The accuracy of preoperative orthognathic predictions.

We assessed the accuracy of preoperative OPAL trade mark orthognathic predictions by retrospective analysis of 25 Class II patients who had had orthodontic treatment combined with mandibular advancement osteotomy. Preoperative and postoperative lateral cephalographs were digitised and surgical predictions generated using OPAL software. Each prediction was compared with the corresponding clinical changes. We also made a method error study by doing a random retracing of 25 cephalographs. Predictions of some of the principal OPAL values (SNA, ANB, LAFH%, OJ, OB) were reasonably accurate in terms of mean values. However, there were large individual variations for most variables and predictions of the vertical skeletal, incisor, and Wits measurements were imprecise. In particular, there was a bias towards under-prediction of the vertical skeletal changes when there was more backward mandibular rotation than anticipated. Immediate postoperative cephalographs were also affected by a 2.1mm mean downward displacement of the mandible as a result of the surgical wafer.