Diagnostic Performance of Isolated Orbital CT Scan for Assessment of Globe Rupture in Acute Blunt Facial Trauma

Abstract Objectives We determine the diagnostic performance of emergent orbital computed tomography (CT) scans for assessing globe rupture in patients with blunt facial trauma. Methods We performed a retrospective cohort study based on prospectively collected trauma registry and acute care surveillance data in a tertiary care hospital. Patients aged at least 18 years who received isolated orbital CT scans for assessing potential ocular trauma were examined. Analyses were performed to evaluate the magnitude of agreement between diagnosis by CT scan and ophthalmic assessment, including of globe rupture. Results Our study cohort comprised 136 patients, 30% of whom (41 patients) sustained orbital wall fractures. Concordance for orbital CT diagnosis and the ophthalmic assessment of globe rupture was substantial (k = 0.708). The relative risk of globe rupture was 0.692 (95% confidence interval [CI] 0.054 to 8.849) for superior wall fractures, 0.459 (95% CI 0.152 to 1.389) for inferior wall fractures, 2.286 (95% CI 1.062 to 4.919) for lateral wall fractures, and 0.637 (95% CI 0.215 to 1.886) for medial wall fractures. According to multivariate analysis, lateral wall fractures was an independent risk factor for globe ruptures (adjusted OR = 12.01, P=.011), and medial or inferior wall fracture was a protective factor (adjusted OR = 0.14, P=.012). In the stratified analysis of diagnostic performance of CT scan, specificity was highest among patients with orbital wall fractures (97.2%), followed by NPV (97%), and accuracy (95.1%). Conclusion Among patients with blunt facial trauma who underwent isolated orbital CT as part of ocular trauma assessment, the diagnostic performance of CT in detecting globe rupture is more accurate in patients with orbital wall fractures. Nevertheless, isolated orbital CT alone does not have high enough diagnostic performance to be reliable to rule out all globe ruptures.