Pelvic fracture pattern predicts pelvic arterial haemorrhage

Background: The association between pelvic fracture pattern and the need fo r pelvic arterial embolization remains controversial. To address this issue , a study of the experience at Auckland Hospital was undertaken. Methods: Review was undertaken of a trauma database, blood bank database, p atient records and pelvic radiographs. Of the 364 pelvic fracture patients admitted over a 4-year period, 76 were transfused with 6 or more units of b lood in the first 24 h and these constitute the study population. Results: Embolized patients were older (median age 42 vs 29.5 years; P < 0. 05) and had a higher abbreviated injury score for the pelvic girdle (median 3 vs 2; P < 0.05) compared to non-embolized patients. Revised trauma score (median 7.69 vs 7.55), injury severity score (median 29 vs 30.5), morbidit y (55 vs 39%) and mortality (45 vs 32.1%) rates did not differ significantl y between embolized and non-embolized groups, respectively. The median bloo d transfusion requirement in the first 6 h from injury was 14 units in embo lized and 8 units in non-embolized patients (P = 0.005). Embolization was r equired in 12 of 27 (44.4%) patients with fracture patterns indicative of m ajor pelvic ligament disruption, whereas seven of 38 (18.4%) patients witho ut these fracture patterns required embolization (P < 0.05). Conclusions: The need for pelvic embolization correlated with fracture patt erns that indicated major ligament disruption, although the relationship wa s not sufficiently strong to warrant change to current indications for pelv ic angiography.