12.46 Pelvic ring fractures: assessment, associated injuries, and acute management

Pelvic ring fractures within the developed world are reported as having a yearly incidence of 23 per 100 000 population. Of 23 patients there will be three prehospital deaths, ten high-energy injuries, and ten low-energy injuries. Only high-energy injuries are discussed here. These are potentially life-threatening injuries and their outcome is very dependent on the early management.

The pelvic ring is formed by the sacrum and two innominate bones. The innominate bones are formed from the ilium, ischium, and the pubis. These bones are connected by the triradiate cartilage of the acetabulum which is fused by approximately 16 years of age.

Axial load is transmitted between the inominate bones and the sacrum via the sacroiliac (SI) joints. The SI joint has an irregular surface and a relatively small synovial cavity. There are small rotatory movements during gait. Posteriorly the joint is spanned by the interosseous and posterior SI ligaments that originate from the posterior iliac crests and insert across the entire sacrum. The sacrum effectively hangs between the innominate bones supported by the very strong posterior ligament complex; this is analogous to a suspension bridge (Figure 12.46.1B). The orientation of the SI joints contributes little to the mechanical stability. The majority of stability comes from the posterior SI ligaments. The anterior SI joint ligaments are weak in comparison.

The anterior part of the pelvis is formed by the pubic rami and a symphysis that is composed of hyaline cartilage, fibrocartilage, and ligament.

The pelvic floor is spanned by the sacrotuberous and sacrospinous ligaments (Figure 12.46.1A) and in combination they resist external rotation and shearing forces. The sacrotuberous ligament also resists flexion (sagittal plane rotation) (Figure 12.46.1C). To understand the contribution of these structures it is important to appreciate the orientation of the pelvis in the anatomical position. The muscles of the pelvic floor contribute little to the stability of the pelvic ring.

In single-leg stance the anterior part of the pelvic ring acts as a strut resisting compression. In double-leg stance the anterior part of the pelvic ring and pelvic floor ligaments are under tension.

The L5/S1 disc provides the connection between the pelvic ring and the lumbar spine this is augmented by the iliolumbar and lateral lumbosacral ligaments that connect the transverse process of L5 to the iliac crest and sacral ala, respectively.

When the pelvic ring is disrupted, the abdominal wall helps to limit pelvic volume; this is breached during a laparotomy.

Stability is defined as the ability of the pelvic ring to withstand physiological forces without abnormal deformation. This is a key concept in the management of pelvic injury.

When a force is applied to the pelvis, both the bone and ligaments fail sequentially in a predictable order. As the pelvis functions mechanically as a ring, disruption of one part necessitates disruption elsewhere. However, that disruption is not always complete and mechanical continuity can be maintained by the ligaments. The transition from stable to unstable patterns is dependent on the integrity of both the bone and ligamentous structures that make up the posterior part of the pelvic ring (Box 12.46.1).

A pelvic injury is described as completely unstable when a hemipelvis is unable to resist deformation in any plane. For this to occur there must be a complete disruption of the anterior and posterior pelvis (bone, joint, or ligament) resulting in a mechanical disconnection of the lower limb from the axial skeleton.

Pelvic injuries can also be considered partially or rotationally unstable when there is incomplete disruption of the posterior complex and pelvic floor ligaments. Vertical stability may be maintained, however there is loss of stability in the horizontal plane, e.g. internal or external rotation.

There are a group of stable pelvic ring fracture patterns that do not compromise the mechanical integrity of the pelvic ring. This group also includes include avulsions (e.g. avulsion of rectus femoris with anterior inferior iliac crest) or direct blows (e.g. transverse sacral fractures below the sacrogluteal arch and iliac crest fractures). The management of these injuries will be dealt with in the next chapter (Chapter 12.4712.47).

How the pelvis fails during loading is dependent on many factors such as bone density, rate of loading, contact surfaces, etc. The direction of force, however, is the key factor in determining injury pattern and allows us to understand the residual stability.

Many classifications have been proposed; however, the two dominant concepts are stability and direction of force. The evolution of a classification based on direction of force has culminated in the Young and Burgess classification. The degree of residual stability has been used in the comprehensive classification system. The most recent version has been produced after collaboration between the AO group and the Orthopaedic Trauma Association (OTA) in 2007.

In addition, the Denis classification is commonly used to describe the sacral component of the pelvic fracture.

The Young and Burgess classification system groups injuries according to the direction of force; anteroposterior compression (APC), lateral compression (LC), vertical shear (VS), and combined mechanism (CMI). APC and LC injuries are further subdivided according to severity. This classification has been used to correlate injury patterns with blood loss and associated injury.

The force is applied directly to the pelvis or through the flexed hip, along the femoral axis, in an anteroposterior (AP) plane. Failure occurs initially at the anterior ring through the symphysis or through vertically orientated fractures of the pubic rami. Direct AP blows may result in fracture of all four pubic rami. This can be seen in ‘straddle’ type of injuries where the rest of the pelvic ring and floor remains intact. More commonly with AP loading the hemipelvis externally rotates resulting in diastasis of the pubic symphysis.

An APC I injury occurs when the pelvic floor ligaments are stretched but maintain their mechanical integrity. With further external rotation the ligaments of the pelvic floor and SI joint start to stretch and then fail from anterior to posterior. With displacement of the symphysis in excess of 2.5cm it can be expected that the anterior SI and ligaments of the pelvic floor will have ruptured, resulting in an APC II injury. The integrity of the posterior SI ligament in isolation will prevent vertical displacement and therefore retain stability in this plane. With further external rotation the last structure to fail in this mechanism is the posterior pelvic ligament complex (interosseous followed by posterior SI), resulting in an APC III injury. Although vertical displacement may not be apparent on the diagnostic imaging this injury is unstable in both the horizontal and vertical plane. Pure AP compression injuries may result in external rotation bilaterally.

Lateral compression is the most frequent mechanism (up to 80%) and is seen with side impacts and crush. Failure of the anterior ring tends to occur through transverse orientated fractures of the pubic rami (ipsilateral, contralateral, or bilateral) (Figure 12.46.3). The most frequent method of failure posteriorly in a lateral compression fracture involves an impacted crush of the sacrum. Due to the impaction, intact posterior pelvic ligament complex (that spans the fracture posteriorly) and pelvic floor ligaments LC I injuries retain both vertical and horizontal stability.

In the case of LC II injuries there may be complete disruption of the posterior pelvis either through the SI joint, posterior ilium, or both. SI disruption occurs when the anterior sacrum acts as a pivot and the lateral compressive force produces a complete disruption of the posterior ligament complex. The ‘crescent’ fracture pattern is a fracture/dislocation through the anterior part of the SI joint and the posterior ilium. The ‘crescent’ refers to the posterior superior iliac crest that remains attached to the superior part of the posterior ligament complex. There is no impaction and therefore these injuries are rotationally unstable, the integrity of the pelvic floor ligaments provides ‘relative’ vertical stability. With significant internal rotation of the hemipelvis there is also the potential for superior rotation and subsequent leg length discrepancy.

LC III injuries are associated with more extreme internal rotation with associated external rotation of the contralateral hemipelvis and increased instability. This injury often results from a roll-over type mechanism rather than falls and lateral impact motor vehicle collisions that are seen commonly in LC I and II injuries. The pattern of associated injuries is also different with less head, chest, and upper abdominal injury.

Lateral compression can also cause isolated extra-articular fractures of the ilium (Duverney fractures).

Axial loading, directly or through an extended leg, results in superior displacement of the hemipelvis and indicates complete disruption of both the anterior (pubic symphysis or vertically orientated pubic rami fractures) and posterior pelvic ring (sacrum, SI joint, or ilium) with rupture of the (pelvic floor) sacrospinous and sacrotuberous ligaments, (Figure 12.46.4). This injury mechanism is usually due to a fall from height or motor vehicle collisions and results in complete instability of the hemipelvis. Vertical displacement is usually apparent on the original x-rays.

The direction of force may not be readily apparent in some circumstances either due to a combination of sequential forces (e.g. ejection from vehicle followed by landing) or resultant force acting in a different plane. Common patterns include lateral compression/vertical shear and AP/vertical shear.

The comprehensive classification is an alpha-numerical code used for coding throughout the body. In relation to pelvic injuries they are categorized according to residual stability of the pelvic ring. The code for pelvic ring is 61 and subdivided into A, B, and C. Group A represents stable pelvic ring injuries. Group B represents partially unstable injuries and Group C represents complete instability. Subclassification further describes the injury.

Details of the classification can be found at: http://www.ota.org/compendium/compendium.html

The Denis classification describes the relationship of the sacral fracture to the sacral foramen and has been correlated with neurologic injury (Figure 12.46.5). Zone 1 injuries are defined as fractures restricted to the sacral ala lateral to the sacral foramina. Zone II injuries involve the sacral foramina but do not extend medially into the sacral canal. Zone II injuries may also result in instability at the lumbosacral junction. This is dependent on the relationship of the vertical fracture line with the L5/S1 facet joint. Fractures running lateral to the joint are not associated with instability, however fractures running into or medial to the facet joint have the potential for instability at L5/S1 as the facet joint is either defunctioned or separated from the axial skeleton.

Any fracture that extends into the sacral canal is defined as Zone III. Zone III fractures also include pure transverse fractures and complex fractures with a transverse component (e.g. ‘U’, ‘Y’, ‘H’ and ‘Λ’ patterns) (Figure 12.46.6). These complex fractures result in complete discontinuity between the spine and pelvis.

Pelvic fracture should be considered a marker for high-energy injuries and associated injuries are very common. All of these associated injuries can have a significant impact on morbidity and mortality.

In addition to the bleeding from fracture surfaces, significant blood loss may arise from venous and arterial vascular injury.

Venous bleeding is much commoner than arterial bleeding. Significant arterial bleeding source is identifiable in 4–15% of cases, and is commoner with more severe injuries of the pelvis. The superior gluteal is the most frequently injured in unstable posterior injuries and the pudendal and obturator vessels are the most commonly injured vessels as a result of the anterior ring injuries.

The frequency of associated injury to the bladder and urethra is reported at 7–25%. Injuries to the urinary tract are frequently multifocal and the overall frequency and anatomic distribution of urinary tract injury is similar in both APC and LC injuries. Although the fracture pattern is not predictive of urinary tract injury, the type of injury is related to the pelvic fracture pattern.

Bladder injury has been reported with a frequency of 9–16%, of these 60% are extra-peritoneal, 30% intraperitoneal, and the remainder are a combination of both. Extraperitoneal injuries result from a shearing force and commonly involve the anterolateral and base of bladder. Intraperitoneal ruptures tend to occur at the dome as the result of blunt injury to a distended bladder. Due to the high energy involved, pelvic fractures associated with bladder rupture have increased mortality rates. Perforations can also occur directly from the fractured bone ends.

Partial and complete urethral injury (Figure 12.46.7) is seen in approximately 10–22% of males and usually involves the posterior part of the urethra, below the level of the sphincter. Female urethral injury is rare and is often secondary to laceration from fracture fragments. Injury involving the bladder sphincter can result in incontinence.

The frequency of associated abdominal injury is 15–26%, and in unstable pelvic fractures the frequency of abdominal injury is up to 55%. Although the incidence of abdominal injury is high, only a proportion will require laparotomy, this can create difficulty in identifying, prioritizing and managing ongoing blood loss. Pelvic fracture with a liver injury requiring packing has been associated with a 60% mortality rate.

The rectum is held within the muscular pelvic floor, and when the pelvic floor is torn it is at risk of injury.

Abdominal compartment syndrome should be considered in patients with pelvic and abdominal injury, this may be secondary to visceral oedema or ongoing bleeding resulting in a tense abdomen, respiratory compromise and renal dysfunction.

Approximately 20% of pelvic fractures will have a thoracic injury greater than AIS (Abbreviated Injury Scale) 2.

Pelvic fractures are frequently associated with central and peripheral neurological injury. In the United Kingdom, AIS scores higher than 2 for head injuries have been identified in 17% of cases. Head injury is particularly common in lateral compression injuries. This has been related to frequently occurring motor vehicle side impacts. Head injury has a significant influence on mortality and outcome.

Up to 25% of high-energy pelvic fractures will have an associated spinal injury.

Peripheral nerves are also susceptible to injury. The anterior lumbar, sacral, and coccygeal rami form nerves which traverse the pelvis providing motor and sensory function to the pelvic viscera, perineum, and lower limbs. Neurological deficit may be subtle and has been reported in up to 21%. The sacral nerve roots and lumbosacral trunk are at risk from posterior pelvic ring disruption either as a result of being crushed within sacrum or stretched due to pelvic or sacral displacement. The location of the sacral injury influences the risk of neurologic deficit: zone I, 5.9%; zone II, 28%; and zone III, 57% (within this group, bowel, bladder, and sexual dysfunction occurs in 76%). In transverse sacral fractures there is a higher incidence of bladder dysfunction at the S1–3 level compared with S4 and below.

Assessment of a pelvic injury should be conducted along ATLS^® principles. With appropriately trained personnel many of the steps can be performed simultaneously rather than sequentially.

The following elements of the assessment are particularly pertinent to pelvic trauma:

The appropriateness and urgency of further assessment and management is dependent on the haemodynamic stability.

There is controversy regarding optimum management. However, there is evidence and consensus that the outcome is improved if the treatment plan is protocol driven. Optimum management will be dependent on the institution and is best considered as a series of standard steps:

Quick, minor interventions can be life saving during the early phase as the majority of pelvic fractures will be amenable to simple interventions performed efficiently. Ongoing bleeding that does not stop with splintage and resuscitation will require surgical or angiographic control and this will take time to organize.

An early decision should be made with regard to transfer. This a contentious issue which requires senior input and a common sense approach. Transfer to an appropriate facility has been demonstrated to increase the probability of survival by 30%.

This should start in the prehospital environment and should influence the choice of appropriate receiving facility, avoiding medical ‘speed bumps’. Many haemodynamically compromised patients will have achieved haemostasis by the time of presentation and will respond appropriately to resuscitation. The group that requires early identification are those patients who are haemodynamically unstable due to ongoing active bleeding. They will present as transient or non-responders to fluid resuscitation and are more likely to require early intervention. Identification of the ‘sick patient’ is the priority as this sets the tempo, direction and location of further aggressive management. 60% of deaths with pelvic fractures occur in the prehospital environment.

A post injury median survival time of 55min has been reported in pelvic fracture fatalities (54% <1h; 11% 1–2h; 16% 2–6^th h). In general the primary survey (including diagnostic peritoneal lavage (DPL) or focused assessment with sonography for trauma (FAST)), and temporary stabilization of the pelvis should be achieved within 15min and the decision with regard to requirements for haemostasis within 20–30min.

Using physiological signs and readily available investigation will allow the correct management decisions to be made in a timely fashion. Pulse, blood pressure (BP), and respiratory rate allow an estimation of blood loss. The response to fluid resuscitation is also useful to determine adequacy of resuscitation and ongoing losses.

Lactate, Ph, and base deficit (BD) are useful indicators of hypovolaemic shock and when monitored will also reflect the response to treatment. Retrospective clinical review has identified BD as a better prognostic indicator than pH and is a sensitive marker for injury severity, blood transfusion requirements within the first 24h, incidence of multiorgan failure (MOF) and death. Patients with a BD higher than 5mmol/L should be considered at risk of having a significant bleed. Lactate levels are also reliable indicators of morbidity and mortality in pelvic trauma. One study demonstrated that survivors had significantly lower initial lactate levels (4.2 ± 1.8mmol/L) in comparison to those early deaths within a few hours (8.6 ± 2.5mmol/L).

The presence of coagulopathy early in the presentation of a trauma patient is a poor prognostic sign. Prothrombin time (PT) and partial thromboplastin time (PTT) are independent predictors of mortality. Initial abnormal PT increases the adjusted odds of death by 35% and an initial abnormal PTT by 326%.

‘Permissive hypotensive’ or ‘balanced’ resuscitation, maintaining a systolic BP of 80–100mmHg and haemoglobin of 7–9g/dL, is appropriate in the initial management to minimize blood loss and the risk of rebleeding. This must only be used as a temporizing measure until the source of bleeding is controlled. This is not appropriate in patients with head injury or spinal cord injury and care should be used in the elderly patient. Under these circumstances it is important to maintain perfusion pressure.

To help control bleeding, a pelvic binder/sheet should be considered and open pelvic wounds should be packed.

The choice of resuscitation fluid is still an area of ongoing development; however, there is increasing evidence to suggest that early resuscitation with blood products is associated with improved outcome. The use of packed red cells (PRC) and fresh frozen plasma (FFP) in a 1:1 ratio is favoured by many trauma centres and the military.

Consideration must also be given to prevention and correction of coagulopathy. Platelets should be administered to maintain a platelet count in excess of 100 000/µL. Clotting can be further promoted with tranexamic acid and Factor VIIa. Factor VIIa is most beneficial when used before the patient becomes acidotic. Fibrinogen levels should be maintained >1g/L with cryoprecipitate.

In addition, it is also important to address hypothermia. A 1° decrease in core temperature is associated with a 10% decrease in coagulation function. All intravenous fluid must be warmed.

In polytrauma, having identified a patient at risk, it may not be clear what contribution the pelvic fracture has made to the haemodynamic status. Accurate assessment is essential as it avoids unnecessary time wasting procedures. Significant bleeding from limb fractures is easily identified through clinical and radiological examination; blood loss volumes can be estimated. Bleeding from open injuries is less easily quantified. The majority of thoracic bleeding can be excluded with the presence of a normal chest x-ray. One of the biggest dilemmas is differentiating between ongoing pelvic and abdominal bleeding. DPL, FAST, and computed tomo graphy (CT) can help to determine the source of bleeding but they all have limitations.

The recorded sensitivity of DPL, FAST, and CT for detecting intra-abdominal injury is 1.0, 0.92, and 0.97 respectively. However, it is the lack of specificity that results in negative laparotomies, reported in one-third of cases. This has the potential to delay treatment of a pelvic bleed.

CT has high sensitivity and specificity but requires the patient to be transported to the CT scan. Emergency CT is extremely valuable in all high-energy pelvic injuries as in addition to information on the fracture pattern it allows the identification of visceral and spinal injury. In addition, CT can identify pelvic haematomas and contrast extravasation. Pelvic haematoma volumes higher than 500mL are associated with a 0.45 probability of ongoing pelvic arterial bleed. Contrast extravasation can indicate ongoing arterial bleeding demonstrable at angiography (positive predictive values 45–80% and negative predictive values of 85–99.6%). Although the negative predictive value is extremely high, a small number of these cases will require embolization and therefore the clinical situation must be monitored.

Although there is no reason for anything other than brief interruptions to resuscitation and monitoring, CT scanning may delay the necessary time critical intervention to stop the bleeding, and this must be balanced against the patient’s physiology. If the patient’s haemodynamic situation requires early intervention and does not permit CT, then DPL and FAST can be used to help in the decision-making and treatment priorities. However, they have greater limitations with regard to specificity.

FAST depends on availability and technical expertise and in general it has a reported sensitivity of 0.92 for detecting intraperitoneal fluid. However, in the context of a patient who is hypotensive (<90mmHg) from abdominal bleeding the sensitivity has been reported as 1.0. In practical terms if a patient is hypotensive, secondary to bleeding in the abdomen, and the FAST scan is negative an extra-abdominal source of bleeding should be sought.

If neither FAST nor CT are available then DPL can be considered, it has been reported as having a sensitivity of 1.0 with poor specificity (false positive rates for DPL have been reported as high as 29%). The accuracy can be improved with supraumbilical placement and good technique. In a hypotensive patient unless gross bleeding is identified on the DPL, bleeding is likely to be originating from an extra-abdominal source.

Treatment should be directed at promoting and maintaining clot formation. The quicker a haemostatic technique is adopted the more likely it is to work. All interventions directed towards controlling pelvic bleeding will fail unless coagulopathy is aggressively corrected.

In addition to the correction of coagulopathy clot formation is enhanced by reducing the pelvis. Reducing the pelvis reduces pelvic volume, increases stability, and apposes pro-thrombotic surfaces. This can be achieved by using non-invasive or invasive external compression devices. Stabilization of the pelvis has been proven to be clinically effective in reducing mortality rates and transfusion requirements. Once stabilization has been achieved the patient should be moved with caution to avoid clot disruption.

This includes pelvic binders, bean bags, and sheets that can be applied to generate non-invasive circumferential pressure (Figure 12.46.8). The principal aim is to minimize movement and clot disruption at the fracture site. To further supplement stability the lower extremities can be held together with padded bandages applied to the thighs and ankles. These devices need to be used correctly (centred on the greater trochanters), with care to avoid neurological, vascular, visceral, and skin injury through over-reduction. They are quick to apply in the prehospital or resuscitation setting and avoid the problems of pin sepsis. It is not advisable to maintain them for long periods of time due to concerns about skin breakdown. The benefit of this technique has yet to be fully quantified and the limited studies available are not conclusive.

External fixation pins can be placed in the ilium through both the iliac crest, (Figure 12.46.9) and supra-acetabular region. The latter has biomechanical advantages but is technically more challenging to apply in the emergent situation. Application through the iliac crests is a familiar technique that can be performed quickly (in the emergency department, operating theatre or angiogram suite) without image intensification. Anterior external fixation has been associated with poor reduction and stability of the posterior ring. Any form of external fixation can be complicated by pin site colonization and sepsis. Ideally pin placement should avoid areas of degloving and if possible the position of future surgical incisions; however, these are secondary considerations in the emergency setting (discussed further in Chapter 12.47).

A pelvic c-clamp can be used for emergency treatment of posterior ring disruption. Pins are applied against the outer table of the posterior ilium, overlying the SI joints. The entry point for these pins is at the intersection of the femoral axis and a vertical line running through the anterior superior iliac spine. Posterior external fixation has the potential to reduce and stabilize the posterior ring more effectively than anterior fixation. The c-clamp must be applied with care to avoid placement of the pins in the sciatic notch and is contraindicated in iliac fractures. Over compression or pin perforation risks vascular, visceral, and neurological injury.

Certain fracture patterns and situations may warrant consideration of early internal fixation. In the presence of an open book injury undergoing a laparotomy it would be appropriate to consider plating the pubic symphysis. Posterior injuries could be considered for percutaneous SI screw placement; however, this is technically demanding and should not be undertaken if there is a risk of incurring an unacceptable time delay. Both of these options need to be considered before draping.

Despite resuscitation, correction of coagulopathy and pelvic stabilization a small proportion of pelvic fractures will have persistent hypotension due to ongoing arterial and venous bleeding. They will require rapid direct haemostatic intervention for survival. In extremis, temporary occlusion of the aorta can be considered by clamping the aorta or insertion of a balloon catheter. Ligation of the internal iliac is of limited benefit due to the retroperitoneal haematoma and considerable anastomoses linking the right and left sides of the arterial and venous circulation.

This procedure involves tamponade of the pre-sacral and paravesical vessels by placing packs into the extraperitoneal space of the true pelvis. This is achieved through a low midline incision. Once the extraperitoneal space is accessed little additional dissection is required due to the injury and the expanding haematoma. Large swabs are placed, from posterior to anterior, either side of the pelvic viscera. For maximum benefit it has been recommended that EPP should be performed within 30min. In those amenable fracture patterns, external fixation is added in conjunction with EPP. The packs are usually removed at second look surgery within 48h.

Routine post-EPP angiography has demonstrated an 80% incidence of ongoing arterial bleed. However, in another study, angiography was only performed in 17% when it was clinically indicated.

Packing has been shown to have a significant effect on haemodynamic status, reduced postpacking blood transfusion, and is suggestive of improved survival rates. It also allows other sources of bleeding to be dealt with by surgical intervention in the same location. It can be performed quickly during the time required to set up for angiography.

This technique requires the immediate availability of an operating theatre and necessitates an open procedure that has implications with regard to infection and abdominal compartment syndrome. It is likely to be less effective with bleeding from larger arteries and under these circumstances may only function as a temporizing measure to limit blood loss. Packing is associated with a 25–29% mortality rate.

AE is an attractive technique for controlling pelvic arterial bleeding with an 85–100% success rate. However, it does not limit venous bleeding. The angiography suite should be equipped and staffed to allow optimum continuation of all the components of resuscitation. In extremis a balloon catheter can be rapidly inserted and used to occlude the aorta. A catheter is passed through the femoral artery at the groin or the left axillary artery if the groin is not accessible. Angiography can also look for sources in the chest and abdomen. Large vessel injury including the thoracic aorta is present in approximately 6% of those requiring embolization of pelvic arteries. Selective embolization techniques can be performed, however non-selective bilateral embolization procedures, including embolization of the internal iliacs, has been reported with good outcome. In some hospitals it is feasible to perform angiography in the operating room.

Review of the literature would suggest only 5% of all pelvic fractures are likely to benefit from embolization. However, there are subgroups that have been identified with an increased incidence of ongoing arterial bleeding: contrast extravasation and large haematoma volumes on CT; hypotensive patients (<90mmHg) who are transient or non-responders to resuscitation have a 44–76% incidence; BD of more than 6mmol/L (when used as a trigger it identified arterial bleeding in 100% of cases); over 60 years of age (over two-thirds of this group of patients will present with normal vital signs). Venography has been recommended if there is no improvement following AE as significant venous injuries have been identified. It is also appropriate to consider placement of an inferior vena cava filter (discussed further in Chapter 12.4712.47).

If hypotension or a BD persists following any intervention, repeat angiography should be considered. Re-embolization has been required in 6–7.5% of reported cases.

The technique is both resource and time consuming, even in expert hands it requires time. It should be expected that AE will take an average of 90min, in addition to the transport and set-up time; however, this is very dependent on the institution. Successful embolization has not been seen to always result in an immediate improvement in haemodynamic status and reduced requirements for blood products. Not all identifiable small arterial bleeds (seen on CT and angiography) will require embolization. These factors have yet to be fully quantified. The extent of ischaemic complications related to AE is believed to be less with selective embolization techniques. However it is difficult to separate the effect of AE from the injury. Risks include:

The cumulative reported mortality rate is 43%.

When the patient is haemodynamically stable, with or without intervention, further assessment is required prior to definitive management. Further imaging is required to define the injury. First-line investigation in most centres still involves inlet and outlet views. The inlet view is taken with the patient supine; with a 40-degree cranial tilt of the beam so that it is perpendicular to the pelvic brim. The beam is then rotated to a 45-degree tilt towards the feet to achieve the outlet view (Figure 12.46.11). In some circumstances obturator and iliac oblique films are useful to define the fracture pattern and exclude acetabular involvement. CT (1–2-mm cuts) is a more sensitive tool for identifying occult injury and with improving technology (e.g. three-dimensional reconstruction) and access, the requirement for inlet and outlet views will become more dependent on surgeon preference.

Under some circumstances, to determine stability, examination under anaesthesia and stress views can be used. Dynamic assessment is contraindicated in haemodynamic instability (wait 3–5 days), lumbosacral plexus injury, ipsilateral vessel injury, and lower limb fractures that prevent axial loading of the hemipelvis.

Provisional reduction and stability is important for pain relief and to facilitate definitive surgery. This can be achieved using the methods previously discussed with or without skeletal traction (skin traction is inadequate) to reduce vertical displacement. If external fixators are considered, application should be discussed with the surgeons performing the definitive fixation.

A history of inability to void, blood at the urethral meatus, haematuria, high riding prostate, and inability to catheterize are suggestive of genitourinary injury. In the presence of these findings urethral catheterization is contraindicated and requires further investigation. Retrograde urethrogram can be performed through a Foley catheter with the balloon sited in the urethral meatus. The presence of a urethral injury demands a urological opinion and may necessitate a suprapubic catheter. Potential bladder ruptures are diagnosed with a retrograde cystogram. The bladder is then emptied and the x-rays are repeated if necessary. Small extraperitoneal bladder ruptures are treated with suprapubic catheter drainage alone, to prevent the accumulation of urine and the risk of sepsis. The majority of tears will be healed by 10 days, the remainder within 4 weeks. Large tears and intraperitoneal tears require surgical repair.

Male urethral tears can be managed with early realignment with a urethral catheter maintained for 4–6 weeks or urinary diversion using a suprapubic catheter for 3–6 months followed by reconstruction. Urethral tears are frequently complicated by stricture formation. In females with proximal urethral tears, early urethral realignment or reconstruction is recommended.

Surgical correction of impotence should be delayed 12–18 months due to delayed return in function.

A vaginal laceration in association with a pelvic fracture should be considered an open fracture. All vaginal injuries should be explored and debrided under a general anaesthetic.

In addition to the associated visceral injury seen with polytrauma involving pelvic fractures it is important to identify rectal injuries that may be in continuity with the pelvic fracture. These fractures should be considered as open due to the contamination of the fracture site by faecal organisms. The only evidence of this injury may be fresh rectal blood identified at PR. Rectal tears may not always be identifiable using standard investigations (proctoscopy, sigmoidoscopy, and contrast studies). However, the consequences of a missed rectal injury are serious. Therefore when there is a high index of suspicion, the general surgeons should be encouraged to perform a complete defunctioning colostomy with distal washout. A double barrelled colostomy is not acceptable as it does not completely divert the faecal stream. The colostomy should be placed sufficiently high on the abdominal wall to avoid interference/contamination of future pelvic surgery. This can be reversed at a later date.

In the haemodynamic unstable patient with pelvic and abdominal bleeding, damage control principles may be appropriate with minimal surgical intervention, packing of the abdomen and pelvis and application of an external fixator. These patients should be monitored for abdominal compartment syndrome. Further pelvic or intra abdominal bleeding may be amenable to AE.

Open fractures including those involving the rectum and vagina constitute 2–4% of all pelvic fractures. The mortality rate for open pelvic fractures is greater than 50%. In an open fracture there is less potential for tamponade and blood loss is limitless. In addition there are the problems with deep sepsis. Early studies (pre-1992) report infection rates of approximately 30%, compared with a 15% incidence in the modern literature. Although the reported incidence of infection has improved the mortality rate of open fractures remains high, 67%.

This problem should be managed, at presentation, according to the guidelines of all open fractures with early antibiotics, tetanus, and wound dressing followed by debridement, stabilization, and early soft tissue coverage. Defunctioning colostomy is required in those patients with an open injury involving the rectum or significant perineal wounds that are at risk of contamination.

Closed degloving injuries (Morel-Lavallée lesions) are associated with necrotic fat, fluid collection and bacterial colonization. These injuries may involve the planned surgical field and must be treated before fixation. Open and percutaneous techniques have been used successfully at both the same operation and as a separate procedure.

Pelvic fracture should be considered a marker of severity and as demonstrated earlier in the chapter there are significant associated injuries. Many of these associated injuries will be fatal. The incidence of pelvic fracture as the sole cause of death (based on postmortem studies) is very low, 3.7%, and approximately 50% of pelvic fracture deaths will have other potential causes of haemorrhage. In general the pelvic fracture is felt to contribute evenly to both early and late deaths.

Generalized inpatient mortality rates, from large studies in the modern literature, are quoted at 3–20% (cumulative data 12.4%). These figures are dependent on the cohort, usually reflecting the referral pattern for the institution, and care should be taken when comparing figures. These figures are of little value in the clinical setting as the true lethality of this injury can be under appreciated. Haemodynamically unstable fractures have a much higher mortality rate, 40–60%.

In the polytrauma patient it is the cumulative physiological burden of the multiple injuries that often contributes to death. An ISS higher than 25 has been determined as an independent determinant of mortality. The level of physiological burden that can be survived is dependent on age with increased mortality rates seen in the over 40 age group. Pelvic injuries occurring in those over 65 years is an independent determinant of mortality.

Pelvic fracture in pregnancy is rare but fetal mortality rates of 35–80% have been reported; many of these deaths occur prehospital.

An understanding of the pathoanatomy of pelvic fractures will allow accurate classification and appropriate treatment. A significant proportion of pelvic fractures will be at additional risk of morbidity and mortality from extravasation and associated injuries. The management of these patients is an area of evolution. However, the treatment principles will remain the same and there will always be a requirement for protocol driven treatment to minimize delay. The indications and thresholds for current and future treatment modalities will need to be refined and ideally this will be based on assessment tools that are practical and can be used with minimal delay in the emergency department.