Does the positioning of the patient in the EOS machine affect the validity of cervical and head parameters calculated on whole-spine radiographs?

The aim was to study the validity of free-standing position using modifiers for assessing cervical spine alignment. A total of 26 asymptomatic subjects (45 ± 11years) and 13 with spinal deformities (30 ± 21years) underwent lateral radiographs using EOS®, while adopting 3 different positions, in 2 sequences, without mirror and then with mirror (m) placed in front of them: first, the neutral standing position (upper extremities positioned on the sides of the body, P1 and P1m), then free-standing (P2 and P2m) and modified free-standing (P3 and P3m) positions (shoulders flexed at 90° and fingers placed on clavicles then cheeks, respectively). Participants were asked not to move their trunk and shoulders when moving from P1 to P2 and then P3. Eighteen cervical radiographic parameters were calculated from the 6 radiographs. The effects of shoulder flexion, mirror placement, and finger positioning on cervical alignment were studied using a repeated-measures ANCOVA. Mirror placement had no effect on cervical alignment. Concerning the effect of shoulder flexion, C2-C7 and T1 slopes were significantly decreased in P2 and P2m (C2-C7 = 10° and 8°; T1 slope = 26° and 24°, respectively) and P3 and P3m (C2-C7 = 9° and 9°; T1 slope = 25° and 26°, respectively) compared to P1 and P1m (C2-C7 = 14°and 14°; T1 slope = 29°and 30°, respectively), without exceeding the uncertainty level. Assessment of cervical alignment is valid on whole-spine radiographs when the shoulders are flexed (free-standing position). Clear instructions regarding the shoulder position at 90°, without moving the trunk, neck, and head before placing the fingers on the clavicles or cheeks, can ensure a cervical alignment comparable to that obtained with the neutral position, which is considered the most physiological one.