Earthquake-related versus non-earthquake-related injuries in spinal injury patients: differentiation with multidetector computed tomography

INTRODUCTION In recent years, several massive earthquakes have occurred across the globe. Multidetector computed tomography (MDCT) is reliable in detecting spinal injuries. The purpose of this study was to compare the features of spinal injuries resulting from the Sichuan earthquake with those of non-earthquake-related spinal trauma using MDCT. METHODS Features of spinal injuries of 223 Sichuan earthquake-exposed patients and 223 non-earthquake-related spinal injury patients were retrospectively compared using MDCT. The date of non-earthquake-related spinal injury patients was collected from 1 May 2009 to 22 July 2009 to avoid the confounding effects of seasonal activity and clothing. We focused on anatomic sites, injury types and neurologic deficits related to spinal injuries. Major injuries were classified according to the grid 3-3-3 scheme of the Magerl (AO) classification system. RESULTS A total of 185 patients (82.96%) in the earthquake-exposed cohort experienced crush injuries. In the earthquake and control groups, 65 and 92 patients, respectively, had neurologic deficits. The anatomic distribution of these two cohorts was significantly different (P < 0.001). Cervical spinal injuries were more common in the control group (risk ratio (RR) = 2.12, P < 0.001), whereas lumbar spinal injuries were more common in the earthquake-related spinal injuries group (277 of 501 injured vertebrae; 55.29%). The major types of injuries were significantly different between these cohorts (P = 0.002). Magerl AO type A lesions composed most of the lesions seen in both of these cohorts. Type B lesions were more frequently seen in earthquake-related spinal injuries (RR = 1.27), while we observed type C lesions more frequently in subjects with non-earthquake-related spinal injuries (RR = 1.98, P = 0.0029). CONCLUSIONS Spinal injuries sustained in the Sichuan earthquake were located mainly in the lumbar spine, with a peak prevalence of type A lesions and a high occurrence of neurologic deficits. The anatomic distribution and type of spinal injuries that varied between earthquake-related and non-earthquake-related spinal injury groups were perhaps due to the different mechanism of injury.