The utility and accuracy of computed tomography in the diagnosis of occipitocervical dissociation.

Background Context: Atlantooccipital dislocation (AOD) results in profound patient morbidity and mortality and is difficult to accurately diagnose using current evaluation techniques.

Purpose: To evaluate the utility of computed tomography (CT) images in the diagnosis of AOD and compare the revised occipital condyle-C1 interval (CCI) and the condylar sum to the current radiographic criteria used to detect AOD.

Study Design: Retrospective review to evaluate the sensitivity, specificity, and the interobserver reliability of eight radiographic criteria as applied to CT imaging.

Patient Sample: Ten cases of clinical AOD and 10 cases of non-AOD cervical injury.

Outcome Measures: Measured values: revised CCI, Wholey basion-dens interval (BDI), and Harris basion-axis interval (BAI). Calculated values: Sun interspinous ratio, Powers ratio, and condylar sum. Assessment of Lee X-line and atlantooccipital joint asymmetry.

Methods: A board certified neuroradiologist, two orthopedic spine surgeons, and two medical students reviewed the CT images for each patient in the series and applied the aforementioned criteria.

Results: Average sensitivity between all reviewers for CCI, condylar sum, and atlantooccipital asymmetry was highest at 1.0, 1.0, and 0.96, respectively. Basion-dens interval, X-line, Sun ratio, BAI, and Powers ratio had sensitivities of 0.72, 0.54, 0.32, 0.26, and 0.26, respectively. Revised CCI and condylar sum had significantly better sensitivity than any other test (vs. BDI, p=.014, all others, p<.001) except atlantooccipital asymmetry (p>.99). Specificity for all measurements was 0.78 or greater, except X-line at 0.38. Interobserver reliabilities were the greatest for CCI, condylar sum, atlantooccipital asymmetry, and BDI.

Conclusions: The revised CCI (>2.5 mm abnormal) and condylar sum (≥5 mm abnormal) are highly sensitive and reliable radiographic criteria for the detection of AOD when applied to CT imaging.