Intraobserver and interobserver reliabilty of measures of kyphosis in thoracolumbar fractures

Abstract

Background context: Consensus documents have recently been developed enumerating the radiographic parameters thought to be most valid in the clinical evaluation of patients with thoracolumbar fractures.

Purpose: The objective of this study was to assess the measurement reliability of plain X-rays, computed tomography (CT), and magnetic resonance imaging (MRI) and their inter-modality agreement, as the three imaging modalities are often clinically interchangeable. This process is an essential reliability evaluation of the measurement parameters being proposed.

Study design: This study evaluated the interobserver and intraobserver reliability of plain radiographs, CT, and MRI measurements of sagittal kyphosis in thoracolumbar fractures.

Patient sample: Suitable plain X-ray, CT, and MRI radiographic imaging of ten cases of thoracolumbar fracture were examined.

Methods: Suitable plain X-ray, CT, and MRI radiographic imaging of ten cases of thoracolumbar fracture were examined by ten independent spine surgery fellowship-trained observers.

Outcome measures: Cobb angle measurement, Gardner segmental deformity angle, and anterior body compression percentage were measured.

Results: Regardless of the imaging modality or the parameter being measured, the intraobserver reliability is always better than the interobserver. Plain radiography has better overall, interobserver and intraobserver reliability, followed by CT and then MRI. Reliability is very high in general, with the highest reliability for intraobserver reliability of the linear measures on plain radiographs. The inter-modality agreement is highest for plain X-ray and CT.

Conclusions: This study demonstrates that Cobb angle measurement, Gardner segmental deformity angle, and anterior body compression percentage are reliable measures of thoracolumbar fracture kyphosis with very high interobserver and intraobserver reliability and very high inter-modality agreement of plain X-ray with CT.