A simple CT score to quantify pelvic and retroperitoneal hematoma associated with pelvic fractures predicts transfusion needs, pelvic hemostatic procedures, and outcome.

ABSTRACT

BACKGROUND: Pelvic and retroperitoneal hematoma (PRH) in case of pelvic fracture may lead to early hemorrhagic shock. Quantifying PRH remains challenging in clinical practice. The goal of this study was to determine the statistical association between a semi-quantitative scoring system for PRH assessed with computed tomography (CT) and transfusion needs, pelvic hemostatic procedures, and outcome. METHODS: All consecutive severe trauma patients with pelvic ring fracture between 2010 and 2015 were included in this retrospective study. PRH was quantified using semi-quantitative analysis on admission CT scan. The pelvis and retroperitoneal cavity was assessed as 10 compartments. Hematoma was counted as 0 (absent), 1 (minimal or moderate), and 2 (large or bilateral) for each compartment (maximum score of 20). The patients were divided into the following 3 groups no or minimal PRH (score 0-5), moderate PRH (6-9), and large PRH (10-20). These groups were compared in terms of initial transfusion needs, massive transfusion, hemostatic procedures, and outcome. Logistic regression and receiver operating characteristic (ROC) curves were analyzed. RESULTS: The study included 311 patients with pelvic fracture (mean age 41.9, [SD] 19.9 years; mean ISS 27.4, [SD] 19.4; unstable fractures, 32%; ≥5 units of packed red blood cells, 37%; massive transfusion, 19%; multiple organ failure, 29%; mortality, 13%), divided into no or minimal PRH group (128 (22%)), moderate PRH group (115 (37%)), and large PRH group (68 (22%)). Increasing PRH was found to increase transfusion needs and massive transfusions, with a higher number of pelvic hemostatic procedures, multiple organ failures, increasing need for mechanical ventilation, and prolonged hospitalization; mortality was also increased. These significant statistical associations were confirmed by logistic regression models (odds ratio, 1.2-12.1 for moderate PRH, 3.1-30.2 for large PRH) and ROC curve analysis (area under the ROC curve, 0.59-0.76). CONCLUSION: Semi-quantitative assessment of PRH on admission CT scan allows to predict transfusion needs, hemostatic procedures, and worse outcome of severe trauma patients with pelvic fracture.