Prediction of thoracic dimensions and spine length on the basis of individual pelvic dimensions: validation of the use of pelvic inlet width obtained by radiographs compared with computed tomography

Abstract

Study design: Retrospective review.

Objective: To validate the pelvic inlet width (PIW) measurement obtained on radiograph as an independent standard used to correlate with thoracic dimensions (TDs) in treated and untreated patients with early-onset scoliosis.

Summary of background data: In children with early-onset scoliosis, the change in TD and spine length is a key treatment goal. Quantifying this change is confounded by varied growth rates and differing diagnoses. PIW measured on computed tomographic (CT) scan in patients without scoliosis has been shown to correlate with TD in an age-independent manner.

Methods: The first arm included 49 patients with scoliosis who had both a CT scan and pelvic radiograph. Agreement between PIW measurements on CT scan and radiograph was analyzed. The second arm consisted of 163 patients (age, 0.2-18.7 yr), with minimal spinal deformity (mean Cobb, 9.0°) and radiographs in which PIW was measurable. PIW was compared with previously published CT-based TD measurements; maximal chest width, T1-T12 height, and T1-S1 height. Linear regression analysis was used to develop and validate sex-specific predictive equations for each TD measurement on the basis of PIW. Interobserver reliability was evaluated for all measurements.

Results: Bland-Altman analysis indicated agreement with no dependence on observed value, but a consistent 8.5 mm (95% CI: 7.2-9.9 mm) difference in CT scan measurement compared with radiographical PIW measurement. Sex and PIW were significantly correlated to each TD measurement (P < 0.01). Predictive models were validated and may be used to estimate TD measurements on the basis of sex and radiographical PIW. Intraclass correlation coefficients for all measurements were between 0.978 and 0.997.

Conclusion: PIW on radiographs and CT scan correlate in patients with deformity and with spine and TD in patients with minimal deformity. It is a fast, reliable method of assessing growth while lowering patient's radiation exposure. It can be reliably used to assess patients with early-onset scoliosis and the impact surgical treatment has on chest and spinal growth.

Level of evidence: 3.