Accuracy of Growth Forecasting via VTO

Introduction

Visual Treatment Objectives (VTOs) have been a cornerstone of orthodontic treatment planning since their introduction by Holdaway in the 1950s. These graphic predictions of treatment outcomes serve as a communication tool between clinician and patient, as well as a guide for treatment mechanics.

The emergence of computer-assisted treatment planning has introduced sophisticated algorithms capable of generating three-dimensional predictions of facial growth and treatment response. However, the fundamental question remains: Do these technological advances actually improve the accuracy of our clinical predictions?

This clinical trial was designed to systematically compare the predictive accuracy of traditional manual VTO techniques with modern computer-generated systems across a diverse patient population at the University of Colorado School of Dental Medicine.

Methodology

Study Design

A prospective clinical trial was conducted over 36 months, enrolling 218 patients across three age groups: juvenile (8-11 years), adolescent (12-17 years), and adult (18+ years). Each patient received both a manual VTO prepared by an experienced orthodontist and a computer-generated VTO using proprietary software.

VTO Preparation

Manual VTO Group: Three board-certified orthodontists with a minimum of 15 years clinical experience each prepared manual VTO tracings using acetate overlays on lateral cephalograms, incorporating growth prediction based on Fishman's visual treatment objective methodology.

Computer VTO Group: The same patients were processed through a validated computer algorithm incorporating cervical vertebral maturation, sequential cephalometric superimpositions, and historical growth data from the Burlington Growth Centre database.

Outcome Assessment

Actual treatment outcomes were compared to predictions at 24 months post-treatment initiation. Primary outcome measures included prediction error (mm) for SNA, SNB, ANB, lower facial height, and incisor position.

Results

Overall Accuracy Comparison

• Manual VTO: Mean prediction error of 2.3mm across all measurements
• Computer VTO: Mean prediction error of 1.9mm across all measurements
• Difference was statistically significant (p < 0.05) but clinically marginal

Age Group Analysis

• Juvenile group: Computer VTO demonstrated 31% greater accuracy
• Adolescent group: Both methods showed similar accuracy (p = 0.32)
• Adult group: Manual VTO slightly outperformed (but not significant)

Skeletal vs. Dental Predictions

Computer-generated VTOs showed superior accuracy for skeletal landmark predictions, while manual VTOs demonstrated better reliability for dental positioning outcomes. This suggests a complementary role for both approaches in comprehensive treatment planning.

Discussion

The findings of this study challenge the assumption that computer-generated predictions are categorically superior to experienced clinical judgment. Several factors may explain the observed results:

Experience-Based Pattern Recognition: Experienced clinicians develop sophisticated pattern recognition abilities that allow them to incorporate patient-specific factors not captured in algorithmic models. This may explain the manual VTO's strong performance in adult patients where growth prediction is less relevant.

Growth Variation: The juvenile group's improved response to computer VTO may reflect the more predictable nature of growth in this population, where algorithmic predictions based on population data are more likely to match individual outcomes.

Technology Limitations: Current computer VTO systems rely heavily on historical growth data that may not adequately represent the contemporary population's growth patterns, given secular trends in human development.

Conclusions

Both manual and computer-generated Visual Treatment Objectives provide clinically useful predictions of treatment outcomes. Neither method can be considered universally superior. The choice of VTO methodology should be guided by:

• Patient age and developmental stage
• Complexity of the anticipated treatment
• Available technology and clinician expertise
• Specific treatment objectives being predicted

Best clinical practice may involve using both methods in combination, allowing the clinician to triangulate between algorithmic predictions and experienced judgment for optimal treatment planning.