Accuracy of CT Scan for Detecting Hollow Viscus Injury in Penetrating Abdominal Trauma

BACKGROUND: In penetrating abdominal trauma, computed tomography (CT) is routinely performed to evaluate stable patients for selective non-operative management (SNOM). Triple-contrast CT (oral, rectal, and IV) has traditionally been used. However, due to its disadvantages, most trauma centres, including our unit at the Charlotte Maxeke Johannesburg Academic Hospital (CMJAH), now perform single-contrast intravenous-only CT scans. We performed a retrospective review to determine the accuracy of single-contrast CT scans for detecting hollow viscus injuries (HVI) in penetrating abdominal trauma. METHODS: A retrospective review of all patients who presented to CMJAH with penetrating abdominal injuries was performed between 01 August 2017 and 31 August 2019 and were evaluated for SNOM with CT (IV contrast only). Patient records were reviewed to determine pertinent demographics, mechanism, and site of injury, as well as metabolic parameters. CT findings were compared to findings at laparotomy. RESULTS: A total of 437 patients met the inclusion criteria. The majority were male (92.7%), with a mean age of 31.5 yrs (SD 8.7). Injuries were predominantly due to stab wounds (72,5%, n = 317). CT scan was negative in 342 patients, of which 314 completed SNOM successfully. A total of 93 patients proceeded to laparotomy. CT had a sensitivity of 95.1%, specificity of 44.2%, positive predictive value of 57.4%, and negative predictive value of 92%. CONCLUSION: Single-contrast CT in penetrating abdominal trauma is a valuable investigative tool in identifying patients for SNOM. Features of HVI on single-contrast CT are not very specific and should be interpreted along with other clinical factors including wound trajectory and serial abdominal examinations. Other associated injuries such as diaphragmatic and solid organ injuries should be considered in the final management plan.