• Yong-Min Kim, Seung-Myung Choi, and Min-Yong Ahn.
• Department of Orthopaedic Surgery, School of Medicine, Chungbuk National University, Chungbuk National University Hospital, 776, 1sunhwan-ro, Heungdeok-gu, Cheongju-si, Chungcheongbuk-do, 361-763, Cheongju, Republic of Korea.
• Arch Orthop Trauma Surg. 2018 Jun 1; 138 (6): 745-755.

ObjectiveRadiologic parameters are important factors for planning the treatment for thoracolumbar fracture. However, we noted that measurements of the degree of kyphosis by lateral decubitus plain radiography were greater than supine CT. The cause of this discrepancy is unclear.MethodsWe retrospectively reviewed the plain radiographs and CT scans of 90 patients with thoracolumbar fractures (fracture group). We measured the segmental sagittal angle (SSA) on lateral decubitus plain radiographs and in the median sagittal plane on CT scans obtained in the supine position. The method agreement (plain radiography versus CT) was determined by utilizing Bland-Altman plots. For the purpose of comparison, the same analyses were performed in a group of age and sex-matched controls (normal group). After establishing the method disagreement in the fracture group, the factors that contributed to the difference in the SSA between plain radiography and CT, as well as their threshold values, were determined.ResultsOn Bland-Altman plots for the fracture group, the mean difference was 4.53° [95% confidence interval (CI) - 4.87° to 13.93°]. For the normal group, the mean difference was - 0.64° (95% CI - 5.87° to 4.58°). On univariate analysis, male sex, thoracolumbar level, and SSA(X) were significant factors associated with ∆SSA (P = 0.03, 0.002, and 0.000, respectively). Multivariable regression analysis showed that SSA(X) was the only significant factor. Receiver operating characteristic curve analysis indicated that the optimal threshold of SSA(X) was 17° with a sensitivity of 78% and a specificity of 75% (area under curve: 0.752).ConclusionsThe mean SSA determined on lateral decubitus plain radiographs indicated significantly more kyphosis than that determined on CT images obtained in supine position. When the SSA on plain radiography is more than 17°, there might be a significant discrepancy between the two imaging modalities. This discrepancy seems to be mainly attributable to the difference in patient positioning (lateral decubitus position for plain radiography versus supine position for CT imaging).

28 keywords...

hide…