Initial nonoperative management of splenic injury is standard in hemodynamically stable patients. Over the past decade, the success rates have climbed by adding angioembolization to the algorithm, according to several published series. However, the objective benefit and specific indications have not been worked out.
A paper published this month by the University of Florida, Jacksonville used the NTRACS registry to try to clarify these issues. They identified 1039 patients undergoing nonoperative management (NOM) over a nearly 10 year period. Patients who died shortly after arrival, those who went directly to OR for hemodynamic reasons, and children were excluded, leaving 539 patients. Only about 1/6 of the patients underwent embolization.
The overall failure rate was about 4%, a little higher in the non-angio patients, a little lower with angio. Incidentally, the angio group had significantly higher injury severity (26 vs 20). Analysis of the lower grade spleen injury group showed no improvement in success rate by adding angio. However, the high grade groups (grades IV-V) did benefit by adding this procedure. Similarly, success improved when performing angio in patients with contrast blush or evidence of slow, ongoing bleeding. If NOM did fail, it usually occurred on day 2.
Bottom line: Although we’ve been adding angio to non-operative management of spleen and liver injury for a decade, here’s the first paper that has been able to define the real indications for doing it. First, all unstable patients go to the OR (don’t even consider nonop management). In the remaining patients, if the CT shows a grade IV or V injury, or a contrast blush, angio is recommended. If neither of these is noted, but the hemoglobin continues to decline “too quickly” (surgeon judgement), then a trip to angio is also warranted. Applying these principles can increase your success rate to about 96%.
Reference: Selective angiographic embolization of blunt splenic traumatic injuries in adults decreases failure rate of nonoperative management. J Trauma 72(5):1127-1134, 2012.