The accuracy of computerized growth prediction in Class II high-angle cases

This study was conducted to test the accuracy of a commercially available forecasting system in predicting the effects of growth and orthodontic treatment. The pretreatment cephalograms and wax bites of mandibular casts of thirty-three consecutively treated Class II patients with high mandibular plane angles, along with twenty-six criteria related to treatment preference, were submitted for analysis. All patients had already been treated on a nonextraction basis by a single practitioner using high-pull face-bow headgear. The computer-generated posttreatment predictions or visual treatment objectives (VTO's) were compared to the actual posttreatment cephalograms, using twenty-one linear and nine angular measurements. Fifteen of the thirty parameters evaluated showed statistically significant (P less than 0.01) differences between the actual posttreatment result and the computer prediction. The computer was found to be accurate in predicting the effects of growth and treatment on maxillary position and rotation, mandibular length, upper face height, and incisor positions. It was found to be inaccurate in predicting the effects of growth and treatment on maxillary length, mandibular rotation, lower anterior and posterior face heights, the horizontal and vertical positions of the molars, and over 50% of the soft-tissue parameters.