Statistical assessment of two methods of measuring scoliosis before treatment.

Regression analysis has shown that the Ferguson and Cobb methods of measuring spinal curvature give comparable results, at least before the scoliosis is treated. When the Cobb angle was calculated by multiplying the Ferguson angle by 1.38, the results were accurate to within 5 degrees in 28 (76%) of 37 cases. Any further debate over the merits of the two methods should focus on their inter- and intraobserver reproducibility and their applicability after corrective treatment.