AT THE AAST ANNUAL MEETING
WAIKOLOA, HAWAII – Blunt trauma patients admitted in shock with anterior posterior compression III or vertical shear fracture patterns, or patients with open pelvic fracture are at greatest risk of severe bleeding requiring pelvic hemorrhage control intervention, results from a multicenter trial demonstrated.
Thirty years ago, researchers defined a classification of pelvic fracture based on a pattern of force applied to the pelvis, Todd W. Costantini, MD, said at the annual meeting of the American Association for the Surgery of Trauma. They identified three main force patterns, including lateral compression, anterior posterior compression, and vertical shear ( Radiology. 1986 Aug;160 [2]:445-51 ).
“They were able to show that certain pelvic fractures were associated with soft tissue injury and pelvic hemorrhage,” said Dr. Costantini , of the division of trauma, surgical critical care, burns and acute care surgery at the University of California, San Diego. “Since then, several single center studies have been conducted in an attempt to correlate fracture pattern with the risk of pelvic hemorrhage. A majority of these studies evaluated angiogram as the endpoint for hemorrhage control. Modern trauma care has evolved to include multiple modalities to control hemorrhage, which include pelvic external fixator placement, pelvic angiography and embolization, preperitoneal pelvic packing, and the use of the REBOA [Resuscitative Endovascular Balloon Occlusion of the Aorta] catheter as an adjunct to hemorrhage control.”
In a recently published study, Dr. Costantini and his associates found wide variability in the use of various pelvic hemorrhage control methods ( J Trauma Acute Care Surg. 2016 May;80 [5]:717-25 ). “While angioembolization alone and external fixator placement alone were the most common methods used, there were various combinations of these methods used at different times by different institutions,” he said.
These results prompted the researchers to prospectively evaluate the correlation between pelvic fracture pattern and modern care of pelvic hemorrhage control at 11 Level I trauma centers over a two year period. Inclusion criteria for the study, which was sponsored by the AAST Multi-institutional Trials Committee , were patients over the age of 18, blunt mechanism of injury, and shock on admission, which was defined as an admission systolic blood pressure of less than 90 mm Hg, or heart rate greater than 120, or base deficit greater than 5. Exclusion criteria included isolated hip fracture, pregnancy, and lack of pelvic imaging.
The researchers evaluated the pelvic fracture pattern for each patient in the study. “Each pelvic image was evaluated by a trauma surgeon, orthopedic surgeon, or radiologist and classified using the Young-Burgess Classification system,” Dr. Costantini said. Next, they used univariate and multivariate logistic regression analysis to analyze predictors for hemorrhage control intervention and mortality. The objective was to determine whether pelvic fracture pattern would predict the need for a hemorrhage control intervention.
Of the 46,716 trauma patients admitted over the two year period, 1,339 sustained a pelvic fracture. Of these, 178 met criteria for shock. The researchers excluded 15 patients due to lack of pelvic imaging, which left 163 patients in the final analysis. Their mean age was 44 years and 58% were male. On admission, their mean systolic blood pressure was 93 mm Hg, their mean heart rate was 117 beats per minute, and their median Injury Severity Score was 28. The mean hospital length of stay was 12 days and the mortality rate was 30%. The three most common mechanisms of injury were motor vehicle crash (42%), followed by pedestrian versus auto (23%), and falls (18%).
Compared with patients who did not require hemorrhage control intervention, those who did received more transfusion of packed red blood cells (13 vs. 7 units, respectively; P less than .01) and fresh frozen plasma (10 vs. 5 units; P = .01). In addition, 67% of patients with open pelvic fracture required a hemorrhage control intervention. The rate of mortality was similar between the patients who required a pelvic hemorrhage control intervention and those who did not (34% vs. 28%; P = .47).
The three most common types of pelvic fracture patterns were lateral compression I (36%) and II (23%), followed by vertical shear (13%). Patients with lateral compression I and II fractures were least likely to require hemorrhage control intervention (22% and 19%, respectively). However, on univariate analysis, patients with anterior posterior compression III fractures and those with vertical shear fractures were more likely to require a pelvic hemorrhage control intervention, compared with those who sustained other types of pelvic fractures (83% and 55%, respectively).
On multivariate analysis, the three main independent predictors of need for a hemorrhagic control intervention were anterior posterior compression III fracture (odds ratio, 109.43; P less than .001), open pelvic fracture (OR, 7.36; P = .014), and vertical shear fracture (OR, 6.99; P = .002). Pelvic fracture pattern did not predict mortality on multivariate analysis.
The invited discussant, Joseph M. Galante, MD , trauma medical director for the University of California, Davis Health System, characterized the study as important, “because it examines all forms of hemorrhage control, not just arterioembolism in the treatment of pelvic fractures,” he said. “The ability to predict who will need hemorrhage control allows for earlier mobilization to resources, both in the operating room or interventional suite and in the resuscitation bay.”
Dr. Costantini reported having no financial disclosures.
