PELVIC BLEEDING: MAKING THE DIAGNOSIS

There are four cavities into which a patient can exsanguinate—thorax, abdomen, retroperitoneum, muscle compartments, as well as externally. While bleeding into the mediastinum or into the brain may be life threatening, only a small volume is required to produce symptoms. Thus, exsanguination into these regions is not possible.

The diagnosis of intrathoracic bleeding can be made with a combination of physical exam and a chest x-ray or ultrasound exam. Tension pneumothorax and/or massive hemothorax can be identified rapidly. Muscle compartment bleeding should be identifiable on physical exam. Finally, external blood loss can generally be diagnosed by history. In addition, external blood loss may become apparent for the second time as the patients are resuscitated and blood pressure increases.

The distinction between intra-abdominal and retroperitoneal bleeding can be most difficult. Combined abdominal and pelvic injuries are common. Physical findings such as abdominal pain, distention, or tenderness do not differentiate between intra-abdominal or retroperitoneal bleeding. In addition, physical findings can be quite nonspecific. Patients can bleed a large volume of blood into the abdomen and/or retroperitoneum with minimal physical findings.

In the past, the diagnosis of intra-abdominal injury was generally made by diagnostic peritoneal lavage in patients who were hemodynamically labile. The advent of the focused ultrasound exam however, has revolutionized the early diagnosis of intra-abdominal injury. The focused assessment with sonography for trauma (FAST) is a rapid bedside technique that can make the diagnosis of intraabdominal injury in several minutes. FAST is portable and can be repeated if results are equivocal.

While FAST is rapid, it is nonspecific. It can identify the presence of blood but is not an organ-specific test. Computed tomography (CT) scanning allows imaging of both the intra-abdominal as well as retroperitoneal structures. It can identify blood loss into both compartments. Unfortunately, this is not nearly as rapid as the FAST exam and is of limited utility in patients who are hemodynamically unstable.

The presence of retroperitoneal hemorrhage should be suspected in any patient with a pelvic fracture. A pelvic x-ray is a rapid screening test that should alert the clinician to the possibility of pelvic hemorrhage. This is usually performed as the screening radiograph with a chest x-ray. After blunt trauma, patients can bleed into the retroperitoneum without a pelvic fracture, but this is exceedingly rare. While a pelvic x-ray is a good screening test, it only describes pelvic anatomy in two dimensions and can vastly underestimate the degree of a pelvic bony injury posteriorly.

Initial physical exam of the pelvis can be helpful in determining skeletal stability even before an x-ray is taken. While some advocate rocking the pelvis vigorously, we believe that this is a potentially dangerous maneuver. In patients with skeletally unstable pelvic fractures, this produces excruciating pain. In addition, displacement of the fracture fragment may exacerbate bleeding that had stopped. Instead, we encourage clinicians to gently compress the pelvis inward at the level of the iliac crest. If the pelvis is skeletally stable, it will not give. If there is give in the pelvis, the patient almost certainly has a skeletally unstable pelvic fracture.

It is important to distinguish between patients with skeletally unstable pelvic fractures and patients who are hemodynamically unstable. Skeletal stability describes the bony architecture of the pelvic fracture. Hemodynamic stability describes the patient’s physiologic response. Not all patients with skeletally unstable pelvic fractures are hemodynamically unstable. In addition, patients who have skeletally stable fractures can still lose a substantial amount of blood into their retroperitoneum.

Patients with a pelvic fracture, a positive FAST, and hemodynamic instability are almost certainly best served by an immediate laparotomy. In most patients, the FAST turns positive with 200–300 cc of fluid in the abdomen (Figure 1). While free fluid could certainly be from a relatively minor intra-abdominal injury or a ruptured hollow viscus such as the bladder, diagnostic laparotomy is probably the most rapid and definitive test in patients who are hemodynamically unstable. If minor injury is found and bleeding is thought to be coming from the pelvis, abbreviated laparotomy should be performed and other plans made to control the pelvic bleeding.

Figure 1 Fluid in the abdomen can signal a positive focused assessment with sonography for trauma (FAST), pelvic fracture, and hemodynamic instability. Such patients are almost certainly best served by an immediate laparotomy.

PELVIC FRACTURE CLASSIFICATIONS

A number of classification schemes are available that describe the bony architecture of pelvic fractures. Probably the most commonly used scheme was described by Young and Burgess in 1986 and classifies pelvic fractures by their vector of force (Table 1). Each classification is subdivided to describe the degree of pelvic instability. The authors originally thought that this classification scheme could predict the need for transfusions. While this may actually not be the case, it is quite useful in describing fracture anatomy and guiding initial attempts at hemostasis.

Table 1 Classification of Pelvic Fractures

Data from Young JWR, Brumback RJ, Poka A: Pelvic fractures: value of plain radiography in early assessment and management. Radiology 160:445, 1986.

Lateral compression (LC) pelvic fractures caused by side impact generally occur after T-bone vehicular crashes or car–pedestrian collisions (Figure 2). LC fractures cause an acute shortening of the pelvic diameter. The pelvis does not open but closes down. The pelvic ligaments generally stay intact. Thus, these fractures generally do not bleed. Hemodynamic instability after a lateral compression fracture more likely results from torso injuries such as intra-abdominal bleeding or intrathoracic bleeding. There is a known association with traumatic aortic injury and LC pelvic fractures.

Figure 2 Lateral compression pelvic fracture.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

Anteroposterior compression (AP) fractures generally occur after head-on vehicular crashes or may occur after equestrian injury, typically when patients are thrown from a horse or the horse lands on them (Figure 3). With this mechanism, pelvic diameter widens and the pelvis opens. The injuries can be purely ligamentous if the sacroiliac (SI) joints rupture, even in the absence of significant bony injury. Pelvic vascular injuries are quite common. AP compression fractures have the highest chance of bleeding, and transfusion requirements are the greatest in patients with these fractures.

Figure 3 Anterior/posterior compression fracture.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

Vertical shear (VS) injuries occur when patients land on an outstretched foot, which generally occurs after falling from a height (Figure 4). VS fractures can also occur in motorcycle crashes, particularly if patients are riding with their legs outstretched. In vertical shear fractures, the force is transmitted up the axial skeleton through the posterior pelvis. Posterior fractures and/or ligamentous ruptures are quite common. If there is a complete disruption of both the anterior and posterior elements (Malgaigne fracture), the psoas muscle pulls the hemipelvis cephalad without opposition. This may be seen on a pelvic x-ray. Vertical shear injuries have an intermediate risk of bleeding.

Figure 4 Vertical shear fracture.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

Pelvic fractures do not always occur in pure form. For instance, a pedestrian may be struck obliquely, not truly from the front or side. Thus, a clear classification may not be obvious. The site of bleeding often correlates with pelvic fracture anatomy. AP compression fractures generally bleed from either the pudendal or obturator artery. Vertical shear fractures most often bleed from the superior gluteal artery. If lateral compression fractures bleed, they can bleed from virtually any vascular structure.

The notion that fracture anatomy could predict bleeding has been questioned recently. The group that Wake Forest reported on the use of angiography and embolization for pelvic hemostasis (Table 2). In their series, low grade AP compression fractures required angiography most often. Seventy percent of AP I injury were treated with embolization. This was substantially more common in AP II and about the same as AP III. In addition, the same was true for lateral compression fractures: nearly 70% of LC I fractures were treated without angiography and required embolization, about the same as LC II or LC III fractures.

It is also clear that pelvic fracture bleeding may be different by age group. In 2001, we demonstrated that lateral compression fractures occurred more common in patients over the age of 55. They were two times more likely to get blood than younger patients and required more blood transfusions (7.5 vs. 5 units). These lateral compression fractures were minor, yet still bled substantially. Not surprisingly, overall mortality was higher in the older patients.

While fracture anatomy may be able to predict both the likelihood and location of bleeding, clearly it is far from perfect. The wise clinician will recognize that any patient with a pelvic fracture can bleed. Patients with evidence of ongoing blood loss should obviously have a search for blood loss in other cavities. If none is found, the patient is probably bleeding from the pelvis. Regardless of the pelvic fracture anatomy, action should be taken to obtain hemostasis.

TREATING PELVIC FRACTURE BLEEDING

A number of techniques are available to stop bleeding from the bony pelvis. Bleeding from the pelvis can occur in a number of ways. The vast majority of bleeding is venous and tends to stop over time as the pelvic hematoma tamponades this low-pressure vascular injury. Patients can bleed from fracture fragments themselves as well as smaller pelvic arterial injuries. Larger arteries such as major branches of the hypogastric distribution (pudendal, obturator, or superior gluteal artery) can bleed. Major vascular structures, such as the proximal hypogastric artery or the external or common iliac artery, rarely bleed.

Hemostatic techniques for pelvic fracture bleeding can include external compressive devices, angiographic embolization, or intraoperative control. External compressive devices reduce the pelvic volume by reducing fracture fragments. This usually stops venous and bony bleeding by limiting pelvic volume. In addition, reducing the fracture fragments limits direct bony bleeding. External compression can help “stabilize the clot” that has formed. Stabilizing the fracture fragments limits recurrent fracture motion and should help prevent recurrent bleeding.

A number of devices are available to help achieve external compression. Perhaps the simplest is using a bed sheet. Ideally, the sheet is placed on the stretcher, even before the patient arrives. Alternatively, the patient can be gently lifted and the bed sheet placed underneath the patient. The bed sheet is then crisscrossed across the pelvis anteriorly and tied down (Figure 5). This low-tech technique has proven to be extraordinarily helpful. It can be especially helpful in small community emergency departments (EDs) that have limited resources and may not have sophisticated technology. We generally advise the emergency physician to tie the bed sheet down snugly but not excessively tightly. Patients can then be transported within the hospital or to a higher level of care.

Figure 5 Achieving external compression by wrapping a bed sheet around the patient, crisscrossing anteriorly across the pelvis.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

The military antishock garment (MAST) was originally used as a resuscitative technique in the prehospital phase. The MAST was originally thought to autotransfuse blood from the capacitance vessels of the lower extremities. It is now clear that increases in blood pressure are caused by increases in systemic vascular resistance, not increases in cardiac preload. Use of the MAST has largely been abandoned. The MAST can, however, be an effective pelvic and lower extremity splint, and can keep pelvic fractures reduced similar to the bed sheet. The lower extremity portion must be inflated if the abdominal and pelvic portion is inflated. The MAST has the advantage of being able to set pressure, which obviously is not possible with the bed sheet. Increases in blood pressure may displace hemostatic clots and cause recurrent bleeding. If the pressure on the MAST is set too high, intra-abdominal pressure can go up, causing abdominal compartment syndrome. Patients must be carefully monitored for either potential complication.

The pelvic C clamp is inserted posteriorly to reduce the posterior fracture fragments. This is inserted percutaneously in the ED in many European trauma centers. In most American trauma centers, fluoroscopic guidance in the operating room is used, which limits its effectiveness as a resuscitation tool. The anterior portion of the C clamp can be rotated out of the way to provide access for angiography or laparotomy. Clearly, if the pin is poorly placed, complications can include gastrointestinal perforation or iatrogenic nerve injury.

In the past, external fixation was very commonly used as a compressive device for pelvic hemostasis. External fixation is the most rigid of the external devices and closes the pelvis down definitively (Figure 6). In centers with expertise, external fixators which can be rapidly applied in the ED. In other centers, this is often placed in the operating room similar to the C clamp. In some cases, external fixation may provide definitive fracture fixation. The anterior portion of the frame can be rotated similar to the C clamp to allow access for angiography and/or laparotomy.

Figure 6 External fixation, commonly used in the past as a compressive device for pelvic hemostasis. External fixation is the most rigid of the external devices and closes the pelvis down definitively.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

Use of external fixation during the resuscitative phase requires that these resources be immediately available. External fixation degrades CT images and may interfere with patient motion through the CT gantry. The pins for external fixation are placed in the iliac crest and the frame is then applied anteriorly. In patients with badly displaced posterior element pelvic fractures, the inward motion produced by tightening down the frame could cause further displacement of the posterior fracture fragments and increase blood loss.

Many American trauma centers have abandoned virtually all other external compressive devices and exclusively use a commercially available pelvic binder. The pelvic binder is a Velcro device that applies even direct pressure on the pelvis (Figure 7). The pressure is set by the Velcro and lace system on the anterior portion of the binder. The binders can be applied rapidly and require a minimum of expertise. The binder should be applied across the femoral trochanters, not across the lower abdomen. Correct placement of the binder can limit access to the groins for angiography. If angiography is required, a hole can be cut in the binder or the binder must be placed slightly higher.

Figure 7 The pelvic binder, a Velcro device that applies even direct pressure on the pelvis, used exclusively by many American trauma centers.

Deciding whether external compression would be helpful is a function of pelvic fracture anatomy. A wide open, AP III pelvic fracture is ideally treated with external compression. A lateral compression pelvic fracture where the pelvis has imploded—not exploded—is unlikely to be helped by external compression. Vertical shear injuries may respond in an intermittent manner. In the days of using external fixators, this discussion was quite germane. Was it wise to delay other hemostatic techniques, even for the 20 minutes it took to place an external fixator in the most sophisticated trauma center? The discussion would be even more important in centers where external fixation took a longer time.

The advent of the pelvic binder has obviated many of these discussions. While a pelvic binder is unlikely to help a lateral compression pelvic fracture, it can be applied quickly and almost certainly not exacerbate bleeding. The pelvic binder can act as a pelvic splint and limit fracture fragment motion during transport to the CT scanner, the operating room, or the angiography suite. Thus, we have become quite liberal in the use of pelvic binders, and place it on virtually any patient with a pelvic fracture who is hemodynamically unstable.

Angiographic embolization for pelvic hemostasis has been used for over 30 years. Diagnostic pelvic angiography should be able to identify all sites of pelvic arterial injury. Embolization with Gelfoam, stainless steel coils, or both can be quite effective in achieving pelvic hemostasis. The more liberal the use of angiography, the less the yield will be. Conversely, withholding angiography until a patient is in extremis, increases the yield of angiography but is probably not the wisest patient care. Our indications for angiography are in Table 3. These were developed nearly 20 years ago and are in current use.

Table 3 Indication for Angiography

Data from Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.

A number of angiographic techniques are available. One would like to be as selective as possible with pelvic embolization to limit potential complications such as impotence or distal ischemia such as pelvic or buttock necrosis. The more selective technique, however, requires a greater amount of time and expertise to use. Another option is to use coil blockade. In this technique first described by Sclafani et al., the vascular injury is bridged and coils are placed distal to the injury. The catheter is then withdrawn and the area flooded with gel foam. This achieves hemostasis and prevents distal migration of the Gelfoam pledgets.

Another technique involves blind proximal hypogastric embolization with stainless steel coils. This is intended to reduce perfusion pressure within the hypogastric distribution and allow spontaneous hemostasis to occur. This could produce suboptimal hemostasis, as there is a rich collateral circulation within the pelvic vasculature. Vascular injuries could be fed by collaterals from the contralateral hypogastric or the circumflex vessels. With the proximal hypogastric artery embolized, the angiographic window to achieve hemostasis is now closed if recurrent bleeding occurs.

Despite this, we have had good success with proximal hypogastric embolization. We are not aware of any patient who has had significant ongoing pelvic hemorrhage from collateral flow. The concern that bilateral proximal hypogastric embolization might cause an increase in problems also seems to be untrue even when bilateral embolization is used.

Operative approaches to achieve hemostasis have generally been discouraged. Laparotomy does not allow the surgeon to directly visualize the injured blood vessel. Opening the retroperitoneum releases tamponade and can restart bleeding that had been stopped, especially venous bleeding. The main hypogastric artery is a very short structure. In most patients, it is only several inches long. It then quickly branches into a large number of much smaller vessels that disappear deep into the pelvis. Identifying the injured blood vessel is extraordinarily difficult, and attempts to do this may only make a bad situation worse.

Hypogastric ligation is an option similar to proximal hypogastric embolization. It does carry risks of unroofing the pelvic hematoma, as well as risks of bleeding from collateral flow. In many centers, hypogastric embolization can be accomplished in approximately the same amount of time that it takes to achieve hypogastric ligation. However, if embolization is not available and patients are exsanguinating, hypogastric ligation remains an option.

Intraoperative hypogastric embolization is another option for patients in extremis. The proximal hypogastric artery is isolated with a vessel loop. The surgeon then mixes a slurry of hemostatic agents, which could include fresh frozen plasma and/or small-particulate Gelfoam. The hypogastric artery is accessed with a large Angiocath and blind embolization performed. There is limited experience in the literature with hypogastric embolization. This course should be avoided unless no other options exist.

Some patients almost certainly benefit from an attempt at operative hemostasis. Patients that present in hemorrhagic shock and have a unilateral absence of a femoral pulse likely have injury to either the common or external iliac artery. Typically, these patients have a traumatic hemipelvectomy (Figure 8). An extreme amount of force is necessary to produce this injury, and these patients usually are in refractory shock. A direct operative approach utilizing medial visceral rotation, combined with a direct approach to the iliac arteries is potentially life saving for these patients.

Figure 8 Traumatic internal hemipelvectomy.

Early rigid fracture fixation is ideal, and definitive reduction of the fracture fragments can be quite helpful in achieving hemostasis. However, early open reduction and internal fixation risks torrential blood loss. Percutaneous SI screw fixation can reduce the posterior pelvis with minimal blood loss. These percutaneous screws are threaded under fluoroscopic guidance and rigidly reduce the pelvis. We typically use this technique with AP II or AP III fractures. The patients must be stable enough to undergo the operative procedure. This technique is not recommended for patients in profound shock. Percutaneous SI screws allow for good reduction of the pelvic fracture, reduction of pelvic volume, and clot stabilization. Once the fracture is definitively fixed, patients can be mobilized out of bed, limiting pulmonary complications.

Anterior pelvic fixation can also be accomplished at the time of concomitant trauma laparotomy. In this scenario, patients undergo laparotomy and treatment for intra-abdominal injuries. When the abdominal portion is finished, the incision can be lengthened. The anterior fracture fragments are manually reduced and then an anterior pelvic plate placed. Again, this provides definitive anterior fracture fixation that obviates the need for other techniques such as external fixation or a binder. We have used this technique approximately 20 times, including in patients who have had gastrointestinal injuries and damage control procedures with an open abdomen. Our complication rate is low.

Pelvic hemostasis can also be achieved by pelvic packing. This technique should be reserved for patients in extremis who failed to respond to resuscitation and cannot be stabilized sufficiently for other techniques such as angioembolization.

Rather than approaching the pelvis via a laparotomy, it is wisest to pack via a retroperitoneal approach and approach the patient through a low midline incision. The pelvic hematoma has lifted the abdominal contents up out of the pelvis (Figure 9). The lower abdominal muscles are split in the midline and the pelvic hematoma entered directly. The pelvic hematoma must be completely evacuated and the pelvis then packed. It is important to stay extraperitoneal. When the packs are applied up against the peritoneum, the peritoneum increases the tamponade effect.

Figure 9 Low midline incision to approach pelvic hematoma.

(From Moore EE, Feliciano DV, Mattox KL: Trauma, 5th ed. New York, McGraw-Hill, 2003.)

Virtually any material can be used for pelvic packing, such as towels, gauze, or laparotomy pack. We have favored the use of homemade fibrin bandages. These mesh packs can be constructed by using fibrin sealant and a Vicryl mesh. The mesh can be folded to the desired size (Figures 10 and 11). We generally pack these deep into the pelvis and supplement hemostasis with lap pads or towels.

Figure 10 Pelvic packing material.

Figure 11 Folded pelvic packing material.

Occasionally, this pelvic packing is used in concert with a laparotomy. In that case, we generally use a transverse incision to avoid having the abdominal and retroperitoneal incisions meet. The muscles can split and the pelvic packed in the same manner. We have used this technique in three patients over the past year. Thus, it is a technique that we apply only in desperate circumstances. However, it has been successful in all three cases.

MANAGEMENT OF OPEN PELVIC FRACTURES

Open pelvic fractures represent a distinct problem. They can be the source of torrential external blood loss. As the blood loss is external, there is often not a large component of retroperitoneal hemorrhage. The clinician must first make the diagnosis and then control bleeding.

The diagnosis should be made on physical exam. As patients present supine, it is important to examine the perineum. It may be necessary to do a vaginal and/or rectal exam to appreciate the laceration. The abdomen should then be evaluated using a FAST. Patients with a positive FAST should proceed to laparotomy, and those with a negative FAST should undergo further diagnostic evaluations.

Bleeding can be controlled initially with packing. If the opening to the outside is small, it may be necessary to widen the external opening in order to get the packs up deep inside the pelvis. Definitive hemostasis can then be achieved later with either angiography or operative control. Once hemostasis is achieved, the patient should undergo debridement of nonviable soft tissue and muscle. Fecal diversion should be attempted later in any patient who has perirectal involvement.

It is important to characterize the anatomy of the pelvic fracture. It can generally be accomplished with a combination of physical exam and pelvic x-ray. The pelvic fracture should be reduced using any of the compressive devices discussed previously. The entire team should be alerted, including the operating room, the orthopedic service, and the angiography suite. Even if the bleeding looks venous and/or is controlled with packing, we advocate angiography for these patients.

The patient should go to the operating room for local exploration, repacking, and whatever hemostasis can be achieved via local control. This can be performed as the angiography team is setting up the angiography suite. If there is torrential bleeding when the patient is unpacked and explored, simply repacking the patient is the wisest course.

Exploring the wound in the ED is not recommended. There is neither adequate light nor appropriate instruments for adequate exploration. Patients almost always need to be placed in lithotomy position to get good access to the perineum. It is also unwise to unpack the patients in the ED. This only risks further blood loss.

Definitive hemostasis is generally achieved via a combination of operative exploration and angiography. There is virtxually no priority higher than achieving hemostasis. Early laparotomy is not necessary. Fecal diversion is not an emergency and can be accomplished later once the patient is hemodynamically stable and bleeding has been controlled.

SUMMARY

The multiply-injured patient with pelvic fractures represents one of the greatest challenges to the trauma service. Various techniques are available to evaluate the patient and help achieve hemostasis. Very few institutions in the United States have immediate access to all of these sophisticated techniques. It is important that an institutional algorithm be devised that will allow efficient, rapid, and effective care for these badly injured patients.