Negative Finding From Computed Tomography of the Abdomen After Blunt Trauma

Despite the focus on time management, cost-efficient health management, resource utilization, and the growing evidence regarding the long-term effects of radiation exposure, the liberal use of computed tomography (CT) for trauma remains common. Determining which patients require CT imaging and what percentage of negative CT findings is an acceptable counterpart to potential missed cases of occult injury is a topic of significant debate.^1,2 The use of negative CT findings after trauma as a trigger for early hospital discharge has been shown to decrease hospital costs.³ In the current era, however, a missed injury after trauma is often regarded as a “never event.” Although CT imaging has become a highly reliable adjunct to a physical examination after trauma, concern remains regarding its sensitivity and specificity in detecting hollow viscus injury.^4,5 Despite the growing number of patients with negative CT findings, it remains unclear at what point it is safe to clear these patients for hospital discharge. Given the sensitivity of physical examinations for posttraumatic intra-abdominal injury and of CT scans for solid organ injury, we hypothesized that a negative CT finding for an asymptomatic patient after blunt abdominal trauma is sufficient to exclude major intra-abdominal injury.

All blunt trauma patients admitted in 2013 who underwent CT of the chest, abdomen, and pelvis on admission were evaluated, and those who underwent CT of the abdomen and pelvis and had negative findings formed the study group. A negative CT finding was defined as a CT scan revealing no abnormalities aside from incidental findings noted on the final report. All images were read by an attending radiologist. During this period, all patients with a mechanism sufficient to trigger a CT scan were observed after imaging to evaluate for delayed injury.

Patients’ demographics, injuries, results of physical examinations, external signs of trauma, and durations of observation were recorded. The primary outcome was a delayed injury diagnosis.

The study was reviewed and approved by the institutional review board of the University of Southern California. The present study was a retrospective study requiring no patient interaction, and, as such, informed consent was waived.

In total, 620 patients had negative findings from abdominal imaging and were admitted to the surgical observation unit. The majority of trauma injuries were due to motor vehicles (303 patients [48.9%]), automobiles vs pedestrians (117 patients [18.9%]), and motorcycle collisions (76 patients [12.3%]). The mean (SD) Glasgow Coma Scale score was 14.7 (0.9), and the mean (SD) Injury Severity Score was 5.0 (4.4), with only 15 patients (2.4%) with an Injury Severity Score higher than 15. External signs of trauma included ecchymosis or hematoma in 71 patients (11.5%), most commonly a “seat belt sign,” or abdominal wounds in 35 patients (5.6%). Abdominal tenderness was noted in 171 patients (27.6%). Overall, 324 patients (52.3%) remained in the hospital for observation for more than 24 hours (Table). The majority of these patients remained in the hospital for orthopedic management (n = 168), continued workup or treatment of comorbid conditions (n = 57), or additional imaging (n = 55). The remaining 296 patients (47.7%) were discharged from the hospital within 24 hours. No delayed abdominal injuries were noted after negative CT findings, and this included patients who initially presented with abdominal pain and the seat belt sign.

A CT scan of the abdomen with a negative finding was highly specific for ruling out intra-abdominal injury. Although our study is limited by being a single-center, retrospective analysis, the large volume and consistent postimaging observation period strengthen the results. Patients were routinely observed after a negative finding with the specific intent to evaluate for delayed presentation of hollow viscus injury with serial laboratory tests and abdominal examinations. Although a significant percentage of patients had abdominal pain or external signs of trauma at presentation, no patient with a true negative CT finding developed delayed abdominal injury. Although further prospective analysis with multicenter participation will be needed to better characterize these results, based on these data, asymptomatic evaluable patients with a negative abdominal CT finding can likely be safely discharged from the hospital or dispositioned to another treating service.

Corresponding Author: Elizabeth R. Benjamin, MD, PhD, Department of Surgery, Los Angeles County + University of Southern California Medical Center, 2051 Marengo St, Inpatient Tower C5L-100, Los Angeles, CA 90033 (elizabeth.benjamin@med.usc.edu).

Published Online: September 9, 2015. doi:10.1001/jamasurg.2015.1649.

Author Contributions: Dr Benjamin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Benjamin, Haltmeier, Lofthus.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Benjamin, Lofthus.

Critical revision of the manuscript for important intellectual content: Benjamin, Siboni, Haltmeier, Inaba, Demetriades.

Statistical analysis: Benjamin, Siboni, Lofthus.

Administrative, technical, or material support: Siboni, Haltmeier, Demetriades.

Study supervision: Benjamin.

Conflict of Interest Disclosures: None reported.