Session I - Pelvis

Hemodynamically Unstable Pelvic Fractures: Retrospective Review of Early Embolization

Daniel N. Segina, MD; Samuel G. Agnew, MD; Tim Daniel, MD; Peter Swischuck, MD; H. Martin Northrup, MD; Robert Booth, MD; F.W. Clevenger, MD; University of Florida-Jacksonville, Jacksonville, FL

Objectives: To determine the incidence of patients requiring embolization and the correlation with their pelvic injury and vascular lesions and the associated mortality.

Design: Retrospective retrieval

Setting: Level One Trauma Center-University Hospital

Methods: Review of records of a consecutive series of patients presenting to the Trauma Center with angiographically confirmed arterial injury and who had subsequent life saving endovascular embolization prior to any other intervention. The time from injury until embolization, mode of embolization and incidence of soft tissue problems were data points sought.

Measures: The patients all sustained some type of pelvic fractures, as classified by the OTA modification of the Tile classification system. Mechanism of injury, location of artery, and mortality rate were correlated.

Results: Three hundred sixty-five consecutive patients sustaining pelvic fractures exclusively from road or industrial trauma between October 1995 and December 1999 constituted the base population. One emergency external fixation frame was applied over the study period, using the modified resuscitation protocol described. Pelvic fractures were temporized with binding of the thigh (11) or skeletal traction (4) applied on arrival, where appropriate for the injury type.

Fifty-six patients (15%) underwent embolization in the Trauma Center as a method of acute resuscitation for persistent hemodynamic instability with a negative abdominal ultrasound. Nineteen patients (33%) underwent therapeutic embolization of bilateral hypogastric system injuries.

Mechanism of injury: motor vehicle collision (22), auto vs. pedestrian (18), crush (7), ejection (6), MCC (4) The average time from injury until embolization completion was 3.3 hr. (1.5-5.2). Multiple named vessel injury occurred in 30 patients (52.6%). Gelfoam and coil combination was utilized in 50/57 procedures. Forty patients (76%, 22 male and 18 female) survived the initial trauma and resuscitation. The nonsurvivors died from non-hemorrhagic sources in all cases: brain injury, multi-system organ failure, and sepsis. All 41patients sustained significant multi-system trauma with an average ISS score of 42 (19-66).

The skeletal profile of pelvic fractures undergoing resuscitative embolization: rotationally unstable OTA B type (24), rotationally and vertically unstable OTA C type (28), and four patients with mechanically stable A type injuries. Those requiring bilateral systems or vessel embolization: Title OTA C type (17) 40%, and B type (5) 8%, and A type (1) 2%.

Mortality data:

OTA Fracture Type	N	Age (range)	ISS	Mortality
A	4	40 (18-81)	24	1 (25%)
B1	2	30 (13-46)	33	0
B2	8	33 (15-66)	33	2 (25%)
B3	14	34 (2-73)	32	4 (28%)
C1	4	38 (22-50)	36	2 (50%)
C2	4	27 (5-55)	24	0
C3	20	35 (18-56)	48	11 (55%)

Discussion: Emergent embolization has been employed routinely for 4 years at our institution prior to any other intervention in hemodynamically unstable patients with a negative abdominal ultrasound. Fifty-two consecutive patients presented with mechanically unstable pelvic injury and remained hemodynamically unstable; the use of emergency external fixation was rare (1). 75% of patients presenting with combination C3 pelvic injuries and bilateral arterial injuries sustained an unsurvivable amount of trauma. Wound problems developed following subsequent hemipelvis operative fixation in only one patient with massive degloving, despite the fact that 33% of the patient cohort sustained a traumatic loss of hypogastric systems bilaterally and 52% had multiple arterial injuries and concomitant massive truncal trauma.

Conclusion: The placement of angiography and embolization in the decision algorithm for the hemodynamically unstable pelvic fracture patient is typically done in the later stages of the decision scheme. These data suggest that embolization in conjunction with binding the thighs or skeletal traction may facilitate the resuscitation process and preclude emergent frame application as well. No evidence of perineal dysvascular changes was detected.