Limitations of computed tomography in the evaluation of acute abdominal trauma: a prospective comparison with diagnostic peritoneal lavage

Abstract

There has been recent enthusiasm for computed tomography (CT) to supplant diagnostic peritoneal lavage (DPL) in the detection of abdominal injuries. We prospectively compared CT to DPL following acute blunt trauma or stab wound to the abdomen. Patients with hemodynamic instability or overt signs of intraperitoneal pathology underwent urgent laparotomy and were excluded from study. Those with indications for DPL had lavage catheter insertion via open technique and attempted aspiration for gross blood. This was followed by contrast CT of the abdomen with a Technicare 2010 scanner. Lavage fluid, when required, was then instilled, recovered, and analyzed. CT interpretations were made in a blind fashion by a single staff radiologist. Decision for laparotomy was based on clinical, DPL, and CT data. In blunt trauma (N = 65), DPL detected 5/5 (100%) injuries discovered at laparotomy and CT 2/5 (40%). Following stab wounds (N = 35), DPL was true positive in 7/7 (100%) and CT in 1/7 (14.3%), with one false positive CT leading to negative laparotomy and one false positive DPL which prompted unnecessary celiotomy. Overall, the sensitivity of DPL was 100% versus 25% for CT and specificity 98.9% for both DPL and CT. In particular, CT missed seven solid visceral (five liver, two spleen), five hollow visceral, one major vascular, and three diaphragmatic lesions requiring operative intervention. In our experience, CT demonstrated an alarming incidence of false-negative studies. Given the widespread variability of CT equipment and personnel we would argue strongly against the use of CT alone in the evaluation of acute abdominal trauma and continue to support DPL as the most accurate and reliable instrument of detection.