Use of flexion and extension radiographs of the cervical spine to rule out acute instability in patients with negative computed tomography scans

Abstract

Objectives: To investigate the usefulness of flexion and extension radiographs of the cervical spine as a screening tool for the acute evaluation of ligamentous injury in cases of awake blunt trauma in patients with a negative cervical computed tomography scan.

Study design: Retrospective study of consecutive patient series.

Setting: Level I trauma center.

Patients: All patients admitted to an academic Level I trauma center over 12 months who sustained a blunt force injury and underwent flexion-extension radiography during hospitalization.

Intervention: The flexion-extension radiographs were interpreted for adequacy and pathology by two independent reviewers who were blinded to patient outcome and the original radiologic interpretation. Adequacy of radiographs was assessed using four criteria: 1) complete visualization of the cervical spine from the occiput to the superior end plate of the first thoracic vertebra; 2) adequate range of flexion and extension was defined as motion greater than 30° from the neutral position; 3) supplementation with a swimmer's view if the cervicothoracic junction was poorly visualized; and 4) no evidence of rotational deformity on neutral, flexion, or extension views. Radiographs were thus deemed either "adequate" or "inadequate." Acute instability was defined as listhesis of greater than 3.5 mm or 11° of relative angulation. Radiologists' interpretation of all studies was noted and any clinical or radiographic evidence of instability on follow-up within 3 months of discharge was also recorded.

Results: A total of 311 patients were included in the study. The intraobserver reliability for the four fixed criteria for adequacy of flexion and extension radiographs was excellent. Only 97 (31%) flexion and extension radiographs were deemed adequate. Two hundred fourteen (69%) patient radiographs were deemed inadequate but were interpreted as normal by the radiologists. Not a single radiograph was identified with evidence of acute instability (true-positive = 0). One hundred seventy-one (55%) of patients had follow-up within 3 months of discharge from the hospital of which one (0.5%) patient developed signs of instability necessitating surgery. The sensitivity was 0%, specificity 99%, positive predictive value 0%, and negative predictive value 31%.

Conclusion: Flexion and extension radiographs do not appear to be clinically useful in assessing acute instability in patients hospitalized with blunt trauma with negative computed tomography scans.