Reliability assessment of cervical spine parameters measured on full-body radiographs in asymptomatic subjects and patients with spinal deformity.

Abstract Background Cervical spinal alignment is usually assessed on full-body radiographs allowing for the concomitant evaluation of possible compensatory mechanisms that may occur at any level in the setting of postural malalignment. Hypothesis Cervical parameters measured on full-body radiographs are reliable. Patients and methods A total of 70 subjects were included and divided in 3 groups: asymptomatic adults (n = 21), adolescents with idiopathic scoliosis (n = 20), and adults with spinal deformity (n = 29), for whom full-body low-dose biplanar radiographs were obtained. Eighteen cervical parameters including gaze and cervical curvature, upper cervical spine, global cervical alignment, thoraco-cervical and cervico-pelvic parameters were measured by 4 operators, three times each. The intraclass correlation coefficient (ICC) and the 95% confidence interval (95% CI) where calculated for each parameter and compared between the 3 groups. Results ICC and the 95% CI were similar between the 3 groups. The measured parameters showed a very high repeatability (ICC > 0.8) except for C0-C2, which presented an average repeatability (ICC = 0.57). The cSVA, CTPA, C2-SPi, cranial offset, T1-SPi, CBVA and cranial tilt had a 95% CI   6°. Discussion The poor visibility of the foramen magnum, hard palate, C7, T1, and the sternum on radiographs could explain why certain parameters showed a higher measurement error. The assessment of these error margins is essential for an accurate evaluation of cervical spinal deformities and a proper therapeutic approach. Level of evidence III; retrospective analysis of prospectively collected data.