Angiography is the gold standard for the diagnosis and treatment of pelvic arterial hemorrhage associated with pelvic fractures. In most cases, a single angiogram with embolization is adequate to control pelvic arterial hemorrhage. However, a small subset of patients, require repeat pelvic angiogram to evaluate and treat recurrent hemorrhage. This study seeks to define this population and determine clinical predictors of recurrent hemorrhage.Methods:
We conducted a retrospective case control study comparing patients with traumatic pelvic fracture undergoing repeated pelvic angiogram versus a single angiogram between the years 1995 and 2000. Stepwise logistic regression was used to identify the independent predictors of recurrent hemorrhage.Results:
In the years studied, 556 patients underwent a pelvic angiogram to evaluate for pelvic arterial hemorrhage associated with pelvic fractures. Among these, 42 (7.5%) patients underwent a second angiogram for suspected recurrent hemorrhage. In comparison to the initial angiogram, the source of bleeding on the repeat angiogram occurred at a new bleeding site in 68%, at a previously embolized site in 18%, and both in 14%. Significant risk factors for recurrent pelvic arterial hemorrhage included hypotension or transfusion of >2 U of blood per hour before the initial angiogram, pubic symphysis widening, and more than two injured arteries requiring embolization (p < 0.05). Of these, more than two injured arteries requiring embolization (odds ratio, 16.0; 95% confidence interval, 2.9–88) and transfusion of >2 U of blood per hour (odds ratio, 6.9; 95% confidence interval,1.9–25) were independent predictors of recurrent hemorrhage.Conclusion:
Angiographic control of traumatic pelvic arterial hemorrhage is highly successful. However, recurrent pelvic arterial hemorrhage does occur. We identified a subgroup of patients with pelvic fractures who are at increased risk of recurrent pelvic arterial hemorrhage and should be considered for early repeat angiography for signs of ongoing hemorrhage.