medline via PubMed, The Cochrane Library, EBSCO-host, and Web of Science. The search was established on a well-defined research question following PICO principle: population, intervention, comparator and outcome. Search evaluation and the assessment of risk of bias were undertaken in each study following its type and design.

Fifteen studies were included for qualitative analysis. Seven studies evaluated the accuracy of soft tissue prediction, seven focused more on the accuracy of hard tissue and one study assessed both hard and soft tissue prediction accuracy delivered by the simulation software. Moreover, three studies were judged to be low risk and four were classified as high risk. Included studies revealed that hard tissue prediction is highly accurate and reliable, leading to clinically acceptable results. Yet, soft tissue prediction is unclear due to various factors that bias its results. Caution should therefore be taken when providing information about the soft tissue planning to patients.

Computer assisted 3D simulation protocols allow for more precise repositioning of the maxilla and/or mandible compared to conventional 2D methods. However, 3D soft tissue prediction using simulation software remains less accurate, especially in the labial region.