As soon as significant pelvic bleeding is suspected, the patient should be resuscitated per routine and a decision should be made for additional diagnostic tests or immediate intervention. The concept of hypotensive resuscitation (ie, allowing a lower than normal blood pressure during the early phases of resuscitation in order to prevent ongoing hemodilution and bleeding) has been adequately established for penetrating trauma.10 However, it is not universally accepted for blunt trauma despite the encouraging reports.11 The coexistence of neurologic injuries, which have been shown to produce worse outcomes in the presence of hypotension, is the main deterrent to allow a low blood pressure in a hemodynamically unstable blunt trauma patient. In a recent prospective randomized study from 19 emergency medical services of the Research Outcomes Consortium, 192 hypotensive trauma patients were assigned in the prehospital setting to receive crystalloid resuscitation in order to maintain a blood pressure above 70 mm Hg (controlled resuscitation group) or above 110 mm Hg (standard resuscitation group). There was no difference in most outcomes but patients with blunt trauma in the controlled resuscitation group had a lower 24-hour mortality (3%) compared to the standard resuscitation group patients (18%). Although we still need phase III studies to evaluate the different resuscitation regimens, this study offers further proof that limiting the initially crystalloid resuscitation does not harm and may benefit hypotensive trauma patients of any type of injury.12 We espouse the principles of hypotensive resuscitation even in blunt trauma and are very cautious with our early resuscitation efforts, rarely using massive crystalloid infusions. If we suspect correctable bleeding sites, we strive to control bleeding as early as possible and then assume full resuscitation. We pay particular attention to substitute lost blood with blood and blood product transfusion rather than acellular fluids, and we decrease the ratio of packed red blood cells to fresh frozen plasma to as close to 1:1 as possible.13 However, it must be noted that after a number of retrospective and prospective uncontrolled studies trumpeted the value of the 1:1 ratio, the only randomized study on this issue failed to show any difference in outcomes. The large, multicenter, National Instituted of Health supported, PROPPR trial randomized 680 severely injured patients to receive a ration of plasma to platelets to red blood cells of 1:1:1 or 1:1:2.14 There was no difference in 24 hours or 30 days mortality between the two groups, nor in the rate of complications. Therefore, one could conclude that it is not the exact 1:1:1 ratio that improves outcomes but the attention to hemostatic resuscitation and the need to enhance coagulation by appropriate infusion of plasma and platelets when major blood loss is predicted. We are not convinced about the effectiveness of recombinant factor VIIa, given that there is no level 1 evidence, supporting the use of this very expensive medication.15 The only prospective randomized study was flawed by excluding early deaths.16 The benefit of patients who received recombinant factor VIIa compared with placebo was modest at best, a reduction of blood transfusions by 2.6 U among blunt trauma patients. There was no benefit among penetrating trauma patients. Therefore, we use recombinant factor VIIa only as rescue therapy in very selected cases, if at all. Tranexamic acid is an antifibrinolytic agent that was shown to prevent all-cause mortality and death from bleeding in a prospective randomized trial of 20,211 trauma patients recruited in 274 hospitals from 40 countries.17 The study has been subjected to criticism, given the relatively small absolute difference of death from bleeding between the tranexamic acid and control groups (4.9% vs 5.7%), which achieved a strong statistical significance (p = 0.0077) due to the very large sample size. In addition, the reverse effect (increase in mortality) was observed when tranexamic acid was given after 3 hours from injury. No US hospital participated in this study.
Unstable pelvic fractures produce bleeding because of ongoing injury to small vessels, as the fractured elements continue to move, and because of the increased volume of the pelvis, as it happens in open-book fractures. Significant bleeding continues unchecked prehospitally and in the emergency department, as the therapeutic choices to counteract these two mechanisms are limited. Pelvic binders address temporarily these two issues by stabilizing the pelvis to stop the movement of the fractured elements and by decreasing the retroperitoneal volume (Fig. 35-6). The former effect can be produced by simply applying a binder with mild to moderate LC of the pelvis. The latter effect is possible only if the device applies significant compression in order to reduce an open pelvis and decrease the volume available for blood to spill. However, significant LC can create the opposite effect, if applied on the wrong type of fracture. For example, a moderately displaced LC fracture can become worse, if excessive compression is applied by a pelvic binder (Fig. 35-7). Under these principles, simple stabilization by a pelvic device is desirable in all unstable fractures but significant compression should only be used in certain fractures, most commonly those of the open-book variety.
A fracture that would be inappropriate for a tight pelvic binder. Displacement with forceful lateral compression by the binder could be exacerbated and result in vascular injury.
The military antishock trousers (MAST) garment became popular in the 1980s after initial reports of improved survival in patients with multiple injuries. However, in 1989 a definitive prospective randomized study found it to be associated with increased mortality.18 The MAST was for the most part abandoned, although some emergency medical systems still use it on patients with pelvic or lower-extremity fractures.
Pelvic binders have been commercialized by different companies along the same principles with only a few differences among them. Typically, a binder consists of a wide belt with a velcro that attaches the two ends of the binder (which can be cut to customize its length according to the patient’s body habitus). On the belt there is a “buckle pulley” mechanism. By pulling the strings the binder tightens and compression increases. The device is radiolucent, which allows radiographic imaging with no artifacts. The evidence on the effectiveness of pelvic binders is poor.19,20 There are three major pitfalls related to its use. First, an inappropriately high placement of the binder can lead to excessive abdominal pressure and minimal pelvic stabilization. It is not uncommon to place the binder too high. Correctly, the binder must be centered around the greater trochanters and not over the iliac spines. Usually, it needs to be gently passed under the patient’s back and then pulled slightly lower and over the buttocks for proper placement. Second, indiscriminate pulling of the strings can lead to greater compression than necessary. The binder is frequently placed before the pelvic fracture is fully characterized by a CT scan or even by plain radiographs. Therefore, the initial step should involve only moderate tightening until the exact type of fracture is diagnosed. The pulley mechanism attached on the binder makes tightening very easy, and with minimal force an enthusiastic operator can squeeze the binder tight, producing on occasions more harm than good. Third, the binder may compromise the viability of skin, subcutaneous tissue, or even muscle if left in place for too long. A general guideline of a maximum 24-hour placement exists but obviously even this may be too long after a tight application of the binder.21 The health care providers should understand that the binder is only an imperfect and temporary tool for bleeding control. Definitive pelvic reduction and cessation of hemorrhage should be planned immediately in order to minimize the need for a binder.
A bed sheet is frequently used as an immediately available and inexpensive way to wrap the pelvis.22 The edges of the sheet are tied together and around a stick, which can be turned to tighten the sheet and apply the desired degree of compression (Fig. 35-8).
A simple sheet wrapped around the pelvis produces inexpensive and adequate reduction of a pelvic fracture. A significant pubic diastasis, as shown in the first image, is reduced by the sheet, as shown in the following image.
External fixation has been popularized as a rapid means of controlling bleeding. In a few institutions this can be accomplished in the emergency room but in most centers the patients are transferred in the operating room. A number of clamps and devices have been used to provide external fixation. The C-clamp was designed for easy placement in the emergency room in the presence of posterior pelvic fracture.23 As opposed to other fixators, it is easy to assemble and apply. Its crossbar rotates around the fixation pins, which are anchored in the cancellous bone in both acetabula. The rotation of the clamp allows other procedures in the abdomen or pelvis to be offered without difficulty. The pins can be placed more anteriorly or posteriorly according to the location of the pelvic fracture and the need to reduce them. It is clearly a temporary method, which needs to be replaced later with either a proper pelvic frame or internal fixation. The C-clamp has been used more frequently in European than American trauma centers, which typically prefer a frame placed in the operating room.
Early external fixation stabilizes the fractured elements, decreases the pelvic volume, and allows clot to form. There is a variety of external fixators. The early systems used small pins and heavy bars, whereas the newer systems are more compact, easy to adjust, and with larger pins. The standard placement of pins is in the superior iliac crest above the superior anterior iliac spine. Lower placement of the pins is also acceptable and can improve the access to the abdominal cavity. In certain designs more than one pin are placed on each side. Pins can be placed by an open or percutaneous technique. All single bar systems require two pins in each hemipelvis, whereas the frames require three pins on each side, except the Pittsburgh system that requires two clusters of two pins in each hemipelvis. In most cases of a true unstable pelvis, external fixators remain a temporary device, which bridges the period to definitive internal fixation. In the supine position, external fixation provides adequate stability. In the standing position the vertical load is usually greater than the capacity of the external fixator to resist these forces. Dislocation of fractured elements can happen, particularly at the sacroiliac complex.24
After placement of the frame, reduction of the pelvic fracture is done by applying opposite forces to the ones that created the fracture. Open-book fractures are corrected by internal rotation of the pins, whereas LC fractures are reduced by external rotation. Vertical shear fractures required skeletal traction by placement of a femoral pin and are the ones least likely to be adequately reduced and stabilized by external fixation.25
If not converted to internal fixation, external fixators usually stay for 6–12 weeks. The most common complication is infection at the pin sites, ranging from mild to severe. Appropriate sterile technique during pin placement and proper care of the pin sites is essential to avoid infection. If the pins become infected or loose, they must be removed and replaced. Other complications are typically associated with placement and include injury to the lateral femoral cutaneous nerve or other neurovascular structures.
In general, it seems that external fixation should be considered in two stages, an early resuscitative and a later definitive stage. In the early stage the fixator is placed to stabilize the fracture and help control the bleeding. At a later stage a decision must be made about the long-term effectiveness of the external frame versus the need to convert to internal fixation. LC fractures are likely to respond to external fixation as the only method, if reduction is satisfactory. Vertical shear fractures are unlikely to be managed without definitive internal fixation.26 Each patient must be carefully assessed to balance the therapeutic choices of fixing the fracture while maintaining hemodynamic stability and inflicting the minimum physiologic insult during the initial critical hours after trauma.
Pelvic ring fracture producing hemodynamic instability is one of the most common indications for angiographic embolization. The ability to control the bleeding by minimally invasive techniques and without the need for an operation, which is routinely unsatisfactory, is very appealing. However, the appeal is hampered by the unavailability of interventional radiology teams around the clock, the poor monitoring available in an angiography suite, and the long times spent on the angiography table. All three reasons have ceased to exist in modern trauma centers.
Interventional radiology teams are now immediately available with short notice in most level 1 trauma centers. In our hospital the team is in house until late at night and expected to be assembled within 40 minutes after that. Monitoring and resuscitation in the angiography suite should be no different in a level 1 trauma center than it is in the operating room. High-rate fluid infusion devices, noninvasive hemodynamic monitoring, mechanical ventilatory support, arterial blood gas assessments, blood transfusions, and aggressive resuscitation efforts should take place during angiography. The surgical team should be present throughout the procedure, exactly as it would be throughout an operation. An anesthesiologist and/or intensivist should be called according to the circumstances. Critical care nurses should participate in tasks with which radiology nurses are not familiar. Emergency angiography for bleeding control is not much different than an emergency laparotomy for the same reason. Shifting the care of the patient from the trauma team to the interventional radiologist during angiography is inappropriate. In our hospital the trauma team, including anesthesiology and critical care nursing when required, remains responsible for the monitoring and resuscitation of the trauma patient in the angiography suite during the emergency interventions for bleeding control. The concept of time spent for embolization is also important and will be discussed later.
The first challenge for the clinician is to identify the correct indications for angiography. Approximately one-fourth of the angiographies performed find no direct or indirect evidence of bleeding,27 and the risk of an unnecessary transfer to the radiology suite could have been avoided in favor of a direct transfer to the intensive care unit. However, there are no controlled studies in the present literature and the precise indications are unknown. In 97 patients with pelvic fractures retrospectively reviewed no factors predicted a positive angiogram with sufficient likelihood.27 Mechanism of trauma, injury severity, hemodynamic presentation, associated injuries, and hemodynamic presentation were similar between patients with and without radiographic evidence of pelvic bleeding. In a later prospective study by the same group, 65 patients with pelvic fractures were included in the study total of 100 consecutive patients evaluated by angiography for bleeding.28 Three independent predictors of bleeding were identified: age older than 55 years, absence of long-bone fractures (indicating that the pelvis was the main source for bleeding), and emergent angiography (indicating that semiacute interventions had a lower likelihood to identify bleeding). The predictive effect of age was confirmed by another prospective observational study.29 Ninety-four percent of patients older than 60 years of age had a positive angiogram as opposed to 52% of younger patients. The authors recommended that angiographic embolization is offered liberally to pelvic fracture patients over 60 years old.
Pelvic fracture pattern is considered a major predictor of bleeding. Traditionally, three types of PI are considered to be associated with hemorrhage: pubic symphysis diastasis of more than 2.5 cm, bilateral superior/inferior pubic rami fractures (butterfly), and posterior fractures (especially of the vertical shear variety).30 However, there is evidence that even anterior fractures can produce bleeding,27 particularly in older patients or those receiving anticoagulants. The presence of contrast extravasation on pelvic CT scan has also been widely used as a predictor of a positive angiogram.31,32 It is suggested that the sensitivity and specificity of a “contrast blush” on CT to identify bleeding that requires embolization is 84% and 85%, respectively, with an overall accuracy of 90%.32 However, our experience has been that the new-generation CT scanners are highly sensitive and—in combination with precise IV contrast infusion protocols—may pick up small bleeds that are potentially self-limited without further intervention. Therefore, the mere presence of contrast extravasation on CT is not an immediate indication for angiography in our institution. We consider contrast extravasation a crucial element of the constellation of symptoms, signs, and findings of PI and consider it in the context of the entire clinical picture. A patient who is hemodynamically labile and has a contrast blush is emergently transferred to the angiography suite. A patient with a contrast blush, who is hemodynamically stable, does not usually receive a preemptive angiogram but is rather placed under close observation. In a study of 296 patients with pelvic fractures, 40% were found to have extravasation on CT but only half eventually required some form of pelvic hemorrhage control.33 Similarly, the size of pelvic hematoma cannot be used as an isolated indication for angiography.34
Interventional radiologists typically seek to identify the precise site of bleeding and control it with coils. This requires subselective intubation of internal iliac artery branches, time, and larger doses of intravenous contrast. For true trauma emergencies the interventional radiologist should be in a different mindset. In alignment with surgical damage control principles, damage control angiography should be offered. The procedure should be rapid, effective, and temporary. Bilateral internal iliac artery embolization embraces these principles. The interventional radiologist does not consume time maneuvering small catheters into small arterial branches. The bleeding is controlled by truncating all the branches of the internal iliac arteries by a temporary agent, such as gelatine sponge particles (Fig. 35-9). There are at least three reasons for performing bilateral internal iliac artery embolization: (1) at the time of embolization, the patient is often in shock and profoundly vasoconstricted. This prevents intravenous contrast extravasation during angiography and offers a misleading impression of bleeding control. Once the patient is resuscitated and vasoconstriction is reversed, bleeding may ensue. Blockage of all the branches of the internal iliac arteries prevents this problem. (2) The pelvic vascular network is so extensive that a bleeder may be fed reversely from the contralateral side. Embolizing only the unilateral internal iliac artery may not offer effective bleeding control (Fig. 35-10). (3) Some bleeders are right in the center and it is very difficult to discern if they are supplied by the right or the left arterial system. Embolizing both is the only way to control the bleeding (Fig. 35-11).
Truncation of all the branches of the internal iliac artery after injection of gelatin particles.
A right pelvic bleed is fed via the extensive pelvic network through the left arterial circulation. In such cases bilateral embolization may be appropriate.
Midline bleeds can be hard to attribute to the left or the right circulation. Bilateral embolization may be appropriate.
The safety of bilateral internal iliac artery embolization has been shown in a study of 30 consecutive patients who received the procedure with no major complications.35 It seems that the extensive vascular supply of the pelvis ensures survival of pelvic tissues and organs during the few days of gelfoam embolization and until the arteries recanalize (Fig. 35-12). Gelfoam pledgets are usually cut to a size not smaller than 2 mm to prevent migration to smaller vessels and allow baseline collateral circulation. Despite isolated reports of serious complications with bilateral embolization for cancer,36,37 such as colon necrosis, perineal wound sepsis, or avascular necrosis of the femoral head, our experience over the last 15 years has been very encouraging and without any significant complications. In a case-matched study of similar male pelvic fracture patients with and without bilateral iliac artery embolization, the incidence of sexual dysfunction 1–2 years after the injury was high but not different between the two groups.38 Although major pelvic fractures affected sexual function, the addition of temporary embolization of both internal iliac arteries did not worsen the outcome. The authors assumed that, if this delicate function was not affected by embolization, it was unlikely that any other pelvic organ would suffer major long-term consequences.
Heavy bilateral internal iliac artery embolization with near complete (but temporary) interruption of the pelvic circulation. The patient did well.
Failure of embolization occurs in approximately 15% of the patients and is typically associated with coagulopathy.39,40 In such patients the thrombogenic potential of the injected gelatin material may not be fully realized and vessels may not be effectively blocked, showing near-full recanalization within only hours of seemingly effective initial embolization. Superselective embolization is associated with a higher risk of rebleeding. Patients who continue to require blood transfusions within 72 hours after embolization should be taken back to the angiography suite, as repeat embolization is typically successful.
Preperitoneal Pelvic Packing
Packing has been an important part of damage control operations for severe abdominal injuries. For the pelvis, packing has been an ineffective method of bleeding control because the incision of the peritoneum to place the packs releases the tamponade. Furthermore, the open funnel that the pelvis presents does not allow for the packs to remain in place and exercise a hemostatic effect by compression; they rather float free back toward the abdomen.
Even if pelvic packing has not been popular in the United States, European trauma surgeons have used it more liberally. Ertel et al41 from Switzerland showed excellent bleeding control after major pelvic fractures using a combination of packing and external fixation by C-clamp. It is not clear which of the two techniques was primarily responsible for the outcomes. Recently, pelvic packing was reintroduced in the United States by the Denver trauma group.42 Their technique of preperitoneal pelvic packing addressed the previous problems of pack displacement and tamponade disturbance. By placing the packs behind the peritoneum through a separate low midline or suprapubic vertical incision, the peritoneal tamponade is not disturbed and the packs cannot float back toward the abdominal cavity. The authors have described the technique, which involves the opening of the fascia, retraction of the recti muscles laterally, and placement of approximately three packs on each side of the bladder deep into the pelvis. Attention is paid at all times to not open the peritoneum. For this reason, if a laparotomy is performed, the incision should be limited to the upper margin of the pelvic hematoma and not below that. Preperitoneal pelvic packing is then performed through a separate incision (Fig. 35-13). In our practice we have found the technique to be lifesaving in patients who are frankly unstable and cannot tolerate transport to the angiography suite. In these patients, the pelvic hematoma is typically very large and more than six packs may be needed to effectively compress all the bleeding sites. One needs to remember that the bleeding typically originates in the posterior pelvis and therefore, the packs should be placed deep into the pelvis and pressed against the sacrum. Given that the preperitoneal pelvis is not an open cavity, it is not an unusual pitfall to put the packs in a relatively superficial plane, failing to produce adequate compression against the bleeding vessels.
Note the Pfannenstiel incision, which is separate and distinct from the midline laparotomy, leading to the open abdomen. An external fixator is also placed to reduce the fracture. Very likely such a patient will also receive angiographic embolization, immediately following the operation.
In a study of 28 patients receiving preperitoneal pelvic packing by the Denver trauma group, 21 (75%) survived. Only 14% of the patients had postoperative angiographic embolization.42 In a similar study by a Norwegian group 13 of 18 patients (72%) with packing survived but postoperative angiographic embolization was used in 80% of the patients.43 In a follow-up study from Denver, 75 patients who received preperitoneal pelvic packing and external fixation were reviewed. Hemorrhage control was successful in most patients and angiographic embolization was offered only in a minority who required additional interventions to control the bleeding.44 One should consider these three procedures (preperitoneal pelvic packing, external fixation, angiographic embolization) complementary rather than competing.45 All of them can be offered on the same patient. We consider angiographic embolization as the preferred method of controlling pelvic bleeding for most patients. However, preperitoneal pelvic packing provides a useful alternative in the following circumstances if (1) there is no angiographic support, as it may happen in non-level 1 trauma centers; (2) there is angiographic support but the team cannot assemble expeditiously; (3) there is profound hemodynamic instability, which makes even mild delays unacceptable and calls for rapid packing in a ready operating room. Following packing, angiographic embolization should still be strongly considered.