The American Journal of Surgery (2016) -, -–-
Patients with pelvic fractures from blunt trauma. What is the cause of mortality and when? Rahul Vaidya, M.D., F.R.C.Sc.*, Alesha N. Scott, D.O., Fred Tonnos, D.O., Ian Hudson, D.O., M.P.H., Adam J. Martin, B.S., Anil Sethi, M.D. 4G University Health Centre, Detroit Receiving Hospital, 4201 St. Antoine Blvd, Detroit, MI 48201, USA
KEYWORDS: Pelvic fracture; Mortality; Survival time
Abstract BACKGROUND: Mortality in patients sustaining pelvic fractures has been reported to be 4% to 15%. We sought to investigate the cause of death based on timing and evaluate if type of fracture and Injury Severity Score have an influence on the survival time. METHODS: Sixty-nine patients of eight hundred sixty seven with a pelvic fracture who died during their hospital admission were included. Fractures were classified using the Arbeitsgemeinschaft Osteosynthesefragen/Orthopaedic Trauma Association system. Cause determined by autopsy in 48/69. RESULTS: The leading cause of death within 6 hours was abdominal and pelvic hemorrhage; 6 to 24 hours head injury, and greater than 24 hours multiple organ dysfunction syndrome. Survival time did not correlate between fracture type (P , .12) or Injury Severity Score. Only 2 patients died of isolated pelvic hemorrhage. CONCLUSIONS: Despite the advances made in acute management of the traumatized patient in the emergency department, mortality is unavoidable in a small group of patients with hemorrhage being the commonest cause of early death but isolated pelvic hemorrhage rare. Ó 2016 Elsevier Inc. All rights reserved.
Mortality in patients sustaining pelvic fractures has been reported to be 4% to 15%.1–7 and are usually related to multiple trauma and massive hemorrhage.8–12
Investigation performed at Detroit Receiving Hospital, Detroit, MI, 48201, USA. There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. The authors declare no conflicts of interest. Presented at the 2015 MSA Annual Meeting Stuart Woods Memorial Lecture * Corresponding author. Tel.: 11-313-966-7852; fax: 11-313-966-8400. E-mail address:
[email protected] Manuscript received August 6, 2015; revised manuscript August 26, 2015 0002-9610/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2015.08.038
Hemorrhage as a cause of death has decreased in the last decade possibly as a result of improved management techniques.10 Various risk factors for mortality have been observed and include increasing age and shock on initial presentation as defined by a systolic blood pressure less than 90 mm Hg.1,2,6,8,10,13,14 Injury Severity Score (ISS) is a measurement of the cumulative injuries per body area and it has been evaluated in relation to mortality with a higher ISS being associated with a higher risk of mortality.3,7,15–18 Despite several studies evaluating patients that died after a pelvis fracture there is no universal consensus as to the cause of death after these injuries. Furthermore, no distinction has been made in the cause of death between early (within 24 hours) and late deaths
2
The American Journal of Surgery, Vol -, No -, - 2016
(after 24 hours). We felt we could do a better job by carefully reviewing charts in our database and autopsy reports rather than relying on papers with cumulative data from multiple centers as their protocols may be varied, the patient population may be heterogeneous and the data could have been collected over decades. We decided to limit our collection over a 12-year period in this century where we have used standardized Advanced Trauma Life Support protocols. It was our goal to evaluate fracture pattern, ISS, and cause of death in patients who died in an imminent manner (,6 hours), those who died early (6 to 24 hours), and those who died late (.24 hours).
surgery performed (cardiothoracic, general, neurosurgical, orthopedic, or vascular), blood products given during hospital course, temporizing pelvic stabilization measures taken, and the course of events leading up to death. Radiographic images and reports were assessed and fractures were classified according to the Arbeitsgemeinschaft Osteosynthesefragen/Orthopaedic Trauma Association (OTA) fracture classification system.20 Patients were divided into causes of death and most patients died of; hemorrhage of pelvis/thorax/abdomen, intracranial hemorrhage, and multiorgan dysfunction syndrome/acute respiratory distress syndrome (MODS/ ARDS)/cardiopulmonary arrest. Time of death was grouped into imminent (,6 hours), early (6 to 24 hours), and late (greater than 24 hours). Given expected nonparametric distribution of survival time, Kruskal-Wallis testing was used with survival time as a function of cause of death. Post-hoc subgroup analysis to assess differences between pairs of groups was performed. ISS, though it gives the impression of being interval data, carries a strongly non-normal, right-skewed distribution inadequately remediated by transformation.21 Relationship between ISS and survival time was therefore explored using ranked ANCOVA, controlling for age as the covariate. Partial correlations were performed between age against survival time in minutes and against ISS. All calculations were performed using SAS 9.3 (SAS Institute, Cary, NC).
Methods Setting and patients We performed an institutional review board approved retrospective review from our level 1 Trauma database Between 1999 and 2013, a total of 867 trauma patients with pelvic fractures were admitted to our level 1 trauma center, 130 of which died. This group does not include fractures resulting from insufficiency, tumor, or infection. From the 130 patients that died, we excluded fractures sustained from gunshot wounds (23) and those patients who presented dead on arrival (38), leaving 69 patients that constituted the study cohort. The mean patient age was 51 6 20 (median 49, range 16 to 99), and there were 55 males and 14 females. On presentation to the emergency department, patients were evaluated and treated per the Advanced Trauma Life Support guidelines.19 The protocol includes initial management with blood transfusion and emergent transfer to the operating room for persistently unstable patients for external fixation followed by embolization if indicated. In the last 3 years, we have been more aggressive with retroperitoneal packing. If the patient demonstrates hemodynamic instability after packing they are considered for arterial embolization.
Procedures and data analysis Basic patient data were obtained from our standardized trauma database for the years 1999 to 2013 and put into customized spreadsheets for analysis. Patients who carried the International Classification of Diseases, 9th Revision diagnosis code 808.x were selected from the database, excluding patients with isolated codes 808.0 and 808.1 (acetabular fracture only). Electronic medical records were reviewed to gather detailed data from emergency department treatment notes, death reports, operative reports (open and angiographic), radiographs (X-ray images, computed tomography scan images, and radiology reads), and official medical examiner autopsy reports or autopsy (48/69) or consensus (21). Demographic data evaluated included age, method of injury, length of stay, cause of death, ISS, the type of
Results A total of 69 patients (8%) of 867 with pelvic fractures died a median of 24 hours and 37 minutes after admission with a range of 10 minutes to 932 hours and 36 minutes. Including the 23 patients that presented with pelvic fracture after gunshot wound, the death rate rose to 10.6%, but these patients were not a part of this cohort. The leading cause of death within 6 hours was abdominal and pelvic bleeding (23/25), between 6 and 24 hours was head injury (8/12), and after 24 hours was MODS (14/32) (Fig. 1). Most patients who died of hemorrhage did so from multiple areas of bleeding and only 2 patients died of isolated pelvic
Survival time Total Etiology
<6h
6-24h
>24h
Hemorrhage
23
4
10
37
Head injury
2
8
8
18
MODS/ARDS
-
-
14
14
25
12
32
69
Thoracic/Abdominal
Figure 1
Cause of death vs survival time.
R. Vaidya et al.
Causes of mortality from pelvic fractures
3
hemorrhage. The most common methods of injury were automobile vs pedestrian collision (26) and motor vehicle collision (MVC; 22). The other methods of injury were motorcycle crash (8), fall (8), and crush injury (5). After evaluation of the course of events of the hospital admission, all surviving patients were triaged to the operating room, intensive care, or the regular floor. Thirty-two patients in the present study were taken to the operating room emergently for immediate exploratory laparotomy with 12 of these undergoing immediate external fixation of the pelvis. Nine patients were transported to the operating room via the angiography suite, and 27 patients were shifted to the intensive care unit from the Emergency Department. Six patients died in the emergency department after arrival.22 Four patients were initially stable enough to be transferred to the regular ward. Five patients died on the operating table during the exploratory laparotomy; of which 1 patient underwent a thoracotomy to address an intercostal artery rupture. Fifty-five patients expired in the surgical or neurosurgical intensive care unit, 2 patients died on the regular hospital floor, and 1 patient expired in a palliative care unit. The mean ISS was 34 6 11.3 (median 35, range 13 to 66). ISS was not determined to be significantly different among patients when dividing them into groups based on length of survival (,6 hours, 6 hours to 24 hours, .24 hours) (Fig. 2). With regard to statistical testing, nonparametric analysis of survival time by cause of death demonstrated a statistically significant difference at the P , .0001 level, with subgroup analysis demonstrating a significantly shorter
survival time among pelvic/trunk bleeding compared with intracranial bleedingdmedian survival 159 minutes vs 1,101 minutes (P , .0035). A nonparametric analysis of survival time by cause of death demonstrated statistically significant difference at the P value lass than.001 level, with subgroup analysis demonstrating significantly shorter survival time among trunk/pelvic bleeding compared with intracranial trauma; median survival in the former group was 1,890 minutes vs head injury 3,459 minutes (P , .015). ARDS/MODS patients died after an average of 24,005 minutes (nearly 17 days). Ranked ANCOVA analysis of survival time as a function of ISS with age as a covariate found no statistically significant relationship between ISS and time of death after controlling for age (P , .73), although ranked age and ISS had a significant negative association (correlation coefficient 2.26, P , .03) (Fig. 3). The fracture patterns were consistent with Arbeitsgemeinschaft Osteosynthesefragen/OTA types 61-C2 (30%), 61-C3 (17%), 61-A2 (16%), 61-C1 (14%), 61-B1 (4%), 61-B2 (1%), and 16% were unclassifiable because of massive hemorrhage and emergent operative treatment. Survival time was not found to be significantly different between OTA fracture groups (P , .12). We were able to ascertain the history of blood products given for 65 of the 69 patients included. The average units of packed red blood cells (PRBCs) given to the patients during their stay was noted to be 12.10 6 12.95 (median 8, range 0 to 39) for patients who expired in less than 6 hours.
Figure 2
ISS vs total minutes survival.
The American Journal of Surgery, Vol -, No -, - 2016
4
Etiology Thoracic/Abdominal/Pelvic
Head
Hemorrhage
injury
MODS/ARDS
Mean survival
57.7h
(SD)
31.5h (66.4h)
(64.8h)
16.7d (16.2d)
ISS
33.8
34.1
34.1
Age
52.8
46.9
51.9
Figure 3 A comparison of mean survival time, ISS, and age among the causes of death.
The average units of fresh frozen plasma (FFP) given to the same patients was 3.86 6 4.63 (median 0, range 0 to 12). The average units of platelets (PLT) given in the same patients were 1.00 6 3.39 (median 0, range 0 to 15). In the patients who survived 6 hours to 24 hours, PRBCs given were 10.00 6 8.06 (median 9, range 1 to 23), FFP was 5.67 6 6.58 (median 6, range 0 to 24), and PLT was 4.58 6 5.82 (median 2.50, range 0 to 15). In patients that survived beyond the 24-hour mark PRBCs given were 12.29 6 11.49 (median 9, range 0 to 56), FFP 6.84 6 7.25 (median 4, range 0 to 22), and PLT 5.81 6 11.11 (median 0, range 0 to 45).
Comments Unstable pelvic ring injuries are associated with a high risk of mortality. In the present study a mortality rate of 8% was recorded. After inclusion of pelvic fractures because of gunshot wound, mortality increased to 10.6%, but these patients did not form a part of this cohort. Our mortality rate of 8% falls within the previously published range.1–7 The primary causes of death include abdominal visceral, thoracic, and pelvic bleeding, head injury, and multiorgan failure. Determining the site of hemorrhage is essential for instituting an early and effective treatment. Several studies have reported that nonpelvic bleeding may be a significant factor leading to death. However, no study has examined the timing of death and its correlation with the cause. Results from the present study demonstrate a shorter survival time among abdominal and pelvic bleeding when compared with head injury. This was statistically significant with the median survival time for hemorrhage being 159 minutes vs that of head injury recorded to be 1,101 minutes (P , .0035). Among the 25 patients that died within the first 6 hours of arrival the majority were noted to be bleeding from the pelvis and viscera. Only 2 patients died from isolated pelvic hemorrhage. Previous reports have emphasized the concept of early identification and control of pelvic hemorrhage in decreasing pelvic fracture-related mortality.23–26 Control of pelvic bleeding may be achieved by
increasing pelvic tamponade with the use of pelvic binders and external fixators. Further intervention may be in the form of embolization or pelvic packing. Patients may present with concomitant abdominal injuries making it imperative to differentiate intraperitoneal bleed from a retroperitoneal collection. The former more likely indicates visceral bleeding and the latter denotes bleeding from the pelvic fracture. In a combined situation, laparotomy and pelvic packing have been advocated with simultaneous assessment and treatment of abdominal injuries.23 Among the patients that had an associated abdominal injury the commonest was a liver laceration followed by a splenic laceration. In 2 patients a bladder rupture was noted. The commonest vascular injury involved laceration of the iliac artery with a complete transection seen in 1 patient. Injury of the iliac veins/inferior vena cava and the aorta was observed in equal numbers. Some studies have reported the spleen as being the commonest organ injured in a blunt abdominal injury.27 However, a more recent autopsy-based report indicates liver injuries in 32% patients and splenic injury in 18% after abdominal injury resulting from road traffic accidents.28 Among patients that died between 6 and 24 hours, a majority had sustained head injury which was determined to be the cause of death. There are reports indicating that pelvic fractures with concomitant head and chest injuries contribute significantly to mortality.2,15 In a recent report, patients who sustained concurrent head and chest injuries had higher odds of dying. The authors noted that head injuries increase intracranial pressure and impair the body’s attempt to cope with hypoxia and hypovolemia.29 Another report observed that presence of pelvic fracture in a trauma patient portends a severe injury but is not necessarily the cause of death.30 None of the studies however, correlated the timing of mortality and its cause in patients of pelvic fractures within the first 24 hours. From our study it is clear that death within the first 6 hours of arrival in the emergency department commonly occurs due to hemodynamic instability from bleeding abdominal viscera and pelvic fracture, and deaths in the 6 to 24 -hour time span are commonly caused by an associated head injury. Patients that died after 24 hours in this study deteriorated to a MODS with or without the presence of ARDS. The patients needed intensive care and their time of death varied from days to weeks. They presented with solid organ damage and/or vascular injuries in addition to pelvic injuries requiring a multispecialty approach to management. In this population, the orthopedic surgery team may need to stabilize the pelvis whereas other associated injuries are addressed by the general and vascular surgeons. ISS has a direct relationship with mortality in patients with pelvic fracture, and Lunsjo et al3 observed that a cutoff of 27 to 28 was a good indicator of risk of death. Gansslen et al16 noted that 61.7% of patients with pelvic fracture who expired had multiple injuries in at least 2 different body areas. Baker et al18 delineated that at ISS greater than 50, no patients survived. All patients with ISS greater than 50
R. Vaidya et al.
Causes of mortality from pelvic fractures
5
in their study died within 1 week, and 75% of these patients expired within 1 hour.31 Our mean ISS was 34 6 11.3 (median 35, range 13 to 66), which is consistent with the results found by Lunsjo et al.3 However, in our study the ISS did not correlate with survival time in those who died imminently, early, or late. The most common methods of injury in our patient population were automobile vs pedestrian collision (26) and MVC (22). There has been an increasing incidence of pelvic fractures as a result of MVC in the past 20 years.31 Rowe et al32 relate pelvic fracture in MVC to female sex, lateral force, vehicle incompatibility, and the severity of the collision. Seatbelts and airbags were not found to protect against pelvic fracture in lateral or head-on collisions.32 Adams et al33 observed that 25% of people who sustained an MVC and died had pelvic fractures, and that the chances of pelvic fracture in pedestrians and motorcyclists increased twofold. In addition, there was a correlation between vehicle speed and severity of injury but it was not statistically significant. Associated injuries and previous comorbidities may play a greater role in patients who were hit by vehicles traveling at lower speeds, but further research would need to be performed to evaluate this hypothesis.4 Severity of fracture pattern has been correlated with death in some reports. In a study on pelvic fractures in motor vehicle accidents, the authors recorded a 52% death rate in patients with a Tile type C fracture; it was observed to be 32% and 16% for type B and A respectively.33 In another study, worse fracture patterns were observed in patients who died.2 In the current report, the 61-C fracture pattern comprised 61% of the cohort with 61-A fractures seen in 16% and 61-B in only 5% of the patients. Furthermore, survival time was not related to the injury classification (P , .225), which supports a previous study.11 In relation to age, Baker et al18 determined that patients aged 50 to 69 years had an increased risk of death, and patients greater than 70 years of age had highly increased mortality with lesser injuries. Gabbe et al13 observed that age greater than 65 years has an 8-fold increase in risk factors for death in their study of 348 patients. Our mean patient age was 51 6 20 (median 49, range 16 to 99). In our study, ranked age and ISS had a significant negative association (P , .03), which showed that the older the patient, the lower the ISS. This reinforces the results of other studies.7,18 Control of hemorrhage in patients of pelvic fractures along with replacement of blood products is known to improve survival rates.34 Blood transfusion carries a risk of well-described adverse reactions including multiorgan failure.35 In an acute trauma setting, the amount of blood transfusion required is undetermined, but the requirement in an intensive care setting is based on a hemoglobin level of 7.0 g/dl as the critical level triggering transfusion.36 Furthermore, data from use of blood products in combat zones have shown that mortality drops by 50% with transfusion of PRBCs, PLT, and FFP in the ratio of 1:1:1.37
Other studies report improved survival with PRBC and FFP transfusion in a ratio less than 2:1.34,36 Our ratio for PRBC’s, FFP, PLT was 3/1/0.3 for less than 6 hrs, than 2:1:1 for patients who died after 6 hours. Which exposes our inability to get our trauma patients’ enough FFP, and any PLT within 6 hours. Limitations of this study include the small number of patients available for analysis, and the retrospective nature of the review. Also, transport time from the scene of injury to the emergency department may be variable. This may result in patients with longer transport time and more severe injuries succumbing enroute. Shorter transport times may allow patients to survive till their arrival to the hospital. Strengths of this study are the use of a standardized comprehensive classification system to evaluate fracture patterns that is based on a system shown to be beneficial for pelvic surgery specialists, the consistency of patient care protocols as all patients were definitively treated at the same institution, and the use of autopsy for cause of mortality.38
Conclusions We found that despite the advances made in acute management of the traumatized patient in the emergency department, mortality was unavoidable in 8% of our patients with pelvic fractures. The demonstration of shorter survival time among thoracic, abdominal, and pelvic bleeding when compared with head injury was statistically significant; the median survival time for hemorrhage was 159 minutes vs that of head injury, which was 1,101 minutes (P , .0035). Survival time was not found to be significantly different between OTA fracture groups (P , .12). ISSs do not correlate with survival time in those who died imminently, early, or late. We also conclude that our trauma center’s acute management for closed pelvic fractures follows standardized acceptable guidelines. Patients who present with pelvic fractures in addition to head injuries, solid organ damage, or vascular injuries require a multispecialty approach to management, and in this population, the orthopedic surgery team may need to play a supportive role while the more life-threatening conditions are addressed. Death from isolated pelvic hemorrhage was rare 2/69 patients.
References 1. Chong KH, DeCoster T, Osler T, et al. Pelvic fractures and mortality. Iowa Orthop J 1997;17:110–4. 2. Holstein JH, Culemann U, Pohlemann T. What are predictors of mortality in patients with pelvic fractures? Clin Orthop Relat Res 2012; 470:2090–7. 3. Lunsjo K, Tadros A, Hauggaard A, et al. Associated injuries and not fracture instability predict mortality in pelvic fractures: a prospective study of 100 patients. J Trauma 2007;62:687–91. 4. O’Brien DP, Luchette FA, Pereira SJ, et al. Pelvic fracture in the elderly is associated with increased mortality. Surgery 2002;132:710–4.
6
The American Journal of Surgery, Vol -, No -, - 2016
5. Poole GV, Ward EF, Muakkassa FF, et al. Pelvic fracture from major blunt trauma. Outcome is determined by associated injures. Ann Surg 1991;213:532–8. 6. Sathy A, Starr A, Smith W, et al. The effect of pelvic fracture on mortality after trauma: an analysis of 63,000 trauma patients. J Bone Joint Surg Am 2009;91:2803–10. 7. Sharma OP, Oswanski MF, Rabbi J, et al. Pelvic fracture risk assessment on admission. Am Surg 2008;74:761–6. 8. Balogh Z, King KL, Mackay P, et al. The epidemiology of pelvic ring fractures: a population-based study. J Trauma 2007;63:1066–73. 9. Giannoudis PV, Grotz MR, Tzioupis C, et al. Prevalence of pelvic fractures, associated injuries, and mortality: the United Kingdom perspective. J Trauma 2007;63:875–83. 10. Pfeifer R, Tarkin IS, Rocos B, et al. Patterns of mortality and causes of death in polytrauma patients: has anything changed? Injury 2009;40: 907–11. 11. Pohlemann T, Bosch U, Ga¨nsslen A, et al. The Hannover experience in management of pelvic fractures. Clin Orthop Relat Res 1994;305: 69–80. 12. Soreide K, Kruger AJ, Vardal AL, et al. Epidemiology and contemporary patterns of trauma deaths: changing place, similar pace, older face. World J Surg 2007;31:2092–103. 13. Gabbe BJ, de Steiger R, Esser M, et al. Predictors of mortality following severe pelvic ring fracture: results of a population-based study. Injury 2011;42:985–91. 14. Gustilo RB, Corpuz V, Sherman RE. Epidemiology, mortality and morbidity in multiple trauma patients. Orthopedics 1985;8:1523–8. 15. 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: 981–1000. 16. Ga¨nsslen A, Pohlemann T, Paul C, et al. Epidemiology of pelvic ring injuries. Injury 1996;27(Suppl 1):S-A13–20. 17. Parreira JG, 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. 18. Baker SP, O’Neill B, Haddon Jr W, et al. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 19. American College of Surgeons. Advanced Trauma Life Support Manual. 8th ed. Chicago, IL: American College of Surgeons; 2008. p. 366. 20. Marsh JL, Slongo TF, Agel J, et al. Fracture and dislocation compendium – 2007. J Orthop Trauma 2007;21:S59–67. 21. Stevenson M, Segui-Gomez M, Lescohier I, et al. An overview of the injury severity score and the new injury severity score. Inj Prev 2001; 7:10–3. 22. Kortbeek JB, Al Turki SA, Ali J, et al. Advanced trauma life support, 8th edition, the evidence for change. J Trauma 2008;64:1638–50.
23. Tanizaki Shinsuke, Maeda Shigenobu, Matano Hideyuki, et al. Time to pelvic embolization for hemodynamically unstable pelvic fractures may affect the survival for delays up to 60 min. Injury 2014;45: 738–41. 24. Papakostidis C, Giannoudis PV. Pelvic ring injuries with haemodynamic instability: efficacy of pelvic packing, a systematic review. Injury 2009;40:S53–61. 25. Gruen GS, Leit ME, Gruen RJ, et al. The acute management of hemodynamically unstable multiple trauma patients with pelvic ring fractures. J Trauma 1994;36:706–13. 26. Mucha P, Farnell MB. Analysis of pelvic fracture management. J Trauma 1984;24:379–86. 27. van der Vlies CH, Olthof DC, Gaakeer M, et al. Changing patterns in diagnostic strategies and the treatment of blunt injury to solid abdominal organs. Int J Emerg Med 2011;4:47. 28. Reddy NB, Hanumantha, Madithati P, et al. An epidemiological study on pattern of thoraco-abdominal injuries sustained in fatal road traffic accidents of Bangalore: autopsy-based study. J Emerg Trauma Shock 2014;7:116–20. 29. Ooi Chee Kheong, Goh Hsin K, Tay Seow Y, et al. Patients with pelvic fracture: what factors are associated with mortality? Int J Emerg Med 2010;3:299–304. 30. Cordts Filho Rde M, Parreira JG, Perlingeiro JA, et al. Pelvic fractures as a marker of injury severity in trauma patients. Rev Col Bras Cir 2011;38:310–6. 31. Inaba K, Sharkey PW, Stephen DJG, et al. The increasing incidence of severe pelvic injury in motor vehicle collisions. Injury 2004;35: 759–65. 32. Rowe SA, Sochor MS, Staples KS, et al. Pelvic ring fractures: implications of vehicle design, crash type, and occupant characteristics. Surgery 2004;136:842–7. 33. Adams JE, Davis GG, Alexander CB, et al. Pelvic trauma in rapidly fatal motor vehicle accidents. J Orthop Trauma 2003;17:406–10. 34. Mauffrey C, Cuellar III DO, Pieracci F, et al. Strategies for the management of haemorrhage following pelvic fractures and associated trauma-induced coagulopathy. Bone Joint J 2014;96:1143–54. 35. McIntyre L, Hebert PC, Wells G, et al, Canadian Critical Care Trials Group. Is a restrictive transfusion strategy safe for resuscitated and critically ill trauma patients? J Trauma 2004;57:563–8; discussion 568. 36. Holcomb JB, Wade CE, Michalek JE, et al. Increased plasma and platelet to red blood cell ratios improves outcome in 466 massively transfused civilian trauma patients. Ann Surg 2008;248:447–58. Erratum in: Ann Surg. 2011 Feb;253(2):392. 37. Borgman MA, Spinella PC, Perkins JG, et al. The ratio of blood products transfused affects mortality in patients receiving massive transfusions at a combat support hospital. J Trauma 2007;63:805–13. 38. Koo H, Leveridge M, Thompson C, et al. Interobserver reliability of the Young-Burgess and Tile classification systems for fractures of the pelvic ring. J Orthop Trauma 2008;22:379–84.