Managing adult degenerative/de novo scoliosis patients presents challenges due to the unique characteristics of each deformity, the complexity of classification. The rotation or translation of vertebrae can lead to the displacement of surrounding vital structures, narrowing the safe zones for pedicle screw insertion. This study evaluates the effectiveness of 3-dimensional (3D) scoliotic models in orthopedic training, specifically for teaching pedicle screw placement in scoliosis surgery. The goal is to validate the efficacy of a 3D model-navigated training system. Twenty life-sized 3D models of de novo scoliotic lumbar deformity were utilized by 20 residency trainees, categorized based on their years of experience: those with <3 years and those with 3 years or more. Each trainee practiced pedicle screw placement across 5 lumbar levels. Their performance was assessed through postinstrumentation CT scans for accuracy and detection of screw malpositions. Instances of screw malposition and breaches of the pedicle’s medial and lateral walls were recorded. The distribution of total cortical and critical penetrations was analyzed based on years of experience and lumbar levels. The training significantly enhanced the speed and accuracy of pedicle screw placement, reducing the incidence of errors. Trainees with less than 3 years of experience were 6 times more likely to misplace screws, particularly at higher levels of lumbar deformity. The average time to complete the simulation was approximately 34 minutes, with a 77.5% accuracy rate for screw placement. Odds ratios indicated that less experience increased the risk of critical cortex penetration 6-fold, and screw placement at the level of deformity increased the risk of poor placement by 4.7 times. The use of 3D models in surgical training enables residents to safely gain practical experience, highlighting potential for broader application in complex orthopedic procedures. This training system supports increasing task complexity and enhancing performance in actual surgical settings.
Öztürk et al. (Fri,) studied this question.