The objectives of this study were to evaluate the repeatability and reproducibility of a method for measuring freehand rod bending and to analyze the relationship between the rod's bend and the resulting sagittal correction.

All the children who underwent correction by posterior translation using pedicle screws at all levels were included prospectively in 2018 and 2019. The rod's sagittal parameters were measured retrospectively by three independent surgeons on two separate occasions using the same protocol. After the rods were bent but before they were inserted, the surgeon traced the contours of the rods on a sheet of paper that was later scanned and analyzed semiautomatically. The spinal parameters were calculated based on biplanar radiographs taken preoperatively, postoperatively and at the final follow-up visit. Patients who had less than 10° thoracic kyphosis (T5–T12) made up the “Lenke N−” subgroup.

Thirty patients were included (14 of whom were Lenke N−) who had a Cobb angle of 59.2±11.3° preoperatively and 13.3±8.4° postoperatively (p<0.00001). The inter- and intrarater ICC for the rod measurements were>0.9 (excellent). The mean kyphosis of the concave rod was 48.4±5.7° (38.3–60.9°). The mean change in T5–T12 kyphosis was 9.7±10.8° (−14.3–30.8°) (p<0.0001) in the entire population, while it was 17.7±7.1° (5.5–30.8°) (p<0.0001) in the Lenke N− subgroup. The change in thoracic kyphosis was positively correlated with the kyphosis of the concave rod (rho=0.52; p=0.003).

This study found excellent reproducibility and repeatability of measuring freehand rod bending. The kyphosis applied to the concave rod is positively correlated to the change in the resulting kyphosis and made it possible to restore satisfactory thoracic kyphosis.

III.