- Email: [email protected]
Arterial embolization for pelvic fractures after blunt trauma: are we all talk?
The American Journal of Surgery (2010) 200, 752–758
Arterial embolization for pelvic fractures after blunt trauma: are we all talk? Todd W. Costantini, M.D., Patrick L. Bosarge, M.D., Dale Fortlage, B.A., Vishal Bansal, M.D., Raul Coimbra, M.D., Ph.D.* Division of Trauma, Surgical Critical Care, and Burns, Department of Surgery, University of California, San Diego School of Medicine, 200 W. Arbor Dr, 8896, San Diego, CA 92103-8896, USA KEYWORDS: Angiography; Angioembolization; Fracture patterns; Pelvic injury mechanism
Abstract BACKGROUND: We hypothesized that arterial embolization for bleeding after pelvic fracture is used relatively infrequently. We sought to identify the true need for arterial embolization and define injury patterns associated with successful therapeutic angiographic embolization. METHODS: A retrospective review identified patients admitted to our urban, Level 1 trauma center with pelvic fractures from 2001 to 2009. Patients requiring pelvic arterial angiogram and embolization of pelvic bleeding were reviewed for pelvic fracture pattern and pelvic injury mechanism. RESULTS: There were 819 patients diagnosed with pelvic fractures, with only 31 patients (3.8%) undergoing diagnostic pelvic angiography. Of those, 18 patients (58.1%) had active bleeding requiring arterial embolization. Complex pelvic fracture patterns were common in patients undergoing angiogram. Patients undergoing pelvic angiography with an anteroposterior compression mechanism were more likely to have negative findings on angiogram. CONCLUSIONS: The actual need for angiography and therapeutic embolization is quite small in patients sustaining pelvic fracture. Although factors associated with the need for pelvic angiography frequently are debated, we may discuss angiography for pelvic fractures more often than it actually is performed. © 2010 Elsevier Inc. All rights reserved.
Traumatic pelvic fractures can result in significant hemorrhage that can be associated with significant morbidity and mortality. Patients with pelvic fractures that cause hemodynamic instability have a mortality rate that exceeds 50% in several series.1–3 Therefore, identifying patients at risk for severe bleeding because of their pelvic fracture and prompt mobilization of resources to treat those patients is paramount. Presented as an oral presentation at the 2010 annual meeting of the Southwest Surgical Congress, March 21 to 24, 2010, Tucson, AZ. * Corresponding author. Tel.: ⫹1-619-543-7100; fax: ⫹1-619-5437202. E-mail address: [email protected] Manuscript received March 5, 2010; revised manuscript June 15, 2010
0002-9610/$ - see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2010.06.006
Bleeding from pelvic fractures can be generated from several sources including arterial and venous injury, and bleeding from fractured cancellous bone within the pelvis. Bleeding from fractured bone within the pelvis can be controlled with prompt stabilization of the fracture, which also can tamponade and control venous bleeding.4 Arterial injury caused by pelvic fracture may require therapeutic arterial embolization, which has been shown to be an important adjunct in the treatment of patients with pelvic hemorrhage.5,6 Therefore, management of pelvic fractures may require care from a multidisciplinary team of trauma surgeons, orthopedic surgeons, and interventional radiologists. Although the risk of significant hemorrhage after pelvic fracture is clear, the incidence and predictors for needing
T.W. Costantini et al.
Arterial embolization for pelvic fractures
therapeutic arterial embolization remain in debate. Criteria exist for obtaining pelvic arterial angiography in patients with severe pelvic fracture based on the presence of hemodynamic instability or the need for ongoing blood transfusion. Clinicians must decide between obtaining further diagnostic testing such as computerized tomography (CT) scan or proceed immediately to angiography. More liberal use of angiography in pelvic fractures based exclusively on hemodynamics or on the amount of blood transfused, in an era of multislice CT scanner with high-definition images and a high likelihood of identifying contrast extravasation in the arterial phase of the examination, may not be the best strategy for patient care. Obviously, the risks of multiple contrast loads and radiation exposure should be taken into account when deciding which therapeutic strategy will be used. Therefore, early prediction of the need for angiography is paramount. In this study, we hypothesized that arterial embolization for bleeding after pelvic fracture is used relatively infrequently, and sought to identify the incidence and true need for arterial embolization. We also sought to define injury patterns associated with successful therapeutic angiographic embolization in an attempt to provide information to clinicians regarding resource mobilization.
753
agnostic pelvic angiogram with possible embolization is attempted. If FAST or DPL are positive, an immediate exploratory laparotomy is performed. If a large or expanding pelvic/retroperitoneal hematoma is found, an immediate postoperative angiogram and possible embolization is performed. Data obtained included demographics, mechanism of injury, injury severity score (ISS), admission vital signs, admission Glasgow Coma Scale, Abbreviated Injury Score (AIS), hospital and intensive care unit (ICU) length of stay (LOS), and mortality. Autopsy records were reviewed for those patients who died after a pelvic fracture requiring angiography. Patients requiring arterial embolization were stratified further for pelvic fracture pattern and pelvic injury mechanism. A senior trauma surgeon blinded to the study groups reviewed all radiographic studies to determine pelvic fracture patterns and pelvic injury mechanism (anteroposterior compression, lateral compression of vertical shear). Patients with multiple pelvic fracture patterns were categorized as having a complex mechanism of injury. Values are expressed as the mean ⫾ standard deviation. Statistical analysis was performed using SPSS Statistics software version 11.5 (SPSS, Chicago, IL). The Student t test, Mann–Whitney U test, and chi-square test were used to determine statistical significance where appropriate. Statistical significance was defined as a P value of less than .05.
Methods A retrospective review identified all trauma patients admitted with pelvic fractures to the University of California, San Diego Medical Center, an urban, Level 1 trauma center between January 1, 2000, and January 1, 2009. Patient data were obtained using our trauma registry database. Medical records and radiographic studies were used to supplement our data collection. Patients requiring pelvic arterial angiogram and arterial embolization of pelvic bleeding were identified. This study received approval from the University of California, San Diego Institutional Review Board. Our institutional protocol since 2000 has been to use angiography, when indicated, early after potentially severe pelvic injury. However, we have relied on CT extravasation as an absolute indication to perform arterial angiography. Our protocol is as follows: hemodynamically stable patients are screened for intra-abdominal injury with focused assessment with sonography for trauma (FAST). If FAST is positive, a contrast CT of the abdomen is obtained. If positive for contrast extravasation (blush) in the pelvis, an angiogram and possible embolization is performed. If FAST is negative and the patient shows a decreased level of hematocrit over time requiring 4 U of blood transfusion within 8 hours, without other possible sources of bleeding, a diagnostic angiography and possible embolization is attempted. Hemodynamically unstable patients undergo FAST or diagnostic peritoneal lavage (DPL) to rule out intra-abdominal hemorrhage. If FAST or DPL are negative, other sources of bleeding have been ruled out, and the patient remains unstable (with a decreasing hematocrit) then a di-
Results Study population During the 9-year study period, there were 819 patients diagnosed with pelvic fracture. Characteristics of this population of patients with pelvic fractures are presented in Table 1. Patients sustaining pelvic fractures were predominantly young with a mean age of 41.4 ⫾ 19.6 years and
Table 1
Demographics of the study population All pelvic fractures
n Age, y Male, n (%) Admission vitals Systolic blood pressure, mm Hg Heart rate Respiratory rate Admission base deficit Admission GCS ISS ICU LOS, d Hospital LOS, d Mortality, n (%) GCS ⫽ Glasgow Coma Scale.
819 41.4 ⫾ 19.6 488 (59.6) 124.5 ⫾ 34.6 94.6 ⫾ 25.8 17.4 ⫾ 7.9 ⫺2.8 ⫾ 4.9 13.0 ⫾ 4.0 19.6 ⫾ 13.2 4.1 ⫾ 8.8 11.9 ⫾ 16.0 64 (7.8)
754
The American Journal of Surgery, Vol 200, No 6, December 2010
Table 2
Characteristics of pelvic fracture patients based on the need for arterial angiography (n ⫽ 819 patients)
n, (%) Age, y Male, n (%) Admission vitals Systolic blood pressure, mm Hg Heart rate Respiratory rate Admission base deficit Admission GCS ISS ICU LOS, d Hospital LOS, d Mortality, n (%)
Pelvic fracture, no angiography
Pelvic fracture, all angiography
788 (96.2) 41.2 ⫾ 19.6 469 (59.5)
31 (3.8) 45.9 ⫾ 19.8 19 (61.3)
.1909 1.000
125.1 ⫾ 31.5 94.0 ⫾ 25.7 17.3 ⫾ 7.8 ⫺2.7 ⫾ 4.7 13.0 ⫾ 3.9 19.2 ⫾ 13.0 3.9 ⫾ 8.5 11.5 ⫾ 15.2 58 (7.4)
109.9 ⫾ 32.5 109.3 ⫾ 24.8 17.6 ⫾ 9.7 ⫺4.1 ⫾ 7.0 12.5 ⫾ 4.2 30.5 ⫾ 15.1 10.5 ⫾ 14.1 22.2 ⫾ 28.7 7 (22.6)
.0086* .0012* 1.000 .1119 .4853 ⬍.0001* ⬍.0001* .0003* .0082*
P value
GCS ⫽ Glasgow Coma Scale. *P ⬍ .05 using the Student t test or chi-square test where appropriate.
male (59.6%). The mean ISS was 19.6 ⫾ 13.2, with an overall mortality rate of 7.8% in this population of patients.
Patients requiring arterial angiogram Of the 819 patients sustaining pelvic fractures during the study period, only 31 patients (3.8%) underwent pelvic angiography. Clinical characteristics of these patients are compared in Table 2. Only 1 patient did not receive a CT scan before being taken to the angiography suite. This patient, who was hemodynamically unstable, was admitted directly to the operating room as part of our operating room resuscitation protocol. After placement of an external pelvic fixator by the orthopedic team, the patient was taken directly to the angiography suite for angioembolization of his left internal pudendal artery. Patients requiring pelvic angiography had significantly lower systolic blood pressure and increased heart rate on admission. Patients with pelvic fractures who underwent pelvic angiography also were injured more severely with an average ISS of 30.5 ⫾ 15.1, compared with an ISS of 19.2 ⫾ 13.0 in the group of pelvic fractures that did not require angiography (P ⬍ .0001). This was associated with an increased hospital and ICU LOS for patients requiring pelvic angiography. Mortality was increased in the group of patients with pelvic fracture who underwent pelvic angiography (22.6% vs 7.4%; P ⬍ .01). Autopsy reports were reviewed for patients who underwent pelvic angiography and died during that hospitalization. Of the 6 patients who underwent angiography and subsequently died, 2 died because of severe pelvic crush injuries after being struck by an automobile. The remainder of the mortalities was due to multiple blunt force injuries.
Patients with positive pelvic angiography Characteristics of patients with pelvic fracture who underwent pelvic angiography are displayed in Table 3. Of the
31 patients who underwent pelvic angiography after pelvic fracture, 18 patients (58.1%) had positive angiogram findings of active arterial extravasation. Patients with a positive finding on pelvic angiogram were more likely to have an abdominal AIS of 3 or greater. Thirteen of the 31 patients (41.9%) undergoing pelvic angiography had a negative pelvic angiogram; however, therapeutic angioembolization was performed in 4 of these patients because of concern for hemorrhage in areas of fracture or hematoma.
Pelvic fracture pattern in patients undergoing pelvic angiography Pelvic fracture pattern and pelvic injury mechanism for patients who underwent pelvic angiography are presented in Table 4. Pubic rami fracture was the most common pelvic fracture seen in patients undergoing angiography, regardless of angiographic findings. Of patients with positive findings on angiogram, sacral fracture and sacroiliac joint disruption were common pelvic fracture patterns. Of patients who underwent pelvic angiography, those with a mechanism of only anteroposterior compression were more likely to have negative angiographic findings. Several patients had complex mechanisms of injury, with the presence of more than 1 pelvic injury mechanism (most commonly a combination of lateral compression and anteroposterior compression). A pure vertical shear mechanism was the pelvic injury mechanism seen least frequently in patients undergoing pelvic angiography.
Comments Pelvic fractures after blunt trauma are common and may represent a wide spectrum of severity. Significant pelvic fractures require a substantial amount of force and often are
T.W. Costantini et al. Table 3
Arterial embolization for pelvic fractures
755
Pelvic fracture patients undergoing arterial angiography (n ⫽ 31 patients) Pelvic fracture, positive angiography (n ⫽ 18)
Age, y Male, n (%) Admission vitals Systolic blood pressure Heart rate Respiratory rate Base deficit Admission GCS ISS Head AIS ⱖ 3 Chest AIS ⱖ 3 Abdominal AIS ⱖ 3 Extremity AIS ⱖ 3 CT scan obtained Blush on CT scan Angiographic intervention ICU LOS, d Hospital LOS, d Mortality, n (%)
Pelvic fracture, negative angiography (n ⫽ 13)
P value
47.6 ⫾ 20.6 13 (72.2)
43.5 ⫾ 19.2 6 (46.2)
.5616 .2623
104.7 ⫾ 29.5 112.3 ⫾ 24.6 19.0 ⫾ 9.7 ⫺5.0 ⫾ 4.3 12.2 ⫾ 4.2 34.5 ⫾ 15.2 9 (50.0%) 9 (50.0%) 15 (83.3%) 18 (100%) 17 (94.4%) 4 (22.2%) 18 (100%) 7.0 ⫾ 7.0 15.0 ⫾ 15.0 6 (33.3)
116.6 ⫾ 36.1 105.2 ⫾ 25.5 15.7 ⫾ 9.8 ⫺3.1 ⫾ 9.4 12.8 ⫾ 4.4 24.9 ⫾ 13.6 2 (15.4%) 7 (53.8%) 5 (38.5%) 11 (84.6%) 13 (100%) 6 (46.2%) 4 (30.8%) 15.4 ⫾ 19.5 32.3 ⫾ 39.5 1 (7.7)
.5864 .4233 .4469 .4544 .7032 .1093 .0656 1.000 .0209* .1677 1.000 .2469 ⬍.0001* .1014 .0990 .1912
GCS ⫽ Glasgow Coma Scale. *P ⬍ .05 using the Mann–Whitney U test or chi-square test where appropriate.
seen in conjunction with other serious injuries. Therefore, early decision making in determining and stabilizing the source of hemorrhage in the severely injured patient with pelvic fracture may significantly affect outcome. Multiple methods have been described to stabilize patients with significant hemorrhage from pelvic fracture, including pelvic fixation devices, external pelvic fixation, direct retroperitoneal packing, transabdominal packing, and arterial angiography.7–10 Because several of these methods require a multidisciplinary team, identifying factors that successfully
Table 4
predict successful management of severe pelvic injuries may be important for resource mobilization and use. Angiography and angiographic embolization have been used for decades to diagnose and treat patients with arterial injury caused by pelvic fracture.11 Common indications for angiography after pelvic trauma include hemodynamic instability, extravasation of contrast seen from pelvic vessels on CT scan, large pelvic hematoma on CT, or ongoing need for blood transfusion as a result of pelvic bleeding. Other life-threatening associated injuries are common in this pop-
Pelvic fracture patterns and pelvic injury mechanisms for patients requiring angiography (n ⫽ 31 patients)
Anterior pelvic injury Pubic rami fracture Widened pubic symphysis Open book injury Lateral pelvic injury Acetabular fracture Impacted sacrum Iliac wing fracture Posterior pelvic injury Sacral fracture Sacroiliac joint disruption Pelvic injury mechanism Lateral compression Anteroposterior compression Vertical shear Complex mechanism
Pelvic fracture, positive angiography (n ⫽ 18)
Pelvic fracture, negative angiography (n ⫽ 13)
P value
14 (77.8%) 9 (50.0%) 3 (16.7%)
10 (76.9%) 3 (23.1%) 2 (15.4%)
1.000 .1581 1.000
10 (55.6%) 12 (66.7%) 9 (50.0%)
6 (46.2%) 4 (30.8%) 4 (30.8%)
.7224 .0732 .4622
13 (72.2%) 12 (66.7%)
8 (61.5%) 5 (38.5%)
.7007 .1572
8 2 0 8
(44.4%) (11.1%) (0%) (44.4%)
*P ⬍ .05 using the Mann–Whitney U test or chi-square test where appropriate.
4 7 0 2
(30.7%) (53.8%) (0%) (15.4%)
.4840 .0166* 1.000 .1285
756
The American Journal of Surgery, Vol 200, No 6, December 2010
ulation because of the high amount of energy that is needed to cause complex pelvic fractures. Hemodynamic instability has been shown previously to be present in up to 25% of patients admitted with grade I and II pelvic fractures.12 This can make identifying the source of blood loss in hemodynamically unstable patients less clear and require the need for additional diagnostic studies before proceeding to pelvic angiography. The use of pelvic angiography and therapeutic angiographic embolization often is discussed in the treatment of severe pelvic fracture. In this series of patient admitted to a Level 1 trauma center we found that only 3.8% of patients with pelvic fracture required pelvic angiography, with an even fewer number of patients undergoing therapeutic embolization. These results shed light on the true need for pelvic angiography in an urban trauma center. Other series have noted more frequent use of angiography for pelvic fractures, possibly owing to treatment algorithms with more liberal indications for angiography. Our relatively infrequent use of pelvic angiography in this series was not owing to a lack of interventional radiology support. Our institutional policy dictates that our interventional radiologists and angiogram suites are available 24 hours a day, 7 days per week, and can be ready to perform a procedure within 30 minutes of being contacted. There is also a dedicated orthopedic trauma service in-house that is available to perform emergent pelvic fixation at all times. Although angiographic embolization is a well-described treatment for pelvic bleeding, identifying the select patients who will benefit from this therapy has long been in debate. Prompt identification of patients who would benefit from angiography is important because angiography requires specialized equipment and expertise that is not available expeditiously in some trauma centers. Salim et al13 studied all trauma patients admitted with blunt pelvic fractures to identify patients who would benefit from therapeutic pelvic angiography. They found that female sex, presence of hypotension, and sacroiliac joint disruption were independent predictors for patients who would benefit from angiographic embolization. As expected, we found that patients requiring pelvic angiography were more severely injured than those patients who did not undergo angiography, with significantly higher ISS, longer hospital and ICU LOS, and increased mortality. This difference in the need for angiogram during our study period does not appear to be due to a less severely injured population of patients. When you compare the patients included in our study with the recent study by Salim et al,13 there is a similar injury severity in both populations of patients with pelvic fracture. The ISS in their pelvic fracture patients was very similar to ours, with an average ISS of 17 in their study compared with an ISS of 19 in our study. Only 1 of the 31 patients who underwent pelvic angiogram did not receive a CT scan before being taken to the angiography suite. This suggests that a majority of patients
undergoing angiogram in this series were not hemodynamically unstable owing to a clear source of bleeding within the pelvis. When comparing patients based on the results of their angiogram, the only difference found was an increased number of patients with an abdominal AIS of 3 or greater in the group that had positive findings on pelvic angiogram. The patients in this study who underwent angiogram were a very severely injured group with a mean ISS of 34. Our data suggest that the increased mortality in that group is not related directly to pelvic bleeding but rather to associated injuries caused by very severe mechanisms of injury. We studied pelvic radiographs in this population to analyze the use of pelvic fracture pattern as a predictor of positive angiography. We found that patients with a pelvic fracture pattern caused by anteroposterior compression who underwent a pelvic angiogram were significantly more likely to have negative angiogram findings. Pubic rami fractures and sacral fractures were the most common pelvic fracture pattern identified. We found that complex fracture patterns predominated over patients with a single pelvic fracture pattern. Interestingly, pelvic injuries caused by a vertical shear mechanism were seen infrequently in patients requiring angiography. Pelvic fracture pattern has been shown previously to be predictive of pelvic bleeding.14,15 However, a recent study by Sarin et al16 analyzed fracture patterns in a retrospective review of hypotensive patients with pelvic fracture, finding that pelvic fracture did not correlate consistently with the need for urgent angioembolization. Eastridge et al17 found that hypotensive patients with unstable pelvic fracture patterns were more likely to have hemorrhage from a pelvic source, suggesting that these patients would benefit from early pelvic angiography before celiotomy. In conclusion, we found that the need for pelvic angiography was relatively rare in 819 patients admitted with pelvic fracture to our urban, Level 1 trauma center during the 9-year study period. Angiographic arterial embolization remains an important treatment adjunct in severely injured patients with bleeding from pelvic fractures. Patients admitted to the trauma bay sustaining complex pelvic fractures with associated hypotension and tachycardia should be considered candidates for angiography. Although factors associated with the need for pelvic angiography frequently are debated, we may discuss angiography for pelvic fractures more often than it actually is performed. We may, in fact, be all talk when it comes to the true need for pelvic angiography in trauma patients admitted with pelvic fracture.
References 1. Demetriades D, Karaiskakis M, Toutouzas K, et al. Pelvic fractures: epidemiology and predictors of associated abdominal injuries and outcomes. J Am Coll Surg 2002;195:1–10. 2. Flint L, Babikian G, Anders M, et al. Definitive control of mortality from severe pelvic fracture. Ann Surg 1990;211:703–7.
T.W. Costantini et al.
Arterial embolization for pelvic fractures
3. Sathy AK, Starr AJ, Smith WR, et al. The effect of pelvic fracture on mortality after trauma: an analysis of 63,000 trauma patients. J Bone Joint Surg 2009;91:2803–10. 4. Suzuki T, Smith WR, Moore EE. Pelvic packing or angiography: competitive or complementary? Injury 2009;40:343–53. 5. Lopez PP. Unstable pelvic fractures: the use of angiography in controlling arterial hemorrhage. J Trauma 2007;62:S30 –1. 6. Velmahos GC, Toutouzas KG, Vassiliu P, et al. A prospective study on the safety and efficacy of angiographic embolization for pelvic and visceral injuries. J Trauma 2002;53:303– 8. 7. Croce MA, Magnotti LJ, Savage SA, et al. Emergent pelvic fixation in patients with exsanguinating pelvic fractures. J Am Coll Surg 2007; 204:935– 42. 8. Jeske HC, Larndorfer R, Krappinger D, et al. Management of hemorrhage in severe pelvic injuries. J Trauma 2010;68:415–20. 9. Osborn PM, Smith WR, Moore EE, et al. Direct retroperitoneal pelvic packing versus pelvic angiography: a comparison of two management protocols for haemodynamically unstable pelvic fractures. Injury 2009;40:54 – 60. 10. Cothren CC, Osborn PM, Moore EE, et al. Preperitoneal pelvic packing for hemodynamically unstable pelvic fractures: a paradigm shift. J Trauma 2007;62:834 – 42. 11. Stein DM, O’Toole R, Scalea TM. Multidisciplinary approach for patients with pelvic fractures and hemodynamic instability. Scand J Surg 2007;96:272– 80. 12. Gustavo Parreira J, Coimbra R, Rasslan S, et al. The role of associated injuries on outcome of blunt trauma patients sustaining pelvic fractures. Injury 2000;31:677– 82. 13. Salim A, Teixeira PG, DuBose J, et al. Predictors of positive angiography in pelvic fractures: a prospective study. J Am Coll Surg 2008; 207:656 – 62. 14. Dalal SA, Burgess AR, Siegel JH, et al. Pelvic fracture in multiple trauma: classification by mechanism is key to pattern of organ injury, resuscitative requirements, and outcome. J Trauma 1989;29:981–1002. 15. Magnussen RA, Tressler MA, Obremskey WT, et al. Predicting blood loss in isolated pelvic and acetabular high-energy trauma. J Orthop Trauma 2007;21:603–7. 16. Sarin EL, Moore JB, Moore EE, et al. Pelvic fracture pattern does not always predict the need for urgent embolization. J Trauma 2005;58: 973–7. 17. Eastridge BJ, Starr A, Minei JP, et al. The importance of fracture pattern in guiding therapeutic decision-making in patients with hemorrhagic shock and pelvic ring disruptions. J Trauma 2002;53:446 –51.
Comments Brian Eastridge, M.D. (Fort Sam Houston, TX): In this presentation, Dr Costantini and his colleagues reviewed nearly a decade of work from University of California, San Diego, with their experience with pelvic fractures. More importantly, he and his colleagues seek to dispel some of the surgical dogma that we all deal with on a day-to-day basis, particularly with respect to the use of angioembolization and the algorithm for significant pelvic fracture management. Of the 819 patients in the analysis, in this series, only 3.9% required angioembolization, which is a substantial departure from other series in the literature. So, with that, I have a couple of questions for the author and his colleagues. Your series had relatively few patients that were unstable, which is again a modest departure from other series. Were your results the effect of lesser-injured pelvic fracture population? Do you have better airbags in San Diego or is it a
757
different pelvic fracture population, different emergency medical service practice? Also, with respect to your mortality difference between the groups, do you think it was related to the hemorrhage from the pelvic fracture or from associated injuries within the population sample in those 2 groups? Classically, the utilization of angioembolization is really contingent upon the level of facility you are at, whether you have interventional radiologists that are willing to come in, particularly in the middle of the night, to help you out. So in your facility, how receptive or how available are your interventional radiology colleagues? Is that possibly a source of bias in your analysis? The last question I have is what effect did technology play in your results? This was a 10-year period in which there has been a substantial technology evolution, so do you think that potentially the technology evolution, particularly with respect to the advanced CT scanners or even pelvic packing would, which the national expert is up there at your left, had anything to do with your results, did you do any pelvic packing? Did that have any effect on the later patients in your analysis? Great work. Todd Costantini, M.D. (San Diego, CA): To answer your first question whether or not pelvic fracture patients in our series were less severely injured than others as a means to account for our lower rate of pelvic angiogram, actually I do not think that that is true. When you compare some of the other series, the recent study by Dr Saleem comes to mind, their ISS and their pelvic fracture patients was very similar to ours; I think theirs was 17 where ours was about 19.13 So I do not think we are looking at a very different patient population sustaining severe pelvic fracture. I think we may have a more stringent criteria for those that undergo angiography. To answer your question regarding the availability of interventional radiology at our institution, our interventional radiology colleagues are immediately available, within 30 minutes or so, to come in to perform angiograms. I do not think we have a bias to not call them because they are far away or difficult to get to come in at any time to perform procedures. I think it is just our institutional policy that we have not used angiography as liberally, but our interventional radiologists are certainly there when we need them. You asked whether or not the pelvic mortality in our patients was due to the pelvic fracture itself or to other injuries. I think when we look at the mortality in the patients who did undergo angiogram, since we are being a little more stringent with who is getting angiograms, we are obviously looking at a very severely injured group of patients with an ISS of 34. I do not think that the increased mortality in that group is directly related to pelvic bleeding because these people sustained such a high mechanism of injury, there are obviously a lot of associated injuries in this population. And, finally, you asked about technology, whether technology advances such as more sensitive CT scanners or the utilization of pelvic packing as a modality to treat these patients affected outcomes later in our analysis. I think newer, multidetector CT scanners may help visualize small areas of pelvic bleeding, however, if you have noticed, both
758
The American Journal of Surgery, Vol 200, No 6, December 2010
in our study and recent studies, the presence of blush on a pelvic CT scan is not a predictor for positive angiography. So I think if we are able to detect pelvic blush or pelvic hematoma more sensitively, I do not think it is necessarily going to help us identify more patients that are going to benefit for angiography. Peter Rhee, M.D. (Tucson, AZ): Great paper. You had a big study here, which you did over a lot of years. I was just wondering did you notice a trend on the frequency of angiogram needed over the years? We found in our institution that the method of resuscitation changed over the years
and this is similar to many institutions in this country. We minimize crystalloids and do not hemodilute as much as before and this seems to decrease bleeding. Did you find a similar trend in your study? Dr Todd Costantini: We really did not see a significant difference in the frequency of pelvic angiogram over the duration of our study. The patients needing angiography were spread out fairly evenly over the 9-year period. There was a little more pelvic angiogram actually earlier on as you suggested, but I cannot really make any conclusions based on our data.
Arterial embolization for pelvic fractures after blunt trauma: are we all talk? Todd W. Costantini, M.D., Patrick L. Bosarge, M.D., Dale Fortlage, B.A., Vishal Bansal, M.D., Raul Coimbra, M.D., Ph.D.* Division of Trauma, Surgical Critical Care, and Burns, Department of Surgery, University of California, San Diego School of Medicine, 200 W. Arbor Dr, 8896, San Diego, CA 92103-8896, USA KEYWORDS: Angiography; Angioembolization; Fracture patterns; Pelvic injury mechanism
Abstract BACKGROUND: We hypothesized that arterial embolization for bleeding after pelvic fracture is used relatively infrequently. We sought to identify the true need for arterial embolization and define injury patterns associated with successful therapeutic angiographic embolization. METHODS: A retrospective review identified patients admitted to our urban, Level 1 trauma center with pelvic fractures from 2001 to 2009. Patients requiring pelvic arterial angiogram and embolization of pelvic bleeding were reviewed for pelvic fracture pattern and pelvic injury mechanism. RESULTS: There were 819 patients diagnosed with pelvic fractures, with only 31 patients (3.8%) undergoing diagnostic pelvic angiography. Of those, 18 patients (58.1%) had active bleeding requiring arterial embolization. Complex pelvic fracture patterns were common in patients undergoing angiogram. Patients undergoing pelvic angiography with an anteroposterior compression mechanism were more likely to have negative findings on angiogram. CONCLUSIONS: The actual need for angiography and therapeutic embolization is quite small in patients sustaining pelvic fracture. Although factors associated with the need for pelvic angiography frequently are debated, we may discuss angiography for pelvic fractures more often than it actually is performed. © 2010 Elsevier Inc. All rights reserved.
Traumatic pelvic fractures can result in significant hemorrhage that can be associated with significant morbidity and mortality. Patients with pelvic fractures that cause hemodynamic instability have a mortality rate that exceeds 50% in several series.1–3 Therefore, identifying patients at risk for severe bleeding because of their pelvic fracture and prompt mobilization of resources to treat those patients is paramount. Presented as an oral presentation at the 2010 annual meeting of the Southwest Surgical Congress, March 21 to 24, 2010, Tucson, AZ. * Corresponding author. Tel.: ⫹1-619-543-7100; fax: ⫹1-619-5437202. E-mail address: [email protected] Manuscript received March 5, 2010; revised manuscript June 15, 2010
0002-9610/$ - see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2010.06.006
Bleeding from pelvic fractures can be generated from several sources including arterial and venous injury, and bleeding from fractured cancellous bone within the pelvis. Bleeding from fractured bone within the pelvis can be controlled with prompt stabilization of the fracture, which also can tamponade and control venous bleeding.4 Arterial injury caused by pelvic fracture may require therapeutic arterial embolization, which has been shown to be an important adjunct in the treatment of patients with pelvic hemorrhage.5,6 Therefore, management of pelvic fractures may require care from a multidisciplinary team of trauma surgeons, orthopedic surgeons, and interventional radiologists. Although the risk of significant hemorrhage after pelvic fracture is clear, the incidence and predictors for needing
T.W. Costantini et al.
Arterial embolization for pelvic fractures
therapeutic arterial embolization remain in debate. Criteria exist for obtaining pelvic arterial angiography in patients with severe pelvic fracture based on the presence of hemodynamic instability or the need for ongoing blood transfusion. Clinicians must decide between obtaining further diagnostic testing such as computerized tomography (CT) scan or proceed immediately to angiography. More liberal use of angiography in pelvic fractures based exclusively on hemodynamics or on the amount of blood transfused, in an era of multislice CT scanner with high-definition images and a high likelihood of identifying contrast extravasation in the arterial phase of the examination, may not be the best strategy for patient care. Obviously, the risks of multiple contrast loads and radiation exposure should be taken into account when deciding which therapeutic strategy will be used. Therefore, early prediction of the need for angiography is paramount. In this study, we hypothesized that arterial embolization for bleeding after pelvic fracture is used relatively infrequently, and sought to identify the incidence and true need for arterial embolization. We also sought to define injury patterns associated with successful therapeutic angiographic embolization in an attempt to provide information to clinicians regarding resource mobilization.
753
agnostic pelvic angiogram with possible embolization is attempted. If FAST or DPL are positive, an immediate exploratory laparotomy is performed. If a large or expanding pelvic/retroperitoneal hematoma is found, an immediate postoperative angiogram and possible embolization is performed. Data obtained included demographics, mechanism of injury, injury severity score (ISS), admission vital signs, admission Glasgow Coma Scale, Abbreviated Injury Score (AIS), hospital and intensive care unit (ICU) length of stay (LOS), and mortality. Autopsy records were reviewed for those patients who died after a pelvic fracture requiring angiography. Patients requiring arterial embolization were stratified further for pelvic fracture pattern and pelvic injury mechanism. A senior trauma surgeon blinded to the study groups reviewed all radiographic studies to determine pelvic fracture patterns and pelvic injury mechanism (anteroposterior compression, lateral compression of vertical shear). Patients with multiple pelvic fracture patterns were categorized as having a complex mechanism of injury. Values are expressed as the mean ⫾ standard deviation. Statistical analysis was performed using SPSS Statistics software version 11.5 (SPSS, Chicago, IL). The Student t test, Mann–Whitney U test, and chi-square test were used to determine statistical significance where appropriate. Statistical significance was defined as a P value of less than .05.
Methods A retrospective review identified all trauma patients admitted with pelvic fractures to the University of California, San Diego Medical Center, an urban, Level 1 trauma center between January 1, 2000, and January 1, 2009. Patient data were obtained using our trauma registry database. Medical records and radiographic studies were used to supplement our data collection. Patients requiring pelvic arterial angiogram and arterial embolization of pelvic bleeding were identified. This study received approval from the University of California, San Diego Institutional Review Board. Our institutional protocol since 2000 has been to use angiography, when indicated, early after potentially severe pelvic injury. However, we have relied on CT extravasation as an absolute indication to perform arterial angiography. Our protocol is as follows: hemodynamically stable patients are screened for intra-abdominal injury with focused assessment with sonography for trauma (FAST). If FAST is positive, a contrast CT of the abdomen is obtained. If positive for contrast extravasation (blush) in the pelvis, an angiogram and possible embolization is performed. If FAST is negative and the patient shows a decreased level of hematocrit over time requiring 4 U of blood transfusion within 8 hours, without other possible sources of bleeding, a diagnostic angiography and possible embolization is attempted. Hemodynamically unstable patients undergo FAST or diagnostic peritoneal lavage (DPL) to rule out intra-abdominal hemorrhage. If FAST or DPL are negative, other sources of bleeding have been ruled out, and the patient remains unstable (with a decreasing hematocrit) then a di-
Results Study population During the 9-year study period, there were 819 patients diagnosed with pelvic fracture. Characteristics of this population of patients with pelvic fractures are presented in Table 1. Patients sustaining pelvic fractures were predominantly young with a mean age of 41.4 ⫾ 19.6 years and
Table 1
Demographics of the study population All pelvic fractures
n Age, y Male, n (%) Admission vitals Systolic blood pressure, mm Hg Heart rate Respiratory rate Admission base deficit Admission GCS ISS ICU LOS, d Hospital LOS, d Mortality, n (%) GCS ⫽ Glasgow Coma Scale.
819 41.4 ⫾ 19.6 488 (59.6) 124.5 ⫾ 34.6 94.6 ⫾ 25.8 17.4 ⫾ 7.9 ⫺2.8 ⫾ 4.9 13.0 ⫾ 4.0 19.6 ⫾ 13.2 4.1 ⫾ 8.8 11.9 ⫾ 16.0 64 (7.8)
754
The American Journal of Surgery, Vol 200, No 6, December 2010
Table 2
Characteristics of pelvic fracture patients based on the need for arterial angiography (n ⫽ 819 patients)
n, (%) Age, y Male, n (%) Admission vitals Systolic blood pressure, mm Hg Heart rate Respiratory rate Admission base deficit Admission GCS ISS ICU LOS, d Hospital LOS, d Mortality, n (%)
Pelvic fracture, no angiography
Pelvic fracture, all angiography
788 (96.2) 41.2 ⫾ 19.6 469 (59.5)
31 (3.8) 45.9 ⫾ 19.8 19 (61.3)
.1909 1.000
125.1 ⫾ 31.5 94.0 ⫾ 25.7 17.3 ⫾ 7.8 ⫺2.7 ⫾ 4.7 13.0 ⫾ 3.9 19.2 ⫾ 13.0 3.9 ⫾ 8.5 11.5 ⫾ 15.2 58 (7.4)
109.9 ⫾ 32.5 109.3 ⫾ 24.8 17.6 ⫾ 9.7 ⫺4.1 ⫾ 7.0 12.5 ⫾ 4.2 30.5 ⫾ 15.1 10.5 ⫾ 14.1 22.2 ⫾ 28.7 7 (22.6)
.0086* .0012* 1.000 .1119 .4853 ⬍.0001* ⬍.0001* .0003* .0082*
P value
GCS ⫽ Glasgow Coma Scale. *P ⬍ .05 using the Student t test or chi-square test where appropriate.
male (59.6%). The mean ISS was 19.6 ⫾ 13.2, with an overall mortality rate of 7.8% in this population of patients.
Patients requiring arterial angiogram Of the 819 patients sustaining pelvic fractures during the study period, only 31 patients (3.8%) underwent pelvic angiography. Clinical characteristics of these patients are compared in Table 2. Only 1 patient did not receive a CT scan before being taken to the angiography suite. This patient, who was hemodynamically unstable, was admitted directly to the operating room as part of our operating room resuscitation protocol. After placement of an external pelvic fixator by the orthopedic team, the patient was taken directly to the angiography suite for angioembolization of his left internal pudendal artery. Patients requiring pelvic angiography had significantly lower systolic blood pressure and increased heart rate on admission. Patients with pelvic fractures who underwent pelvic angiography also were injured more severely with an average ISS of 30.5 ⫾ 15.1, compared with an ISS of 19.2 ⫾ 13.0 in the group of pelvic fractures that did not require angiography (P ⬍ .0001). This was associated with an increased hospital and ICU LOS for patients requiring pelvic angiography. Mortality was increased in the group of patients with pelvic fracture who underwent pelvic angiography (22.6% vs 7.4%; P ⬍ .01). Autopsy reports were reviewed for patients who underwent pelvic angiography and died during that hospitalization. Of the 6 patients who underwent angiography and subsequently died, 2 died because of severe pelvic crush injuries after being struck by an automobile. The remainder of the mortalities was due to multiple blunt force injuries.
Patients with positive pelvic angiography Characteristics of patients with pelvic fracture who underwent pelvic angiography are displayed in Table 3. Of the
31 patients who underwent pelvic angiography after pelvic fracture, 18 patients (58.1%) had positive angiogram findings of active arterial extravasation. Patients with a positive finding on pelvic angiogram were more likely to have an abdominal AIS of 3 or greater. Thirteen of the 31 patients (41.9%) undergoing pelvic angiography had a negative pelvic angiogram; however, therapeutic angioembolization was performed in 4 of these patients because of concern for hemorrhage in areas of fracture or hematoma.
Pelvic fracture pattern in patients undergoing pelvic angiography Pelvic fracture pattern and pelvic injury mechanism for patients who underwent pelvic angiography are presented in Table 4. Pubic rami fracture was the most common pelvic fracture seen in patients undergoing angiography, regardless of angiographic findings. Of patients with positive findings on angiogram, sacral fracture and sacroiliac joint disruption were common pelvic fracture patterns. Of patients who underwent pelvic angiography, those with a mechanism of only anteroposterior compression were more likely to have negative angiographic findings. Several patients had complex mechanisms of injury, with the presence of more than 1 pelvic injury mechanism (most commonly a combination of lateral compression and anteroposterior compression). A pure vertical shear mechanism was the pelvic injury mechanism seen least frequently in patients undergoing pelvic angiography.
Comments Pelvic fractures after blunt trauma are common and may represent a wide spectrum of severity. Significant pelvic fractures require a substantial amount of force and often are
T.W. Costantini et al. Table 3
Arterial embolization for pelvic fractures
755
Pelvic fracture patients undergoing arterial angiography (n ⫽ 31 patients) Pelvic fracture, positive angiography (n ⫽ 18)
Age, y Male, n (%) Admission vitals Systolic blood pressure Heart rate Respiratory rate Base deficit Admission GCS ISS Head AIS ⱖ 3 Chest AIS ⱖ 3 Abdominal AIS ⱖ 3 Extremity AIS ⱖ 3 CT scan obtained Blush on CT scan Angiographic intervention ICU LOS, d Hospital LOS, d Mortality, n (%)
Pelvic fracture, negative angiography (n ⫽ 13)
P value
47.6 ⫾ 20.6 13 (72.2)
43.5 ⫾ 19.2 6 (46.2)
.5616 .2623
104.7 ⫾ 29.5 112.3 ⫾ 24.6 19.0 ⫾ 9.7 ⫺5.0 ⫾ 4.3 12.2 ⫾ 4.2 34.5 ⫾ 15.2 9 (50.0%) 9 (50.0%) 15 (83.3%) 18 (100%) 17 (94.4%) 4 (22.2%) 18 (100%) 7.0 ⫾ 7.0 15.0 ⫾ 15.0 6 (33.3)
116.6 ⫾ 36.1 105.2 ⫾ 25.5 15.7 ⫾ 9.8 ⫺3.1 ⫾ 9.4 12.8 ⫾ 4.4 24.9 ⫾ 13.6 2 (15.4%) 7 (53.8%) 5 (38.5%) 11 (84.6%) 13 (100%) 6 (46.2%) 4 (30.8%) 15.4 ⫾ 19.5 32.3 ⫾ 39.5 1 (7.7)
.5864 .4233 .4469 .4544 .7032 .1093 .0656 1.000 .0209* .1677 1.000 .2469 ⬍.0001* .1014 .0990 .1912
GCS ⫽ Glasgow Coma Scale. *P ⬍ .05 using the Mann–Whitney U test or chi-square test where appropriate.
seen in conjunction with other serious injuries. Therefore, early decision making in determining and stabilizing the source of hemorrhage in the severely injured patient with pelvic fracture may significantly affect outcome. Multiple methods have been described to stabilize patients with significant hemorrhage from pelvic fracture, including pelvic fixation devices, external pelvic fixation, direct retroperitoneal packing, transabdominal packing, and arterial angiography.7–10 Because several of these methods require a multidisciplinary team, identifying factors that successfully
Table 4
predict successful management of severe pelvic injuries may be important for resource mobilization and use. Angiography and angiographic embolization have been used for decades to diagnose and treat patients with arterial injury caused by pelvic fracture.11 Common indications for angiography after pelvic trauma include hemodynamic instability, extravasation of contrast seen from pelvic vessels on CT scan, large pelvic hematoma on CT, or ongoing need for blood transfusion as a result of pelvic bleeding. Other life-threatening associated injuries are common in this pop-
Pelvic fracture patterns and pelvic injury mechanisms for patients requiring angiography (n ⫽ 31 patients)
Anterior pelvic injury Pubic rami fracture Widened pubic symphysis Open book injury Lateral pelvic injury Acetabular fracture Impacted sacrum Iliac wing fracture Posterior pelvic injury Sacral fracture Sacroiliac joint disruption Pelvic injury mechanism Lateral compression Anteroposterior compression Vertical shear Complex mechanism
Pelvic fracture, positive angiography (n ⫽ 18)
Pelvic fracture, negative angiography (n ⫽ 13)
P value
14 (77.8%) 9 (50.0%) 3 (16.7%)
10 (76.9%) 3 (23.1%) 2 (15.4%)
1.000 .1581 1.000
10 (55.6%) 12 (66.7%) 9 (50.0%)
6 (46.2%) 4 (30.8%) 4 (30.8%)
.7224 .0732 .4622
13 (72.2%) 12 (66.7%)
8 (61.5%) 5 (38.5%)
.7007 .1572
8 2 0 8
(44.4%) (11.1%) (0%) (44.4%)
*P ⬍ .05 using the Mann–Whitney U test or chi-square test where appropriate.
4 7 0 2
(30.7%) (53.8%) (0%) (15.4%)
.4840 .0166* 1.000 .1285
756
The American Journal of Surgery, Vol 200, No 6, December 2010
ulation because of the high amount of energy that is needed to cause complex pelvic fractures. Hemodynamic instability has been shown previously to be present in up to 25% of patients admitted with grade I and II pelvic fractures.12 This can make identifying the source of blood loss in hemodynamically unstable patients less clear and require the need for additional diagnostic studies before proceeding to pelvic angiography. The use of pelvic angiography and therapeutic angiographic embolization often is discussed in the treatment of severe pelvic fracture. In this series of patient admitted to a Level 1 trauma center we found that only 3.8% of patients with pelvic fracture required pelvic angiography, with an even fewer number of patients undergoing therapeutic embolization. These results shed light on the true need for pelvic angiography in an urban trauma center. Other series have noted more frequent use of angiography for pelvic fractures, possibly owing to treatment algorithms with more liberal indications for angiography. Our relatively infrequent use of pelvic angiography in this series was not owing to a lack of interventional radiology support. Our institutional policy dictates that our interventional radiologists and angiogram suites are available 24 hours a day, 7 days per week, and can be ready to perform a procedure within 30 minutes of being contacted. There is also a dedicated orthopedic trauma service in-house that is available to perform emergent pelvic fixation at all times. Although angiographic embolization is a well-described treatment for pelvic bleeding, identifying the select patients who will benefit from this therapy has long been in debate. Prompt identification of patients who would benefit from angiography is important because angiography requires specialized equipment and expertise that is not available expeditiously in some trauma centers. Salim et al13 studied all trauma patients admitted with blunt pelvic fractures to identify patients who would benefit from therapeutic pelvic angiography. They found that female sex, presence of hypotension, and sacroiliac joint disruption were independent predictors for patients who would benefit from angiographic embolization. As expected, we found that patients requiring pelvic angiography were more severely injured than those patients who did not undergo angiography, with significantly higher ISS, longer hospital and ICU LOS, and increased mortality. This difference in the need for angiogram during our study period does not appear to be due to a less severely injured population of patients. When you compare the patients included in our study with the recent study by Salim et al,13 there is a similar injury severity in both populations of patients with pelvic fracture. The ISS in their pelvic fracture patients was very similar to ours, with an average ISS of 17 in their study compared with an ISS of 19 in our study. Only 1 of the 31 patients who underwent pelvic angiogram did not receive a CT scan before being taken to the angiography suite. This suggests that a majority of patients
undergoing angiogram in this series were not hemodynamically unstable owing to a clear source of bleeding within the pelvis. When comparing patients based on the results of their angiogram, the only difference found was an increased number of patients with an abdominal AIS of 3 or greater in the group that had positive findings on pelvic angiogram. The patients in this study who underwent angiogram were a very severely injured group with a mean ISS of 34. Our data suggest that the increased mortality in that group is not related directly to pelvic bleeding but rather to associated injuries caused by very severe mechanisms of injury. We studied pelvic radiographs in this population to analyze the use of pelvic fracture pattern as a predictor of positive angiography. We found that patients with a pelvic fracture pattern caused by anteroposterior compression who underwent a pelvic angiogram were significantly more likely to have negative angiogram findings. Pubic rami fractures and sacral fractures were the most common pelvic fracture pattern identified. We found that complex fracture patterns predominated over patients with a single pelvic fracture pattern. Interestingly, pelvic injuries caused by a vertical shear mechanism were seen infrequently in patients requiring angiography. Pelvic fracture pattern has been shown previously to be predictive of pelvic bleeding.14,15 However, a recent study by Sarin et al16 analyzed fracture patterns in a retrospective review of hypotensive patients with pelvic fracture, finding that pelvic fracture did not correlate consistently with the need for urgent angioembolization. Eastridge et al17 found that hypotensive patients with unstable pelvic fracture patterns were more likely to have hemorrhage from a pelvic source, suggesting that these patients would benefit from early pelvic angiography before celiotomy. In conclusion, we found that the need for pelvic angiography was relatively rare in 819 patients admitted with pelvic fracture to our urban, Level 1 trauma center during the 9-year study period. Angiographic arterial embolization remains an important treatment adjunct in severely injured patients with bleeding from pelvic fractures. Patients admitted to the trauma bay sustaining complex pelvic fractures with associated hypotension and tachycardia should be considered candidates for angiography. Although factors associated with the need for pelvic angiography frequently are debated, we may discuss angiography for pelvic fractures more often than it actually is performed. We may, in fact, be all talk when it comes to the true need for pelvic angiography in trauma patients admitted with pelvic fracture.
References 1. Demetriades D, Karaiskakis M, Toutouzas K, et al. Pelvic fractures: epidemiology and predictors of associated abdominal injuries and outcomes. J Am Coll Surg 2002;195:1–10. 2. Flint L, Babikian G, Anders M, et al. Definitive control of mortality from severe pelvic fracture. Ann Surg 1990;211:703–7.
T.W. Costantini et al.
Arterial embolization for pelvic fractures
3. Sathy AK, Starr AJ, Smith WR, et al. The effect of pelvic fracture on mortality after trauma: an analysis of 63,000 trauma patients. J Bone Joint Surg 2009;91:2803–10. 4. Suzuki T, Smith WR, Moore EE. Pelvic packing or angiography: competitive or complementary? Injury 2009;40:343–53. 5. Lopez PP. Unstable pelvic fractures: the use of angiography in controlling arterial hemorrhage. J Trauma 2007;62:S30 –1. 6. Velmahos GC, Toutouzas KG, Vassiliu P, et al. A prospective study on the safety and efficacy of angiographic embolization for pelvic and visceral injuries. J Trauma 2002;53:303– 8. 7. Croce MA, Magnotti LJ, Savage SA, et al. Emergent pelvic fixation in patients with exsanguinating pelvic fractures. J Am Coll Surg 2007; 204:935– 42. 8. Jeske HC, Larndorfer R, Krappinger D, et al. Management of hemorrhage in severe pelvic injuries. J Trauma 2010;68:415–20. 9. Osborn PM, Smith WR, Moore EE, et al. Direct retroperitoneal pelvic packing versus pelvic angiography: a comparison of two management protocols for haemodynamically unstable pelvic fractures. Injury 2009;40:54 – 60. 10. Cothren CC, Osborn PM, Moore EE, et al. Preperitoneal pelvic packing for hemodynamically unstable pelvic fractures: a paradigm shift. J Trauma 2007;62:834 – 42. 11. Stein DM, O’Toole R, Scalea TM. Multidisciplinary approach for patients with pelvic fractures and hemodynamic instability. Scand J Surg 2007;96:272– 80. 12. Gustavo Parreira J, Coimbra R, Rasslan S, et al. The role of associated injuries on outcome of blunt trauma patients sustaining pelvic fractures. Injury 2000;31:677– 82. 13. Salim A, Teixeira PG, DuBose J, et al. Predictors of positive angiography in pelvic fractures: a prospective study. J Am Coll Surg 2008; 207:656 – 62. 14. Dalal SA, Burgess AR, Siegel JH, et al. Pelvic fracture in multiple trauma: classification by mechanism is key to pattern of organ injury, resuscitative requirements, and outcome. J Trauma 1989;29:981–1002. 15. Magnussen RA, Tressler MA, Obremskey WT, et al. Predicting blood loss in isolated pelvic and acetabular high-energy trauma. J Orthop Trauma 2007;21:603–7. 16. Sarin EL, Moore JB, Moore EE, et al. Pelvic fracture pattern does not always predict the need for urgent embolization. J Trauma 2005;58: 973–7. 17. Eastridge BJ, Starr A, Minei JP, et al. The importance of fracture pattern in guiding therapeutic decision-making in patients with hemorrhagic shock and pelvic ring disruptions. J Trauma 2002;53:446 –51.
Comments Brian Eastridge, M.D. (Fort Sam Houston, TX): In this presentation, Dr Costantini and his colleagues reviewed nearly a decade of work from University of California, San Diego, with their experience with pelvic fractures. More importantly, he and his colleagues seek to dispel some of the surgical dogma that we all deal with on a day-to-day basis, particularly with respect to the use of angioembolization and the algorithm for significant pelvic fracture management. Of the 819 patients in the analysis, in this series, only 3.9% required angioembolization, which is a substantial departure from other series in the literature. So, with that, I have a couple of questions for the author and his colleagues. Your series had relatively few patients that were unstable, which is again a modest departure from other series. Were your results the effect of lesser-injured pelvic fracture population? Do you have better airbags in San Diego or is it a
757
different pelvic fracture population, different emergency medical service practice? Also, with respect to your mortality difference between the groups, do you think it was related to the hemorrhage from the pelvic fracture or from associated injuries within the population sample in those 2 groups? Classically, the utilization of angioembolization is really contingent upon the level of facility you are at, whether you have interventional radiologists that are willing to come in, particularly in the middle of the night, to help you out. So in your facility, how receptive or how available are your interventional radiology colleagues? Is that possibly a source of bias in your analysis? The last question I have is what effect did technology play in your results? This was a 10-year period in which there has been a substantial technology evolution, so do you think that potentially the technology evolution, particularly with respect to the advanced CT scanners or even pelvic packing would, which the national expert is up there at your left, had anything to do with your results, did you do any pelvic packing? Did that have any effect on the later patients in your analysis? Great work. Todd Costantini, M.D. (San Diego, CA): To answer your first question whether or not pelvic fracture patients in our series were less severely injured than others as a means to account for our lower rate of pelvic angiogram, actually I do not think that that is true. When you compare some of the other series, the recent study by Dr Saleem comes to mind, their ISS and their pelvic fracture patients was very similar to ours; I think theirs was 17 where ours was about 19.13 So I do not think we are looking at a very different patient population sustaining severe pelvic fracture. I think we may have a more stringent criteria for those that undergo angiography. To answer your question regarding the availability of interventional radiology at our institution, our interventional radiology colleagues are immediately available, within 30 minutes or so, to come in to perform angiograms. I do not think we have a bias to not call them because they are far away or difficult to get to come in at any time to perform procedures. I think it is just our institutional policy that we have not used angiography as liberally, but our interventional radiologists are certainly there when we need them. You asked whether or not the pelvic mortality in our patients was due to the pelvic fracture itself or to other injuries. I think when we look at the mortality in the patients who did undergo angiogram, since we are being a little more stringent with who is getting angiograms, we are obviously looking at a very severely injured group of patients with an ISS of 34. I do not think that the increased mortality in that group is directly related to pelvic bleeding because these people sustained such a high mechanism of injury, there are obviously a lot of associated injuries in this population. And, finally, you asked about technology, whether technology advances such as more sensitive CT scanners or the utilization of pelvic packing as a modality to treat these patients affected outcomes later in our analysis. I think newer, multidetector CT scanners may help visualize small areas of pelvic bleeding, however, if you have noticed, both
758
The American Journal of Surgery, Vol 200, No 6, December 2010
in our study and recent studies, the presence of blush on a pelvic CT scan is not a predictor for positive angiography. So I think if we are able to detect pelvic blush or pelvic hematoma more sensitively, I do not think it is necessarily going to help us identify more patients that are going to benefit for angiography. Peter Rhee, M.D. (Tucson, AZ): Great paper. You had a big study here, which you did over a lot of years. I was just wondering did you notice a trend on the frequency of angiogram needed over the years? We found in our institution that the method of resuscitation changed over the years
and this is similar to many institutions in this country. We minimize crystalloids and do not hemodilute as much as before and this seems to decrease bleeding. Did you find a similar trend in your study? Dr Todd Costantini: We really did not see a significant difference in the frequency of pelvic angiogram over the duration of our study. The patients needing angiography were spread out fairly evenly over the 9-year period. There was a little more pelvic angiogram actually earlier on as you suggested, but I cannot really make any conclusions based on our data.