Objectives: By using abdominal computed tomographic scans in the evaluation of blunt splenic trauma, we previously identified the presence of vascular blush as a predictor of failure, with a failure of nonoperative management of 13% in that series. This finding led to an alteration in our management scheme, which now includes the aggressive identification and embolization of splenic artery pseudoaneurysms.
Methods: The medical records of 524 consecutive patients with blunt splenic injury managed over a 4.5-year period were reviewed for the following information: age, Injury Severity Score (ISS), American Association for the Surgery of Trauma splenic injury grade (SIG), method and outcome of management.
Results: Of the patients, 66% were male with a mean age of 32 +/- 16, and mean ISS of 25 +/- 13. A total of 180 patients (34%) were managed with urgent operation on admission (81% splenectomy (SIG 4.0), 19% splenorrhaphy (SIG 2.6)). The remaining 344 patients (66%) were hemodynamically stable and underwent computed tomographic scan and planned nonoperative management. Of these patients, 322 patients (94%) were successfully managed nonoperatively (61% of total splenic injuries). In 26 patients (8%), a contrast blush identified on computed tomographic scan was confirmed as a parenchymal pseudoaneurysm on arteriography. Twenty patients (SIG, 2.8) were successfully embolized. In six patients, technical failure precluded embolization; all required splenectomy (SIG, 4.0). A total of 22 patients (6%) failed nonoperative management, including the six with unsuccessful embolization attempts. Sixteen patients (SIG, 3.0) who had no evidence of pseudoaneurysm were explored for a falling hematocrit, hemodynamic instability, or a worsening follow-up computed tomography: 13 patients had splenectomy, and three patients had splenorrhaphy.
Conclusions: Aggressive surveillance for and embolization of posttraumatic splenic artery pseudoaneurysms improved the rate of successful nonoperative management of blunt splenic trauma to 61%, with a nonoperative failure rate of only 6%. In comparison with our previous work, this reduction in failure of nonoperative management is a significant improvement (p < 0.03).