There were two groups in our study including the earthquake trauma and control group. Patients entered the present study as earthquake trauma group if they had blunt abdominal crush injury complicated with or without other system crush injuries such as injury in thorax and pelvis; if clinically suggested abdominal injury was initially confirmed by CT not greater than 14 days after the outbreak of this earthquake; and if the etiology of abdominal injury was crush injury in the 2008 Sichuan earthquake on the basis of the history and the findings of rescue. Patients were excluded from the earthquake trauma group if the etiology of the abdominal injury was not crush injury but jump or accidental fall from buildings, based on the history of injury in the earthquake. Patients with blunt abdominal trauma were enrolled into control group according to the inclusion criteria similar with those in earthquake trauma group except that the etiology was not associated with an earthquake.

In the 2008 Sichuan earthquake, an undamaged key university hospital 92 kilometers away from the epicenter, equipped with 4,300 beds, working as the largest-scale urgent care center in the earthquake-affected areas, received and treated a total of 2,728 victims with earthquake-related injury over a period of 14 days. Among these victims, 51 patients (22 men and 29 women; mean age, 41.9 yr; age range, 7-86 yr) with blunt abdominal crush injury in the Sichuan earthquake, who met the inclusion criteria, entered the present study as earthquake trauma group. In this group, 49 patients with abdominal crush injury were in combination with suggested crush injuries in one or more of other anatomic systems including the thorax, pelvis, extremity, and neck in 39, 40, 11, and 4 patients, respectively. In order to timely detect the injury for appropriate emergency treatments in a great number of injured patients as soon as possible, two of five MDCT scanners in this Department of Radiology were utilized to image the injury. The MDCT scans were followed with a routine renal function test and a measurement of the myoglobin in blood and urine shortly after the admission, and acute renal failure were confirmed in 36 patients in earthquake trauma group. The mean time from crush injury to the CT scans was 5.4 days with a range from 6 hr to 14 days. During the time from injury to rescue, few dangerously ill patients survived before being evacuated to this university hospital to receive effective treatments. Some of them had received antibiotics in disaster areas to prevent infection. Based on the image findings and clinical data, 37 patients with intra- or retroperitoneal injury, fracture of lumber vertebra, or sever injury of abdominal wall underwent surgical treatments; and the remaining patients received conservative treatments. The retro- or intraperitoneal injury and severer injury of abdominal wall in patients who had received surgical treatments were reconfirmed by the operative findings. Owing to the appropriate treatments, 49 patients were cured, and 2 died of crush injuries.

Between 1st May and 6th June 2009, 41 consecutive patients (27 men and 14 women; mean age, 34.5 yr; age range, 2-60 yr) with abdominal non-earthquake blunt injury, who were admitted in this key university hospital and met the inclusion criteria, were enrolled into the present study as the control group. In the control group, patients had abdominal injury due to traffic accidents (n = 12), assault (n = 10), falls (n = 8), work-related accidents (n = 5), crush injury (n = 4) and collision injury (n = 2). In the control group, 34 patients had abdominal injury in combination with suggested injuries in one or more of other anatomic systems including the pelvis (n = 21), thorax (n = 16), extremity (n = 5), neck (n = 6), maxillofacial region (n = 9), and head (n = 10). In this group, emergency MDCT scans were performed similarly with that in the earthquake trauma group, and followed by appropriate emergency treatments. The mean time from the injury to the CT scans was 4.6 days with a range from 0.5 hr to 12 days. Owing to timely and appropriate treatments, 40 patients were cured, and 1 died of the injury.

In earthquake trauma group, 38 and 13 victims with abdominal crush injury underwent emergency scans with Philips Brilliance 16-section MDCT (Philips Healthcare, Eindhoven, Netherland) and Siemens Somatom Sensation 16-section MDCT (Siemens Medical Systems, Forchheim, Germany), respectively. Due to suspicion of acute renal failure in the massive earthquake, an emergency CT scan of abdomen was performed without intravenous contrast media from right diaphragmatic dome to pelvic basement as soon as possible. In the control group, all patients underwent similar non-enhanced emergency scans with Siemens Somatom Sensation 16-section MDCT. If the victims were suggested with other system injury, MDCT scans of the corresponding anatomic regions were performed along with scans of the abdomen to image all injuries. The scanning parameters used for both scanners were as follows: 120 kV, 250 mAs, gantry rotation time of 0.5 sec, pitch of 0.85, collimation of 16 × 0.75 mm, reconstructed section thickness of 5 mm, field of view of 380 mm, and matrix of 512 × 512 mm. If fractures of lumbar vertebrae were found, the reconstructed section thickness was 1 mm. Image data were transferred to a picture archiving communication system (Syngo-Imaging, Siemens Medical Solutions, Forchheim, Germany).

Image data were retrospectively reviewed by two radiologists (with 12 yr and 11 yr of experience in radiology, respectively) working in consensus focusing on the detailed anatomic regions involved and types of the abdominal traumatic lesions, and injuries of other anatomic regions. The abdominal regions involved were classified as retroperitoneum, intraperitoneal space, lumbar vertebrae, and abdominal-wall soft tissue. The fractures of lumbar vertebrae were reviewed in detail by multiplanar reconstruction with a slab of 5-7 mm and three-dimensional reconstruction. So as to better understand overall types of abdominal injuries, traumatic lesions in abdominal parenchyma organ, peritoneal pathological changes (seroperitoneum and pneumoperitoneum), and injuries of abdominal aorta were reviewed. In order to better demonstrate severity of abdominal injury, combined injuries of other anatomic regions such as the thorax and pelvis were reviewed according to the traumatic lesions demonstrated in the published reports (10, 11).

Data analysis was performed with a statistical software package (version 13.0 for Windows, SPSS Inc., Chicago, IL, USA). To investigate the features of abdominal earthquake-related crush injury, we compared lumbar vertebral fractures, injury of abdominal-wall soft tissue, retroperitoneal injury, intraperitoneal injury, and injuries of other anatomic regions between earthquake trauma group and control group by chi-square tests. Comparison between retro- and intraperitoneal injury was performed by the similar test. A P value of less than 0.05 was considered significant.

The present study was approved by the institutional ethic review board of our university hospital, and informed consent was waived due to the retrospective nature of this cohort study.

In earthquake trauma group, 24 of 51 patients (47.06%) had lumbar vertebral fractures (Fig. 1) in 49 vertebrae in comparison with 9 of 41 patients (21.95%) with the fractures in 25 vertebrae in control group. Lumbar vertebral fractures in earthquake trauma group occurred more frequently than in control group (P < 0.001). In groups, the fractures occurred in one or more vertebrae of L1-5 vertebra based on the anatomic levels involved. According to the structure of a lumbar vertebra involved, the fractures occurred in the transverse process, body, spinous process, or vertebral plate of the lumbar vertebra. Of these patients, the number of the lumbar vertebral fractures per patient ranged from 1 to 5. The involved vertebra levels, the structure, and the number of lumbar vertebrae are listed in Table 1.

Injuries of abdominal-wall soft tissue such as intramuscular hematoma, muscular swelling, fatty edema, or subcutaneous air collection were found in 54.9% patients (28/51) in earthquake trauma group, and in 31.7% patients (13/41) in control group. Injuries of abdominal-wall soft tissue were more common in earthquake trauma group than in control group (P = 0.019).

In earthquake trauma group, retroperitoneal injuries appeared as renal or perirenal injury, pancreatic injury, rupture of abdominal aortic aneurysm, swelling of perirenal fascia, and air collection in posterior perirenal space in 29 patients. In this group, renal or perirenal injuries appeared as hematoma, subcapsular hematoma, hemorrhage of renal cyst, and perirenal hematoma with or without swelling of perirenal fascia in 9 patients. Pancreatic injuries appeared as traumatic pancreatitis (Fig. 2A), or pancreatic rupture in combination with swelling of perirenal fascia in 3 patients. Renal injury in combination with pancreatic injury and abdominal aortic aneurysm (Fig. 2A, B) occurred in 1 patient, respectively. The remaining 15 patients in earthquake trauma group had swelling of perirenal fascia. In control group, 14 patients with retroperitoneal injuries manifested renal or perirenal injuries (n = 12), traumatic pancreatitis (n = 1), and renal injury in combination with traumatic pancreatitis (n = 1). Retroperitoneal injuries were more often seen in earthquake trauma group than in control group (P = 0.042).

Intraperitoneal injuries were found in 16 patients in earthquake trauma group, and were composed of liver injury in 7 patients, splenic injury in 2, hemoperitoneum in 5, and pneumoperitoneum in 2 (Fig. 2C). In control group, intraperitoneal injuries in 15 patients were composed of liver injury in 5 patients, splenic injury in 1, hemoperitoneum in 5, pneumoperitoneum in 3, and hydropneumoperitoneum in 1. In the two groups, incidences of intraperitoneal injuries were not different (P = 0.087). Retroperitoneal injury was more common than intraperitoneal trauma in earthquake trauma group (P = 0.008).

In earthquake trauma group, 43 patients had abdominal injury in combination with injuries of one or more other organs in the thorax (n = 29) and pelvis (n = 30). In control group, 29 patients had abdominal injury in combination with injuries of the thorax (n = 13) and pelvis (n = 15). The combined injuries of other organs are shown in Table 2. Abdominal injury in combination with injuries of other organs occurred more frequently in earthquake trauma group than in control group (P < 0.001).