1.18 Complications of pelvic trauma
1 Introduction
“It has always been said that major surgery creates major complications, and any surgeon willing to undertake this type of surgery must be willing to look after these complications” [1].
Injury to the pelvic bones and associated soft tissues remains a significant contributing cause of morbidity and mortality in patients who have sustained an injury. Achieving an optimal treatment outcome requires a detailed understanding of the potential complications of injury and treatment. Appropriate early management of pelvic ring injuries and application of the principles detailed throughout this chapter can limit the resulting morbidity and mortality associated with these complex injuries. Education of the entire treatment team, patient, and patient′s family is a necessary component of any reliably successful treatment regimen. The orthopedic trauma surgeon must lead this endeavor. In so doing he/she will often be required to coordinate care among various team members while balancing competing concerns with long-term functional outcome and acute care needs. This chapter considers the prevention and treatment of the short- and long-term complications associated with injuries to the pelvic ring.
2 Early complications
2.1 Mortality
The mortality rate in patients sustaining pelvic fractures remains high. Current data from major trauma centers suggests an overall mortality ranging from 5–10%. A review of 10 years’ experience at one center in the United States revealed overall mortality of 6%, with an incidence of 19% following open fractures [2]. Parkland Hospital in Dallas, Texas, demonstrated a 9.1% overall mortality in 1,248 patients presenting with lateral compression injuries over a 7-year period. The mortality following treatment of > 22,000 pelvic fractures at R Adams Cowley Shock Trauma Center over a 10-year period was also 9% [3]. In the report from Parkland [3], it was clear that unstable patterns of injury (lateral compression type II and type III, AP compression type II and type III, vertical shear and combined mechanism) were associated with a higher mortality than stable patterns (11.5% versus 7.9%). A trend toward higher mortality rate for lateral compression type III, AP compression type II and type III injuries was identified (13%, 12%, and 16%, respectively) [4]. Findings taken from The German Pelvic Trauma Registry Initiative [5] are consistent and reveal a 5% mortality rate for all fractures with a significantly higher mortality of 18% following complex pelvic injuries, defined as “unstable fractures with severe peripelvic soft tissue and organ laceration.” Vascular injury, injury to the enveloping soft tissues, and infection are potential causes of mortality and must be optimally managed if complications, including death, are to be avoided.
2.2 Vascular injury
The venous system is the most common source of hemorrhage following pelvic fracture (see Chapter 1.4). Mechanical stabilization with anterior external fixation, placement of pelvic C-clamp, or placement of pelvic binder or sheet have been shown to be effective modalities for controlling low-pressure hemorrhage. Proposed mechanisms include decrease in pelvic volume leading to tamponade, and immobilization of bleeding surfaces and vessels allowing for clot formation. Some institutions [6] have included pelvic packing as a protocol to control early hemorrhage, also with reported success. A significant percentage of patients who fail early resuscitative efforts, including such intervention, do so as a result of specific arterial bleeding that can be effectively treated using transcatheter arterial embolization (TAE) [7], which may be used selectively or for damage control. Selective embolization is a targeted approach to terminal vessels responsible for hemorrhage. A partial list of bleeding vessels that are reported to have undergone successful TAE includes corona mortis [8–10], superior gluteal artery, internal iliac, [11] external iliac (including one case of a dissection treated with stent placement), [12] common iliac artery [13] sacral arteries, and pudendal arteries [14]. Though often effective, repeated embolization is necessary in 10–22% of cases [15, 16]. The need for repeated embolization may be the result of an initial false-negative arteriogram occurring as a result of vasospasm subsequently relieved by adequate resuscitation, or the result of embolic material being dislodged following vessel relaxation. In contrast, the damage-control technique of embolization addresses pelvic hemorrhage in a manner analogous to the cross-clamping of major vessels, most commonly at the proximal internal iliac level [17]. The likelihood of recurrent bleeding is higher when selective TAE is performed, leading some centers to favor the damage-control modality of TAE [7]. This intervention is not without risk. Massive gluteal necrosis has been reported [18, 19] following damage-control embolization. Selective TAE is a safe alternative in terms of maintaining perfusion of the gluteal musculature [20]. Poorly considered selective TAE, however, carries a risk of potentially catastrophic sequelae, such as following embolization of the ramus profundus with resulting iatrogenic devascularization of the femoral head ( Fig 1.18-1 ). Such occurrences indicate that direct communication between the orthopedic trauma surgeon and interventional radiologist or the presence of the orthopedic traumatologist in the angiography suite is mandatory to ensure optimal patient care during resuscitation. The development of well-defined protocols [21] for the emergent treatment and resuscitation of patients presenting with pelvic trauma has been shown to decrease mortality.
2.2.1 Low-energy injuries and vascular damage
Most practitioners are aware of the association between hemorrhage and high-energy pelvic ring injury. Recent literature is replete with cautionary tales of low-energy pelvic fractures resulting in pronounced bleeding. Bleeding can occur as a result of underlying coagulopathy (eg, liver failure or pharmacological hypocoagulable states) but often occurs in the elderly without those preconditions. Loffroy et al [22] reported the case of an 83-year-old man who presented with a stable pelvic ring injury and developed life-threatening hemorrhage because of inferior epigastric artery injury. His blood pressure dropped to 70/40 mm Hg in the emergency department despite transfusion of 3 units of packed red blood cells. The patient underwent selective embolization of the inferior epigastric artery, rapidly stabilized, and survived and was discharged home [22]. Henry et al [23] indicated a mortality rate of 7.6% for elderly patients with stable fracture patterns mostly caused by simple mechanical falls. Elderly patients in his series were four times more likely to die and nearly twice as likely to receive a blood transfusion, despite having significantly less severe osseous injury. Krappinger et al [24] presented a series of 11 patients older than 65 years who required TAE for control of hemorrhage following stable pelvic fractures caused by a low-energy mechanism. Using clinical criteria [24] of increasing local pain and either hypotension or tachycardia, or relevant drop in hematocrit level, contrast computed tomography (CT) was ordered for evaluation. Of 328 patients, 15 underwent contrast CT in response to those criteria with a finding of contrast blush in 11, leading to successful TAE. The authors [24] recommended that every elderly patient with a low-energy pelvic fracture be observed and that serial hematocrit levels be obtained to diagnose hemorrhage in a timely manner.
2.2.2 Iatrogenic vascular injury
Some degree of hemorrhage occurs due to all surgical procedures performed on the pelvis, regardless of the approach. In most cases hemorrhage occurs in response to the unavoidable manipulation of clotted venous beds, debridement of fracture surfaces, or injury to relatively small vessels, however, iatrogenic injury to larger vessels is also possible. Specific fracture patterns and approaches place particular vessels at risk. Anterior approaches place the external iliac vessels at risk. Of particular concern when dissecting anterior to the obturator foramen is the corona mortis, an aberrant vessel connecting the obturator artery and external iliac or inferior epigastric artery. The superior gluteal artery is at risk with any posterior approach and with manipulation of fractures involving the greater sciatic notch or when placing retractors within the notch. Obtaining control of hemorrhage which occurs because of injury to the superior gluteal artery is difficult via the available present day surgical approaches and may necessitate angiography. Preoperative angiographic assessment should thus be considered for any patient with a fracture involving the greater sciatic notch for whom reduction maneuvers are being considered [25]. Damage to the superior gluteal artery resulting in the need for emergent TAE has also been reported during use of the Stoppa approach [26]. Any procedure where drill bits, clamps, or screws may enter the true pelvis can lead to damage to the internal iliac artery or its tributaries. Karkare et al [27] described an approach to the internal iliac vessels if vascular control is needed and if a vascular surgeon is not immediately available. Rapid control can be easily accomplished via the lateral window of the ilioinguinal approach. The interval between the iliopsoas and the peritoneal cavity is rapidly developed. The external iliac artery and vein which lie anteromedial to the psoas can be identified and followed proximally to their origin from the common iliac vessels. Care must be taken to leave the soft tissue adjacent to the external iliac vessels intact so as to avoid injury to the sympathetic nerves and lymphatic system. Continuing the dissection medially will lead to the ureter and the internal iliac artery. Dissection terminates at the bifurcation of the common iliac vessels located at the level of the anterior superior iliac spine. Internal iliac vessels can be manually compressed against the sacral promontory, or cross-clamp applied. Placement of a cross-clamp requires additional dissection of the vessel, taking care to protect the ureter. Distal pulses should be assessed once internal iliac vessels have been controlled to ensure that there has been no damage to the external iliac vessels during the course of dissection [27].
2.3 Infection
In Fractures of the Pelvis and Acetabulum, 3rd edition, there is a single index citation pertaining to infection and pelvic ring fractures:
“The trend away from large posterior wounds has helped to greatly reduce the infection rate in patients with a crush injury to the pelvis. Infection still occurs with disastrous effects in some patients” [1].
An overall incidence of infection ranging from 2–13% has been reported following pelvic ring injuries but this seems low [28–32]. The incidence varies significantly among subgroups. These differences can be attributed in part to the variable nature of associated soft-tissue injuries, heterogeneity of fracture patterns, and variety of available surgical approaches and fixation options. Posterior ring injury, associated soft-tissue injury, and obesity are factors which have been shown to correlate with an increased risk. The likelihood of infection following surgical procedures in those groups appears to be at least 10–30% [3, 31, 33].
2.3.1 Infection related to soft-tissue injury/devascularization
Subcutaneous stripping is commonly seen in association with pelvic ring injury. Devitalized tissue and dead space occurring due to trauma or surgical approach can lead to wound-healing difficulties and serve as a site for colonization and infection. The presence of an associated vascular injury may further complicate treatment. One case report [3] details the clinical course of a patient who sustained a vertical shear injury resulting from traction on the limb in contrast to the more common mechanism of axial loading. Multiple sites of disruption of the internal and external iliac arteries led to embolization with subsequent need for hemipelvectomy due to extensive soft-tissue necrosis [3]. The case seen in Fig 1.18-2 is a dramatic example of internal degloving and resulting necrosis of skin, subcutaneous tissue, and gluteal musculature. This patient required initial percutaneous posterior stabilization of the bilateral sacroiliac joints and anterior external fixation with serial debridement, and free-tissue transfer of a fasciocutaneous abdominal flap based on bilateral inferior epigastric artery. Optimal treatment of patients presenting with these complex combined osseous and soft-tissue injuries requires the coordinated care of interventional radiologists, plastic surgeons, orthopedic surgeons, and general trauma surgeons to avoid the complications of sepsis, multisystem organ failure, and death.
2.3.2 Posterior approaches/fixation
In the absence of significant vascular injury or soft-tissue compromise, the incidence of infection following posterior fixation ranges from 0–4% [28, 30, 34, 35]. Surgical approaches should avoid bone prominences and obvious areas of soft-tissue compromise to minimize this risk. Full-thickness flaps should be developed and adequate skin bridges maintained. Wounds should be irrigated copiously, closed suction drains placed, and multilevel closure performed. The routine use of negative-pressure wound therapy in patients considered at elevated risk for infection, such as those who are obese, may further minimize the incidence of infection following posterior approaches [36]. Early deep infections should be treated with irrigation and debridement, retention of stable hardware, and culture-specific antibiotic therapy. Duration of antibiotic therapy must be based on the initial severity of infection, virulence of organism, host health, and clinical response, including trend of infection markers such as sedimentation rate and C-reactive protein level. Consultation with an infectious disease expert is generally advisable. When the status of treated infection is in question and the risk of continued suppressive antibiotics minimal, it is recommended that therapy be maintained until stable union has been obtained. Recommendations pertaining to the timing of hardware removal must also be specific to the patient. In the acute situation, stable hardware should be retained until adequate intrinsic pelvic ring stability is present to counteract physiological forces. Retention of hardware may necessitate serial debridement and/or coverage procedure(s). When considering removal of hardware subsequent to healing of the pelvic ring the surgeon must evaluate the likelihood of recurrent infection, the general health of the patient, and the morbidity of the surgical approach. Hardware removal is recommended when clinical evidence of infection persists or laboratory markers of latent infection (eg, elevated C-reactive protein level) remain elevated and the patient is capable of tolerating surgery. Higher rates of infection have been associated with some methods of fixation including triangular osteosynthesis. In the largest reported series [32] triangular osteosynthesis performed for the treatment of vertically unstable sacral fractures was associated with a 13% incidence of wound complications requiring surgical debridement. Two patients in this series required removal of hardware to eradicate infection. Notably, all those fractures that subsequently became infected were associated with a significant posterior soft-tissue injury [32]. It is likely those patients who have sustained an osseous injury complex enough to warrant triangular osteosynthesis have also sustained a significant soft-tissue injury resulting in an increased risk of infection unrelated to modality of fixation. The benefits of stable fixation in providing an optimal environment for host mechanisms to combat infection are generally believed to outweigh any detrimental effects of placing hardware.
2.3.3 Obesity
Obesity is a risk factor for numerous postoperative complications including infection. In one series deep infection occurred in 7.9% of nonobese patients and 30% of obese patients following open reduction and internal fixation of the posterior ring via a posterior approach. Only two of nine patients undergoing anterior fixation of the posterior pelvic ring developed an infection (one from each group). Patients with a body mass index > 30 were 6.8 times more likely to have complication and 4.6 times more likely to require additional surgery than patients with a lower BMI [31]. Complications of all types were more common in the obese group (39% versus 19%) with most related to wound problems [33].
2.3.4 Gunshot wounds
The risk of infection associated with fractures occurring as a result of low-velocity gunshot wounds to the extremities is relatively low. Watters et al [29], in a retrospective review of 40 patients who sustained a low-energy injury to the pelvis, documented similar findings in patients without associated gastrointestinal tract violation. Uncomplicated pelvic fractures without gastrointestinal tract violation (n = 3) were treated without debridement or antibiotic therapy and none developed infection. An association between the development of infection and presence of gastrointestinal tract injury was identified only when irrigation and debridement were not undertaken. The presence of retained shrapnel following injuries associated with violation of the gastrointestinal tract did not increase the risk of infection if debridement was performed. Overall mortality was 11%. A total of three infections developed including two intraabdominal abscesses. Both patients responded to a combination of drainage and/or antibiotic therapy. The third infection involved the trochanteric bursa and also responded to debridement and antibiotic therapy [29].
2.4 Neurological complications
The association between poor outcome and neurological complication in patients with pelvic fractures is exceeded only by the association between nonunion/malunion and poor outcome. Suzuki et al [37] evaluated outcome using the Majeed score, physical component of Short-Form-36 and the Iowa Pelvic score. The authors [37] demonstrated a significant association between long-term functional outcome and neurological complication. The incidence of objectively identified neurological dysfunction following nonoperative treatment of major pelvic disruptions has been estimated between 42% and 64% [38, 39]. Operative treatment yields more favorable results. In 1992 Majeed [40] applied a scoring system he had previously described to 73 patients with pelvic fractures and identified 33% with associated neurological deficits. Majeed [41] reported that nerve recovery continues for up to 2 years and that complete recovery is not obtained in “severe” injuries. Chiu et al [28] reported a 24.6% irreversible neurological deficit in patients with type C pelvic fractures. Reilly et al [42] reviewed 83 patients treated over a 3-year period and noted a 21% incidence of neurological deficit. Combined motor and sensory deficits were most common; however, some patients exhibited only sensory symptoms. The author [42] reported improvement in muscle strength of at least one grade in all patients, and complete recovery in 53%. The L5 nerve root appeared to have the least capability to recover fully.
Denis et al [43] described a commonly used classification system and correlated the incidence of neurological injury with location of sacral fracture. The sacrum was divided into three zones with respect to the sacral foramina. The highest incidence of neurological injury was seen in fractures medial to the sacral foramina and involving the spinal canal. These so-called zone III injuries were associated with a neurological deficit in 56% of patients. Fractures involving the foramina—zone II injuries—were associated with a 28% incidence of neurological deficit and fractures occurring lateral to the foramina—zone I injuries—were associated with a 6% incidence. Siebler et al [44] evaluated outcome in eleven nonoperatively treated patients with Denis type III fractures. Two patients had worsening symptoms including one who had been noncompliant with recommendation for protected weight bearing. That patient′s symptoms (loss of bowel and bladder function) recovered within 24 hours of resuming weight-bearing restrictions. The second patient developed worsening parasthesia and weakness which partially responded to decompressive laminectomy. Three patients had complete loss of bowel and bladder control including two on presentation and the patient noted above. All deficits resolved spontaneously. Five of eleven patients reported continued urinary frequency, six reported constipation, and four of those six reported occasional uncontrolled bowel movements. Three of six patients with complaints of constipation had no bowel symptoms on presentation. Seven of eleven patients exhibited some degree of hypoesthesia or parasthesia involving the lower extremities. Four patients exhibited sexual dysfunction [44]. In an attempt to favorably alter functional outcome, some practitioners have elected to decompress the sacral nerves. Taguchi et al [45] identified neurological deficits in 7 (58%) of 12 patients presenting with sacral fractures including seven unilateral Denis type II and five Denis type III “H” fractures. All patients presenting with a neurological deficit underwent decompressive laminectomy with five of seven demonstrating neurological recovery [45]. Chapman et al [46] reviewed 19 consecutive patients who presented with sacral fractures and associated cauda equina deficits over a 15-year period. All patients were treated with decompression, fracture reduction, and lumbopelvic stabilization. All patients healed and 83% demonstrated improvement or normalization of bladder and bowel function; however, the authors [46, 47] noted a 31% incidence of hardware failure and 26% incidence of infection. The decision to proceed with reduction and decompression of sacral nerve roots is based on the concept that compression of neurological structures is a cause of deficit; however, autopsy studies [48] have shown a high percentage of sacral root lesions associated with pelvic ring fractures to be the result of traction or rupture.
2.4.1 Iatrogenic neurological injury
Injury associated with iliosacral screw placement
Placement of image intensifier-guided iliosacral screws alone or in combination with alternative modalities of pelvic ring stabilization has become routine. With increased use have come numerous reports of damage to the cauda equina or isolated nerve roots. Reported rates of malpositioned screws and resulting neurological compromise range from 6–8% [49, 50]. Reilly et al [51] have shown that residual displacement of 1–2 cm following zone II sacral fractures significantly reduces the corridor available for screw placement. The literature mostly suggests improvement is likely following screw removal; however, deficits persist in some patients. Preoperative templating and placement of screws under high-quality image intensifier guidance using sequential assessment in the inlet, outlet, and true lateral projection, only after an acceptable reduction has been obtained, results in reliably accurate positioning (see Chapter 1.8.6). The use of routine neurodiagnostic monitoring (discussed further below) has been shown to decrease the incidence of problematic screw placement and resulting neurological symptoms from 7.5–0% [50]. The use of CT guidance particularly when combined with computer-assisted navigation may further improve accuracy and minimize the risks of placement. Zwingmann et al [52] were capable of decreasing the incidence of cortical perforation from 60–31% with computer-assisted screw placement, though no neurological deficits occurred in either group in their series. The risk of neurological damage associated with placement of S1 screws in dysmorphic sacra has been presumed higher; nevertheless, Conflitti et al [53] demonstrated an ability to safely place screws within the S2 body in 24 dysmorphic sacra using standard techniques and noted that significantly longer S2 screws were possible in these patients. There were neither foraminal penetrations nor extraosseus screw placements in this series [53].
Related posts:
1.16.1 Outcomes after pelvic ring injuries: general concept and conclusion
1.16.2 Outcomes after pelvic ring injuries: critical review of the world experience
1.17 Venous thromboembolism in pelvic trauma
1.19 Malunion and nonunion of the pelvis: posttraumatic deformity
1.8.4 The management of the injured pelvic ring: internal fixation of the anterior pelvic injuries—open book (type B1)
2.23 Results of treatment for fractures of the acetabulum
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
