CT has become an integral part of the evaluation of patients with blunt abdominal trauma. CT has been shown to be accurate in defining solid and hollow visceral injuries as well as associated hemoperitoneum [1,2,3].

With the increasing use of contrast-enhanced CT in the emergency setting, a number of authors have reported their experiences using CT to detect active abdominal hemorrhage in patients with blunt abdominal trauma [4,5,6,7,8,9,10,11,12,13]. On CT, active hemorrhage originating from various organs including liver, spleen, pancreas, kidneys, bowel, mesentery, and abdominal soft tissues can be detected [5]. Identification of active hemorrhage is of paramount importance because this finding may indicate a life-threatening hemorrhage and has implications for proper emergency treatment [7, 9, 10, 12, 14, 15].

Active extravasation of contrast material is considered a rare finding on conventional CT images of patients with blunt abdominal trauma [4,5,6, 8, 15]. The prevalence of active hemorhage was only 0.2% in a study conducted by Taylor et al. [6] in which they reviewed conventional CT images of 3000 patients who had sustained trauma. Detection of active hemorrhage is reported to occur more frequently when patients with blunt abdominal trauma are scanned with single-detector helical CT— in particular, when blunt splenic or hepatic injuries are present [7, 10, 12, 13].

Since April 2000, at our level I trauma center, multidetector CT has been used to examine patients with blunt abdominal trauma. Our experience with multidetector CT in the emergency trauma area suggests that a considerable number of active hemorrhages are detected in these patients.

The purpose of this retrospective study was to determine the imaging findings and the prevalence of active hemorrhage on contrast-enhanced multidetector CT in patients who sustained blunt abdominal trauma.

CT has been proven to be an excellent imaging modality for evaluation of hemodynamically stable patients with blunt abdominal trauma. The rapid diagnostic capability afforded by CT has contributed to a decrease in morbidity and mortality from traumatic abdominal injuries [1].

Active hemorrhage can be diagnosed on CT on the basis of increased radiodensity compared with surrounding tissue, which results from the extravasation of intravascular contrast agent [4,5,6,7,8,9,10,11,12,13]. The exact bleeding rate required for this finding is unknown, but extravasation of IV-administered contrast agent in a patient with blunt abdominal trauma represents a significant finding that may require immediate surgical or interventional therapy [7, 9, 10, 12, 14, 15].

Active extravasation of contrast material during CT in patients with blunt hepatic trauma is reported to be an important sign in predicting the necessity of surgery or angiography. In a series of 150 hemodynamically stable patients with blunt hepatic injury who were evaluated with contrast-enhanced CT, Fang et al. [11] showed that extravasation of contrast material in the peritoneal cavity or in the ruptured hepatic parenchyma indicates active hemorrhage. Emergency surgery was necessary in six (75%) of eight patients from this series to control the bleeding site of the liver.

On the basis of a retrospective evaluation of CT scans of 150 patients with traumatic splenic injuries, Federle et al. [7] concluded that the absence of active bleeding on CT scans may help predict successful nonsurgical management of splenic injuries. Conversely, 26 (93%) of 28 patients with splenic injury and active extravasation of blood as evidenced on CT underwent immediate surgical intervention [7]. Gavant et al. [12] reported that 38 (83%) of 46 patients with vascular abnormalities on initial CT scans after blunt splenic injury required immediate or delayed surgical repair of the splenic injury.

The common belief that detection of active hemorrhage on CT is an important prognostic factor necessitating immediate surgery or interventional angiographic therapy is also supported by the results of our study. We found that detection of active hemorrhage on multidetector CT was followed by immediate surgical and angiographic intervention in 16 (73%) of 22 patients. One patient underwent angiographic intervention and successful embolization of a renal pseudoaneurysm hours after multidetector CT. Because of their overall clinical situations, the remaining five patients with active hemorrhage on multidetector CT were treated conservatively by trauma surgeons. All five of these patients had severe closed head injuries and died from multiple trauma within 3 hr after undergoing multidetector CT.

The appearance of active hemorrhage varies depending on the bleeding rate and the applied CT technique [7]. IV administration of contrast material using an automated power injector allows rapid administration of contrast material, which is of paramount importance [7]. However, the optimal scanning protocol may vary depending on the type of CT scanner and the suspected organ injury. An imaging protocol for scanning patients with blunt abdominal trauma using multidetector CT has, to our knowledge, not yet been established. The multidetector CT protocol used for our study, including the dose of the contrast agent, the scanning delay, slice width, and reconstruction interval, is mostly based on our experiences after initial tests with multidetector CT. We chose a scanning delay of 85 sec after contrast material administration because the imaging protocol for single-detector helical CT in the emergency department recommends a scanning delay of 70-90 sec for sufficient visceral enhancement [2, 7].

In contrast to other authors [1, 2], we consider the administration of an oral contrast agent not to be useful in patients presenting with blunt abdominal trauma to the emergency department because the extravasated contrast agent in a patient with active hemorrhage may not be detected when a high concentration of contrast material is present within the intestine. In addition, oral application of contrast material—even via a gastric tube—is another time-consuming procedure that we believe is impractical in the acute clinical setting of patients presenting to the emergency department.

Since the first report of active hemorrhage detection using CT [4], a few studies have addressed the imaging findings of contrast-enhanced CT for the detection of active hemorrhage in patients with blunt abdominal trauma [5, 6, 8,9,10]. On conventional or single-detector helical CT, active bleeding was identified when a focal or diffuse high-attenuation collection of contrast material surrounded by hematoma or damaged parenchyma was noted on CT scans. Other CT findings of active hemorrhage include the presence of high-density contrast-enhanced blood mixed with hemoperitoneum and a clot [6, 8].

The presence of a jet of extravasated contrast material as an indicator of active bleedling is infrequently noted using conventional or single-detector helical CT [6, 10]. In a series conducted by Taylor et al. [6] describing conventional CT findings in seven patients with active hemorrhage from trauma, the presence of a jet of contrast material as an indicator of active bleeding was noted in only one patient. In contrast to the findings of these previous studies, we found that active hemorrhage presented most frequently as a jet of extravasated contrast material on multidetector CT scans. We believe that this phenomenon is most likely caused by the fast data acquisition capability associated with multidetector CT. Combined with the use of IV-administered contrast material, the faster scanning speed of multidetector CT and its ability to use a narrow collimation increase opacification of contrast material in the mesenteric, retroperitoneal, and portal vasculature [19,20,21]. The improved spatial resolution provided by multidetector CT enables a more detailed assessment of the parenchymal organs and abdominal vasculature than is possible with conventional and single-detector helical CT [22]. This higher spacial resolution probably results in a higher detectability of a jet of extravasated contrast material when vascular integrity has been disrupted.

In this study, the CT attenuation values measured in the aorta or adjacent major arteries were significantly higher than the CT attenuation values measured in the extravasated hemorrhage. This finding contradicts previous findings of similar measurements performed in active hemorrhages detected on conventional or single-detector helical CT [7, 9]. On conventional or single-detector helical CT, attenuation values of extravasated blood have been reported to exceed those of blood vessels. The difference between these measurements performed on conventional or single-slice helical CT and those performed on multidetector CT may be caused by the faster z-axis speed of multidetector CT that allows the entire abdominal vasculature to be scanned in several seconds. This faster scanning time results in high intravascular contrast even in the presence of ongoing continuous dilution of the extravasating contrast material.

Our data show a broad range of CT attenuation values for the abdominal aorta or adjacent major arteries and the extravasated contrast material. Because all patients were examined with the same imaging protocol and because the delay time was standardized, this range in attenuation values may result from variations in hemodynamic stability or cardiac output of the patient population. In addition, because CT attenuation values were obtained at different locations of active hemorrhage (at a distance ranging from the liver to the internal iliac artery), the measurements were likely obtained at different time points during the multidetector CT data acquisition.

The reported prevalence of active hemorrhage detected on CT in patients with blunt abdominal trauma is variable. The variability of the reported prevalence may depend on several factors including severity of injury, the criterion used for hemodynamic stability in patients undergoing CT evaluation after trauma, and the scanning technique used. In a retrospective review of conventional CT records during a 44-month period, Jeffrey et al. [8] identified 18 patients with a CT diagnosis of active bleeding resulting from various causes, including 12 patients with blunt abdominal trauma. Shanmuganathan et al. [9] diagnosed active hemorrhage in 26 patients after blunt abdominal trauma during a 44-month period. Unfortunately, in both of these studies the prevalence of active bleeding cannot be calculated because the total number of patients with blunt abdominal trauma during the study period was not reported.

In their retrospective review of 3000 patients, Taylor et al. [6] found that active hemorrhage was present on conventional CT scans in seven children, which resulted in a calculated prevalence of 0.2% for active hemorrhage. With single-detector helical CT, the prevalence of active hemorrhage is reported to be substantially higher. Federle et al. [7] reviewed the single-detector helical CT images of 150 patients with blunt splenic injury over a 65-month period. The authors detected active hemorrhage in 28 patients (19%). Gavant et al. [12] found vascular abnormalities, including posttraumatic splenic pseudoaneurysm and posttraumatic splenic hemorrhage, on the initial CT scans, collected over a 5-year period, of 46 (25%) of 181 patients. Shanmuganathan et al. [13] reported intraparenchymal, subcapsular, or intraperitoneal splenic vascular contrast material extravasation in seven (9%) of 78 patients who had sustained blunt splenic injury over a 15-month period. During a 42-month period, pooling of IV-administered contrast material as a marker of active bleeding [10] was identified in 15 (7%) of 212 patients with blunt hepatic injury in a study by Fang et al. [10].

In our study using multidetector CT, the overall prevalence of active hemorrhage (including all bleeding sites) was 13%. In patients with blunt splenic or hepatic injury, the prevalence of active bleeding was 14% and 9%, respectively. With the use of multidetector CT in an emergency setting, we collected data that confirm the findings reported for single-detector helical CT; active hemorrhage may no longer be a sporadic finding in the emergency setting, as has been reported for conventional CT. However, because we did not perform a paired comparison of multidetector CT versus conventional CT in our study, we are not able to determine the statistical difference between the techniques with regard to depiction of active bleeding after blunt abdominal trauma.

At our hospital, only patients who require 2 L or less of fluid to achieve hemodynamic stability qualify for abdominal multidetector CT. The prevalence of active hemorrhage may be even higher in a series of patients with a less strict definition of hemodynamic stability.

Similar to the variability in the reported prevalence of active hemorrhage on CT scans, the prevalence of organ injuries in patients with blunt abdominal trauma is variable. Singh et al. [23] reported injuries of liver, spleen, or kidney in 23 (32%) of 73 patients with blunt abdominal trauma. Udekwu et al. [24] identified 64 patients (24%) with spleen, liver, or kidney injuries in a series of 265 patients. In contrast, in a large series of 11,188 patients who sustained blunt trauma during a 33-month period, traumatic intraabdominal injuries were identified in only 4% [25]. In our study, parenchymal injuries confined to the liver, spleen, or kidney were identified in 58 (35%) of 165 patients. The relatively high number of organ injuries reported in our study is probably explained by the fact that most of our patients sustained blunt abdominal trauma combined with multiple injuries, thus reflecting the severity of the trauma.

We acknowledge the following limitations to our study. The retrospective nature of this study and the small number of patients with active contrast extravasation may limit the findings. We did not evaluate the sensitivity of contrast-enhanced multidetector CT for the detection of active hemorrhage. However, determining the diagnostic performance of multidetector CT in the detection of active hemorrhage is crucial because a possible inclusion bias will always influence the results. In addition, with the protocol used in this study, including a fixed delay of 85 sec, differentiation between arterial and major venous hemorrhage may be difficult. Furthermore, patients were not evaluated on unenhanced multidetector CT before the IV administration of contrast material. The lack of unenhanced studies may limit the ability to distinguish hematoma from active bleeding. However, by measuring the CT attenuation values on contrast-enhanced multidetector CT scans, radiologists can differentiate active hemorrhage from surrounding hematoma because the CT attenuation values of active bleeding are significantly higher than those of surrounding hematoma. Finally, our results have not been evaluated using interobserver agreement.

In conclusion, our study has shown that active hemorrhage in patients who have sustained blunt abdominal trauma is most frequently visible as a jet of extravasated contrast agent using contrast-enhanced multidetector CT. When extravasation is detected, immediate surgical or angiographic therapy is required.

Address correspondence to D. Weishaupt.