INTRODUCTION: The advent of pedicle screws which provide distraction and derotation has led to higher correction of major curves. Newer methods have been devised to evaluate preoperative coronal flexibility, including lateral-bending (LB), push-prone (PP) and fulcrum-bending (FB) radiographs. Documentation of a consistent radiographic method predictive of correction rate has not been established.

OBJECTIVE: To determine the most predictive radiographic method for evaluating spine flexibility and correction by comparing the correction rate (CR), flexibility rate (FR) and correction index (CI) of the Cobb's angle using the different radiographic methods.

METHODS: Preoperative radiographs of 20 patients who underwent spinal fusion for adolescent scoliosis were obtained using the LB, PP and FB method and compared with postoperative radiographs.

RESULTS: Comparing the mean Cobb angles using the different methods to that of postoperative standing showed that only the FB method is not significantly different from the latter (p=0.669). There was a significant difference between the Cobb's angle measured on the LB and PP and that measured on postoperative standing (p=0.043, p=0.008). Comparing the mean flexibility of the different methods with the mean CR also showed that the mean FR of LB (p=0.007) and PP (p=0.00013) were significantly different from the CR while that of FB is not significantly different from the CR (p=0.687).

CONCLUSION: The FB radiograph demonstrated no statistical difference compared to postoperative radiograph, FR, and CI.