The cervical spine is an essential element in the assessment of global posture and it has been demonstrated that compensatory mechanisms can occur to maintain horizontal gaze [1Diebo B., et al. Spine 2016 ;

Cliquez ici pour aller à la section Références]. To our knowledge, errors of measurement on head and cervical spine parameters are not yet known. The aim was to quantify the reliability of head and cervical parameters measured on full body radiographs.

Seventy subjects (30 with adult spinal deformity, 30 with adolescent idiopathic scoliosis and 20 asymptomatic subjects) underwent full body low dose biplanar radiographs in free-standing position and were included in this study. Four well-trained operators measured digitally, 16 cervical spine and head parameters on the lateral radiographs, 3 times each. The collected parameters were: radiographic chin brow vertical angle (CBVA), slope of line of sight (SLS: angle between the horizontal and Frankfort plane), angle between McGregor line and the lower endplate of C2 (OC2), C1–C2 angle, C2 slope, C2–C7 angle (cervical lordosis/kyphosis), T1 slope, cervical sagittal vertical axis (cSVA), cranial tilt, cervical tilt, cranial offset, neck tilt (NT), thoracic inlet angle (TIA), C2 pelvic angle (CPA), thoracic pelvic angle (TPA) and cervico-thoracic pelvic angle (CTPA). General reliability (including both inter and intra-operator) was assessed by calculating the intraclass correlation coefficient (ICC), the errors of measurement were evaluated as the confidence interval at 95% (CI95%) for each parameter (ISO 5725-2, IC95% calculated with a precision of 5%).

ICC values were high (>0.85) for all parameters. CSVA, CTPA, CPA, Cranial offset, TPA, CBVA, cranial tilt showed low CI95% values (0.30mm, 0.32^o, 0.36^o, 0.42mm, 0.44^o, 1.44^o, 1.64˚ respectively). More elevated CI95% were found for C2 slope, SLS, T1 slope, C1–C2, OC2, cervical tilt, C2–C7 (4.1^o, 4.3^o, 5.0^o, 5.1^o, 5.2^o, 5.1^o, 5.8^o respectively). TIA and NT showed the highest CI95% values (8.22^o and 6.16^o respectively).

All head and cervical parameters were deemed highly repeatable when measured on full body low dose radiographs. The TIA and NT had higher uncertainties due to the poor visibility of the sternum because of the superimposition of the bony structures on the lateral radiographs. This study established the thresholds on head and cervical spine parameters, below which any differences should be considered due to uncertainties and should not be interpreted as a consistent variation.