Diagnosis and Treatment of Complications

We define a complication as a disease process that occurs in addition to a principal illness. In the lexicon of diagnosis-related groupings, complications are comorbidities. However, a broken implant complicating the healing of a radius shaft fracture hardly seems to fit either of these definitions. In orthopaedic trauma terminology, the term complication has come to mean an undesired turn of events specific to the care of a particular injury.

Complications can be local or systemic and are caused by, among other things, physiologic processes, errors in judgment, or fate. Codivilla described complications as “inconveniences.” A colleague once described a pin tract infection with external fixation as a problem, not a complication. Preventing pin tract drainage is indeed a problem that needs a solution, but when it occurs in a patient, it becomes a complication. Since 1995, The Joint Commission (formerly the Joint Commission on the Accreditation of Healthcare Organizations) has required the mandatory reporting of “sentinel events,” a type of major complication that involves unexpected occurrences such as limb loss, surgery on the wrong body part, and hemolytic transfusion reaction. Another term, “never events,” was introduced in 2001 by Dr. Ken Kizer and includes medical errors that should never occur (e.g., wrong site surgery, a retained foreign object after surgery, intraoperative or postoperative death in an American Association of Anesthesiologists class I patient, any stage 3 or 4 pressure ulcer acquired after admission). These definitions beg the question, is death from any arrhythmia caused by a congenital irritable cardiac focus any more preventable on an operating table than it is on a basketball court?

Fracture care today expects perfection. These unrealistic expectations have led in part to the current adverse medicolegal situation surrounding the care of broken bones. There are many scales for judging the quality of results but none for complications. As a starter, operative misadventures can be classified as follows: (1) unexpected events that just slow things down—such as contaminating a reamer; (2) events that change the operation but have no long-term consequences—such as breaking a drill bit; and (3) events that cause long-term harm—such as cutting a nerve.

This chapter presents current knowledge about three systemic complications (fat embolism syndrome [FES], thromboembolic disorders, and multiple organ system dysfunction and failure) and five local complications (soft tissue damage, vascular problems, posttraumatic arthrosis, peripheral nerve injury, and complex regional pain syndrome [CRPS] or reflex sympathetic dystrophy [RSD]).

Systemic Complications

Fat Embolism Syndrome

Fat embolism syndrome is the occurrence of hypoxia, confusion, and petechiae a few days or even hours after a long bone fracture. FES is distinct from posttraumatic pulmonary insufficiency, shock lung, and acute respiratory distress syndrome (ARDS). When known etiologic factors of posttraumatic pulmonary insufficiency such as pulmonary contusion, inhalation pneumonitis, oxygen toxicity, and transfusion lung are excluded, there remains a group of patients who have FES with unanticipated respiratory compromise several days after a diaphyseal fracture.

Fat embolism was first described by Zenker in 1861 in a railroad worker who sustained a thoracoabdominal crush injury. It was initially hypothesized that the fat from the marrow space embolized to the lungs and caused the pulmonary damage. Fenger and Salisbury believed that fat embolized from fractures to the brain, resulting in death. Von Bergmann first clinically diagnosed fat embolism in a patient with a fractured femur in 1873. The incidence of this now recognized complication of long bone fracture was extensively documented by Talucci and coworkers in 1913 and subsequently studied during World Wars I and II and the Korean conflict. Mullins described the findings in patients who died as “lungs that looked like liver.” Wong and colleagues reported on the use of continuous pulse oximeter monitoring and daily intermittent arterial blood gas to define the incidence pattern and severity of long bone fractures compared with control participants; they found that long bone fracture patients had more desaturation episodes, longer duration of total desaturation, and larger total area under desaturation curves in both the prefracture repair and the postfracture periods.

Although the fat in the lungs comes from bone, other processes are required to produce the physiologic damage to lung, brain, and other tissues. Although the term fat embolism syndrome does not describe the pathomechanics of this condition as was originally hypothesized, embolization of active substances and fat from the injured marrow space has traditionally been thought to be the source of embolic fat. Recent studies suggest otherwise. Mudd and associates did not observe any myeloid tissue in any of the lung fields at autopsy in patients with FES and suggested that the soft tissue injury, rather than fractures, was the primary cause of FES. Husebye and colleagues also noted that FES might also result from “an abnormal patient reaction to the fat intravasation.” ten Duis in a review of the literature stated that “future attempts to unravel this syndrome . . . should pay full attention to differences in the extent of accompanying soft tissue injuries that surround a long bone fracture.” In a laboratory rabbit model, Aydin and colleagues found that pulmonary contusion had more deleterious effects than fractures in the formation of cerebral fat embolism.

Although there are many unanswered questions about FES, several issues are apparent. It strikes the young patients; older patients with significant upper femoral fractures do not seem at risk. It usually occurs after lower, not upper, limb fractures and is more frequent with closed fractures. Russell and associates reported a case of fat embolism in an isolated humerus fracture. McDermott and colleagues reported three cases of patients with tibial fractures from football injuries who also had dehydration and developed FES, and they concluded that adequate preoperative hydration, especially if injuries were sustained during heavy exercise, may reduce the risk of developing FES. In a prospective study, Chan and associates found an incidence of 8.75% of overt FES in all fracture patients, with a mortality rate of 2.5%. The incidence rose to 35% in patients with multiple fractures. Other investigators reported the incidence of FES between 0.9% and 3.5% in patients with long bone fractures.

Early recognition of the syndrome is crucial to preventing a complex and potentially lethal course. Clinically, FES consists of a triad of hypoxia, confusion, and petechiae appearing in a patient with fractures. The disease characteristically begins 1 to 2 days after fracture after what has been called the latent or lucid period. Sixty percent of all cases of FES are seen in the first 24 hours after trauma, and 90% of all cases appear within 72 hours. Gurd and Wilson’s criteria for FES are commonly used, with the clinical manifestations grouped into either major or minor signs of FES. The major signs are respiratory insufficiency, cerebral involvement, and petechial rash. The minor signs are fever, tachycardia, retinal changes, jaundice, and renal changes. Petechiae are caused by embolic fat. They are transient and are distributed on the cheek, neck, axillae, palate, and conjunctivae. The fat itself can be visualized on the retina. A fall in hematocrit levels and alterations in blood clotting profile, including a prolongation of the prothrombin time, can be observed. The diagnosis of FES is made when one major and four minor signs are present ( Table 23-1 ) along with the finding of macroglobulinemia. The most productive laboratory test is measurement of arterial oxygenation on room air. When the Po 2 is less than 60 mm Hg, the patient may be in the early stages of FES.

TABLE 23-1


Major Criteria Minor Criteria
Hypoxemia (Pao 2 <60 mm Hg) Tachycardia >110 beats/min
Central nervous system depression Pyrexia >38.3°C
Petechial rash Retinal emboli on funduscopy
Pulmonary edema Fat in urine
Fat in sputum
Decreased hematocrit

Source: Gurd AR, Wilson RI: The fat embolism syndrome, J Bone Joint Surg Br 56:408–416, 1974.

* A positive diagnosis requires at least one major and four minor signs.

Lindeque and colleagues believe that Gurd and Wilson’s criteria are too restrictive and should also include the following: (1) Pco 2 of more than 55 mg Hg or pH of less than 7.3; (2) sustained respiratory rate of more than 35 breaths/min; and (3) dyspnea, tachycardia, and anxiety. If any one of these is present, then the diagnosis of FES is made. Other supportive findings include ST segment changes on electrocardiography and pulmonary infiltrates on chest radiography.

Neurologic changes have been noted in up to 80% of patients. It is important to assess the neurologic status of the patient to differentiate among fat embolization, frontal concussion or contusion, and intracranial mass lesions. Although hypoxia alone can cause confusion, in FES, petechial hemorrhages, particularly in the reticular system, may alter consciousness. These changes persist despite adequate oxygen therapy. Focal neurologic findings should be investigated to rule out lesions caused by associated head trauma. Persistent alterations of consciousness or seizures are a bad prognostic sign.

Clinically, fat embolism is a diagnosis of exclusion. In the first few days, sudden pulmonary compromise can also result from pulmonary embolism (PE), heart failure, aspiration, and medication reaction. When these possible causes have been excluded along with many other less likely conditions, fat embolism becomes the leading cause of morbidity in the injured patient with a long bone lower limb fracture.

Fat globules are found in blood, sputum, urine, and cerebrospinal fluid. The urine or sputum can be stained for fat using a saturated alcoholic solution of Sudan III. Sudan III stains neutral fat globules yellow or orange. The Gurd test, in which serum is treated with Sudan III and filtered, is also diagnostic. These tests are of historical interest when house staff actually handled specimens.

The specificity of these tests is in question. Fat droplets are normally found in sputum. In addition, Peltier believes that detection of fat droplets in circulating blood and urine is too sensitive a test for the clinical diagnosis of FES. Furthermore, because the embolic phenomena associated with FES are transient and may not be detected on spot testing, these laboratory investigations are of research interest only and are not part of the usual clinical workup.

The experimental study of FES is linked historically to the study of the circulation of blood, the development of intravenous (IV) therapy, and transfusion. As early as 1866, Busch experimented with marrow injury in the rabbit tibia and showed that fat in the marrow cavity would embolize to the lungs. Pulmonary symptoms have been produced in the absence of fracture by the IV injection of fat from the tibia of one group of rabbits to another.

There are several reasons for uncertainty about the role of bone fat in producing FES. First, researchers have failed to develop an animal model that reproduces the human syndrome. Moreover, injection of human bone marrow fat into the veins of experimental animals has shown that neutral fat is a relatively benign substance, and it is not certain that the bones contain enough fat to cause FES. One hypothesis is that the fat that appears in the lungs originated in soft tissue stores and aggregated in the blood stream during posttraumatic shock. However, chromatographic analysis of pulmonary vasculature fat in dogs after femoral fracture has shown that the fat most closely resembles marrow fat. In contrast, Mudd and colleagues reported that there was no evidence of myeloid elements on postmortem studies of lung tissue in patients with FES. Furthermore, extraction of marrow fat from human long bones has shown that sufficient fat is present to account for the observed quantities in the lungs and other tissues. The relative lack of triolein in children’s bones may explain why they have a significantly reduced incidence of FES compared with adults.


Although the precise pathomechanics of FES are unclear, Levy found many nontraumatic and traumatic conditions associated with FES. The simplest hypothesis is that broken bones liberate marrow fat that embolizes to the lungs. These fat globules produce mechanical and metabolic effects culminating in FES. The mechanical theory postulates that fat droplets from the marrow enter the venous circulation via torn veins adjacent to the fracture site.

Peltier coined the term intravasation to describe the process whereby fat gains access to the circulation. The conditions in the vascular bed that allow intravasation to take place also permit marrow embolization. Indeed, marrow particles are found when fat is found in the lungs ( Fig. 23-1 ).

Figure 23-1

Histologic appearance of fat from a pulmonary fat embolism in a vessel of the pulmonary alveoli. C, Capillary; F, fat globules (arrowheads).

(From Teng QS, Li G, Zhang BX: Experimental study of early diagnosis and treatment of fat embolism syndrome, J Orthop Trauma 9:183–189, 1995.)

Mechanical obstruction of the pulmonary vasculature occurs because of the absolute size of the embolized particles. In a dog model, Teng and coworkers found 80% of fat droplets to be between 20 and 40 µm. Consequently, vessels in the lung smaller than 20 µm in diameter become obstructed. Fat globules of 10 to 40 µm have been found after human trauma. Systemic embolization occurs either through precapillary shunts into pulmonary veins or through a patent foramen ovale.

The biochemical theory suggests that mediators from the fracture site alter lipid solubility, causing coalescence, because normal chylomicrons are smaller than 1 µm in diameter. Many of the emboli have a histologic composition consisting of a fatty center with platelets and fibrin adhered. Large amounts of thromboplastin are liberated with the release of bone marrow, leading to activation of the coagulation cascade.

Studies of the physiologic response to the circulatory injection of fats have shown that the unsaponified free fatty acids are much more toxic than the corresponding neutral fats. Peltier hypothesized that elevated serum lipase levels present after the embolization of neutral fat hydrolyzes this neutral fat to free fatty acids and causes local endothelial damage in the lungs and other tissues, resulting in FES. This chemical phase might in part explain the latency period seen between the arrival of embolic fat and more severe lung dysfunction. Elevated serum lipase levels have been reported in association with clinically fatal FES. Alternative explanations are also possible for the toxic effect of fat on the pulmonary capillary bed. The combination of fat, fibrin, and (possibly) marrow may be sufficient to begin a biochemical cascade that damages the lungs without postulating enzymatic hydrolysis of neutral fat. Bleeding into the lungs is associated with a decrease in the hematocrit level. The resulting hypoxemia from the mechanical and biochemical changes in the lungs can be severe—even to the point of death of the patient.

Pape and associates demonstrated an increase in neutrophil proteases from central venous blood in a group of patients undergoing reamed femoral nailing. In another study, Pape and colleagues demonstrated the release of platelet-derived thromboxane (a potent vasoconstrictor of pulmonary microvasculature) from the marrow cavity. Peltier demonstrated the release of vasoactive platelet amines. These humeral factors can lead to pulmonary vasospasm and bronchospasm, resulting in vascular endothelial injury and increased pulmonary permeability. Indeed, thrombocytopenia is such a consistent finding that it is used as one of the diagnostic criteria of FES. Barie and coworkers associated pulmonary dysfunction with an alteration in the coagulation cascade and an increase in fibrinolytic activity.

Autopsy findings in patients who died of FES do not, however, show a consistent picture. This may be caused by a lack of clear-cut criteria that define patients included in a given series but may also be because manifestations of FES depend on a wide number of patient, accident, and treatment variables.

In light of the incidence of fat emboli and FES in trauma patients, it is likely that other precipitating or predisposing factors such as shock, sepsis, or disseminated intravascular coagulation are needed for the phenomenon of embolized fat to cause FES. Müller and associates summarized that “fat embolism syndrome is likely the pathogenetic reaction of lung tissue to shock, hypercoagulability, and lipid mobilization.”

Two clinically related treatment questions arise: (1) Is there an association between intramedullary (IM) nailing, FES, and other injuries? (2) Is there an effect from different nailing methods on the incidence of FES? In 1950, Küntscher described FES as a complication of IM nailing. Pape and associates found that early operative fracture fixation by nailing was associated with an increased risk of ARDS in patients with thoracic injury. These results are in contrast with those of the group without thoracic injury. Thoracic trauma is associated with direct pulmonary injury. The pathogenic mechanisms were examined by Lozman and colleagues. Thus, the timing and the associated injuries are crucial in deciding when and how to use a nail.

In a prospective study, Pape and associates showed a significant impairment of oxygenation in multiple trauma patients who underwent reamed nailing. A group of similar patients who had unreamed nailing did not have the same signs of pulmonary dysfunction. These investigators reasoned that the most likely difference between the two groups was a lower degree of fat embolization in the unreamed group. In sheep, Pape and colleagues demonstrated intravasation of fat associated with reaming of the IM canal. They concluded that the unreamed procedure caused substantially less severe lung damage than the reamed procedure. However, Heim and associates found that there was a significant increase in IM pressure associated with unreamed nail insertion and that both reamed nailing and unreamed nailing lead to bone marrow intravasation ( Fig. 23-2 ). Thus, the use of an unreamed nail does not solve the problem of bone marrow embolization and resultant pulmonary dysfunction.

Figure 23-2

Intramedullary pressure during reamed nailing of the femur.

(From Heim D, Regazzori P, Tsakiris DA, et al: Intramedullary nailing and pulmonary embolism: does unreamed nailing prevent embolization? An in vivo study in rabbits, J Trauma 38:899–906, 1995.)

What influences the degree of fat embolization? The answer has not been fully elucidated. High IM pressures have been linked with fat embolization and FES. Wozasek and coworkers looked at the degree of fat intravasation during reaming, and IM nailing and correlated this with IM pressure changes and echocardiographic findings. They found peak IM pressures in both the tibial and the femoral nailings in the first two reaming steps. Insertion of the nail caused only minimal pressure rises (but this was after reaming). Echocardiography, however, demonstrated that the maximal infiltration of particles occurred when the nail was inserted. They concluded that the phenomenon of fat intravasation did not depend on the rise in IM pressure. Pinney and associates studied 274 patients with isolated femur fractures and found that waiting more than 10 hours after injury was associated with a 2.5-fold increase in FES. Bulger and coworkers noted that early IM fixation did not seem to increase the incidence or severity of FES.

Prevention and Treatment

The risk of FES can be decreased by several measures. Proper fracture splinting and expeditious transport, use of oxygen therapy in the postinjury period, and early operative stabilization of long bone fractures of the lower extremities are three important measures that can be taken to reduce the incidence of this complication. Blood pressure, urinary output, blood gas values, and—in the more critically injured—pulmonary wedge pressures should be monitored to evaluate fluid status and tissue perfusion more precisely. Dramatic advances in emergency medical transport have resulted in increasing survival of patients with complex polytrauma and high injury severity scores. This has led in some instances to a tendency to “scoop and run” without traction splinting. Unsplinted long bone fractures in patients transported over long distances are a setup for IV fat intravasation. Oxygen therapy by mask or nasal cannula lessens the decrease in arterial oxygenation after fracture and appears to have value in the prevention of FES.

If surgery is delayed, the patient’s arterial oxygen on room air is measured daily, and supplemental oxygen therapy is continued until the posttraumatic decrease in oxygen tension is complete and the PaO 2 on room air returns toward normal. Alternatively, if inspired oxygen tension (FiO 2 ) can be measured accurately, the shunt equation can be used to monitor pulmonary performance. Teng and coworkers completed preliminary development of a dog model of FES to establish diagnostic criteria sufficiently sensitive and specific enough for diagnosis of FES in the early stages. They correlated blood gas analysis samples with computer image analysis of oil red O-stained pulmonary artery blood samples. Although fracture fixation and particularly medullary nailing cause a transient decrease in oxygenation, the immediate stabilization of fractures before the development of low arterial saturation may prevent the occurrence of FES.

In a prospective randomized study of 178 patients, Bone and associates confirmed that early fracture stabilization, within the initial 24 hours after injury, decreased the incidence of pulmonary complications. Likewise, Lozman and colleagues, in a prospective randomized study, concluded that patients receiving immediate fixation had less pulmonary dysfunction after multiple trauma and long bone fractures than did patients receiving conservative treatment.

Although IM nailing is the preferred method of stabilization, the timing of nailing is a point of controversy. There was concern that immediate nailing of long bone fractures early in the postinjury period would increase the incidence of pulmonary complications, including FES. External fixation of long bone fractures can be used as a temporizing alternative to IM nailing. Earlier studies showed no evidence to support the view that the effect of reaming on intravascular fat is additive or that immediate reamed IM fixation causes pulmonary compromise. In fact, the opposite is true, probably because fracture stabilization removes the source of intravascular marrow fat and decreases shunting in the lung because the patient can be mobilized to an upright position.

No cases of FES were seen in a retrospective study by Talucci and associates in which 57 patients underwent immediate nailing. Similarly, Behrman and colleagues reported a lower incidence of pulmonary complications for patients undergoing early fixation among 339 trauma patients who underwent either immediate or late fixation of femoral fractures. In the study by Lozman and colleagues, patients who had delayed fracture fixation had a higher intrapulmonary shunt fraction compared with that in the early fixation group.

Early IM nailing of long bone fractures is not without complications. Pell and coworkers, using intraoperative transesophageal echocardiography, demonstrated varying degrees of embolic showers during reamed IM nailing. FES developed postoperatively in three patients, and one patient died. Other studies showed an increased number of pulmonary complications associated with early, reamed IM nailing of femoral shaft fractures.

Specific therapy has been used in an attempt to decrease the incidence of FES. No clinical effect on the rate of FES has been found with increased fluid loading or the use of hypertonic glucose, alcohol, heparin, low-molecular-weight dextran, or aspirin. Various studies have looked at the efficacy of corticosteroids in reducing the clinical symptoms of FES. Large doses of steroids immediately after injury do have a beneficial effect. Corticosteroids most likely decrease the incidence of FES by limiting the endothelial damage caused by free fatty acids. Babalis and colleagues, in a randomized, prospective study of 87 patients with long bone fractures allocated to either a placebo group or a group treated with IV, low-dose methylprednisolone, found that methylprednisolone deceased posttraumatic hypoxemia and probably fat embolism in patients with isolated lower limb long bone fractures, especially when early fracture stabilization is not possible. They concluded that the prophylactic use of methylprednisolone in small doses was useful in preventing posttraumatic hypoxemia and FES. Although this is encouraging, routine use of steroids is not without significant risk of complications and is not routinely employed.

Fat embolism syndrome is primarily a disease of the respiratory system, and current treatment is therefore mainly with oxygen and meticulous mechanical ventilation. Treatment of FES remains mainly supportive.

Finally, in no way should it be construed that either clinical experience or scientific investigation provides a sure pathway to prevent the appearance of this significant postinjury and potentially lethal problem. Although careful review of the medical record might suggest how things could have been done alternatively, there is no certainty, for example, that waiting another day as the PaO 2 returned toward normal before performing a nailing would have prevented the complication of FES—indeed, it might have invited another one.

Thromboembolic Disorders


In 1846, Virchow proposed the triad of thrombogenesis: increased coagulability, stasis, and vessel wall damage ( Fig. 23-3 ). These are all factors that are unfavorably affected by trauma. Virchow also linked the presence of deep venous thrombosis (DVT) with PE and deduced that a clot in the large veins of the thigh embolized to the lungs. Laennec, in 1819, was the first to describe the clinical presentation of an acute pulmonary embolus. The pathogenesis of a proximal DVT was first described by Cruveilhier in 1828. Venous thrombi have been shown to develop near the valve pockets on normal venous endothelium and are not necessarily related to inflammation of the vessel wall.

Figure 23-3

Virchow’s triad.

Trauma creates a hypercoagulable state. Vessel wall injury with endothelial damage exposes blood to tissue factor, collagen, basement membrane, and von Willebrand factor, which induce thrombosis through platelet attraction and the intrinsic and extrinsic coagulation pathway. Antithrombin (AT-III) activity, which decreases the activity of thrombin and factor Xa, was found to be below normal levels in 61% of critically injured trauma patients. Also, fibrinolysis is decreased and appears to be from increased levels of plasminogen activator inhibitor type 1 (PAI-1), which inhibits tissue plasminogen activator and thus decreases the production of plasmin.

The presence of heart disease alone increases the risk of PE by 3.5 times, and this is further increased if atrial fibrillation or congestive heart failure is present. The risk of DVT is increased during pregnancy and is especially great in the postpartum period. Spinal cord injury is associated with a threefold increase in lower extremity DVT and PE.

A meta-analysis by Velmahos and coworkers looked at DVT and risk factors in trauma patients. The following variables studied did not have a statistically significant effect for increasing the development of DVT: gender, head injury, long bone fracture, pelvic fracture, and units of blood transfused. The variables that were statistically significant included spinal fractures and spinal cord injury, increasing the DVT risk by two- and threefold, respectively. They could not confirm that the widely assumed risk factors of pelvic fracture, long bone fracture, and head injury affected the incidence of DVT but did note that the multiple trauma patients may have already been at the highest risk of DVT.

For immobilized trauma patients with no prophylaxis, the incidence of venography-proven thigh and iliofemoral thrombosis is between 60% and 80%. Even with full prophylaxis, the incidence of DVT is as high as 12%. Stannard and colleagues reported a significant rate of DVT in high-energy skeletal trauma patients despite thromboprophylaxis. They noted in a series of 312 patients with high-energy trauma that 11.5% developed venous thromboembolic disease with an incidence of 10% in those with nonpelvic trauma and 12.2% in the group with pelvic trauma despite thromboprophylaxis. Some investigators have concluded that “there is no adequate prophylaxis against DVT in the trauma patient.”

Trauma to the pelvis and lower extremities greatly increases the risk of DVT and PE. In an autopsy study of 486 trauma fatalities, Sevitt and Gallagher found 95 cases of PE for an incidence of 20%. At autopsy, the rate of PE after hip fracture was 52 in 114 (46%); for tibia fractures, six in 10 (60%); and for femoral fractures, nine in 17 (53%). The rate of DVT for hip fractures increased to 39 in 47 (83%) and for femur fractures to six in seven (86%) when supplemental special studies of the venous system were done at autopsy.

Pulmonary embolism is a significant cause of death after lower extremity injury. Two-thirds of patients having a fatal pulmonary embolus die within 30 minutes of injury ( Fig. 23-4 ). Whereas the incidence of fatal PE without prophylaxis after elective hip surgery is from 0.34% to 3.4%, the incidence after emergency hip surgery is from 7.5% to 10%.

Figure 23-4

A large embolus in the pulmonary artery, which was the cause of death.

(Courtesy of James E. Parker, MD, University of Louisville, Louisville, KY.)

Solheim reported a 0.5% incidence of fatal PE in a series of tibia and fibula fractures. Similarly, Phillips and coworkers reported one of 138 patients (0.7%) with severe ankle fractures developed a nonfatal PE. In a study of 15 patients with tibia fractures, Nylander and Semb found that 70% had venographic changes compatible with DVT.

The types of DVT that are at high risk for causing a PE are those that originate at the popliteal fossa or more proximally in the large veins of the thigh or pelvis. Moser and LeMoine found the risk of pulmonary embolization from distal lower extremity DVT to be relatively low. Of DVTs that are first limited to the calf, about 20% to 30% extend above the knee. Those that extend above the knee carry the same risk as femoral and popliteal thrombi. Kakkar and colleagues speculated that thrombi in the calf are securely attached and resolve rapidly and spontaneously. However, embolization from “calf only” venous thrombi does occur. Calf vein thromboses are responsible for 5% to 35% of symptomatic PE, 15% to 25% of fatal PE, and 33% of “silent” PE.

In addition to PE, complications of DVT include recurrent thrombosis and postthrombotic syndrome. Symptoms of postthrombotic syndrome are edema, induration, pain, pigmentation, ulceration, cellulitis, and stasis dermatitis. Symptoms are present in up to 20% to 40% of those having had a DVT.

Upper extremity DVT is much less common (2.5%) and can be from primary or secondary causes. The primary causes are idiopathic and effort thrombosis (Padget-Schroetter syndrome). Effort thrombosis is most common in athletes and laborers who do repetitive shoulder abduction and extension. Predisposing causes of thoracic outlet obstruction should be investigated. Secondary causes are venous catheters, venous trauma, extrinsic compression or malignancy, and hypercoagulable conditions.


The clinical signs and symptoms of DVT are nonspecific. DVT was clinically silent in two thirds of cases in which thrombosis was found at autopsy or the findings on leg venography were positive.

Clinically, the diagnosis of DVT and PE is frequently difficult. With PE, although some patients experience sudden death, many more present with gradual deterioration and symptoms similar to pneumonia, congestive heart failure, or hypotension. Symptoms can be intermittent with episodes of transient pulmonary compromise caused by clusters of small emboli. Because the clinical diagnosis is difficult, diagnostic studies are necessary so that early treatment can be instituted. Various tests are described along with limitations and advantages.

Traditionally, venography has been the diagnostic test of choice for DVT, but it is no longer the gold standard. The major drawbacks of venography are that it is usually a one-time test that cannot be done on a serial basis and has been reported to cause phlebitis in about 4% to 24% of patients and may cause thrombosis. Furthermore, the venogram may be uninterpretable because of technical considerations and can be a cause of serious allergic reactions to contrast agents.

Radioactive fibrinogen is effective in detecting thrombi in the calf but is less effective in the thigh. Fibrinogen I-125 is incorporated into a forming thrombus and can be detected. DVT in the thigh is poorly detected with this technique, and so it is not used as a screening tool for trauma patients.

Impedance plethysmography (IPG) detects the presence of DVT by measuring the increased blood volume in the calf after temporary venous occlusion produced by a thigh tourniquet and the decrease in blood volume within 3 seconds after deflation of the cuff. IPG is sensitive for diagnosing proximal DVT but is not sensitive for distal DVT. It is a poor screening tool for trauma patients.

Noninvasive venous Doppler examinations are the current standard for imaging DVTs. Continuous-wave Doppler (CWD) or Doppler ultrasound examination is easy to do and can be done at the bedside, but it requires experience to reduce the false-positive result rate. Venous thrombosis is characterized by the absence of venous flow at an expected site, loss of normal fluctuation in flow associated with respiration, diminished augmentation of flow by distal limb compression, diminished augmentation of flow by release of proximal compression, and lack of change on Valsalva maneuver. Barnes and coworkers found that Doppler ultrasonography was 94% accurate, and no errors were made in diagnosis above the level of the knee. However, for isolated calf vein thrombosis, CWD is insensitive. An additional disadvantage is that CWD may fail to detect nonobstructive thrombi even in proximal lesions.

Color-flow duplex ultrasonography (CFDU) uses a Doppler component that is color enhanced and detects blood flow by the shift in frequency from the backscatter of high-frequency sound. The frequency is shifted by an amount proportional to the flow velocity. The color saturation is proportional to the rate of flow. A black image indicates an absence of flow, flow velocities less than 0.3 cm/sec, or flow vectors at a right angle to the second beam. The addition of color allows for the easier and faster detection of vascular structures. Blood flowing away from the transducer appears blue, and blood flowing toward the transducer appears red. This has provided improved imaging of the iliac region, the femoral vein in the adductor canal, and the calf veins. CFDU is superior to duplex scanning and B-mode imaging in detecting nonocclusive thrombi because the flow characteristics in the vessels are readily detected. Several studies have reported high sensitivity and specificity in symptomatic patients.

Serial ultrasonography has been used as surveillance screening to detect DVT in trauma patients, but it was thought not to be cost effective. When DVT develops in the calf, about 25% extend to the thigh if left untreated. If the initial ultrasound missed the asymptomatic DVT and no treatment is given, approximately 2% of cases will have an abnormal proximal scan on testing 1 week later.

Magnetic resonance imaging (MRI) has been recently applied to the detection of DVT in the pelvis. Rubel and colleagues have reported on the use of MR venography to evaluate DVT in patients with pelvic and acetabular trauma. Stannard and colleagues reported that ultrasound had a false-negative rate of 77% for diagnosing pelvic DVT compared with MR venography. Stover and colleagues, in a prospective study of MR venography and contrast-enhanced computed tomography (CT), reported that the false-positive rate for MR venography was 100%, and the false-positive rate for contrast-enhanced CT was 50%. They stated that they cannot recommend the sole use of either CT venography or MR venography to screen and direct the treatment of asymptomatic thrombi in patients with fracture of the pelvic ring because of these high false-positive rates. An additional disadvantage of MR venography is the cost, which is typically 2 to 2.5 times the cost of an ultrasound scan and 1.4 times the cost of venography.

Multidetector computed tomography pulmonary angiography (CTPA) is now the primary imaging modality for imaging of patients suspected of having an acute PE. The data indicate that multidetector CTPA is more accurate than single-slice CT or ventilation/perfusion (V/Q) scans. In fact, CTPA has fewer nondiagnostic scans than V/Q scans. Conventional CT with contrast, not performed as a dedicated CTPA, is no longer indicated in the workup of acute chest pain presumed to be secondary to acute PE.

Venous Thromboembolism Protection in Orthopaedic Trauma

If DVT or PE occurs before definitive management of the fracture, the method of treating the fracture may have to be modified because of the use of therapeutic anticoagulants. There are three major approaches to the treatment of DVT: protect the patient from ongoing risk of thrombosis (prevention is really a misnomer in trauma), ignore it if it occurs, or treat it. Implicit in each of these approaches is a consideration of (1) the risk of the intervention and (2) the risk if no intervention is taken. For protection of the patient from thrombosis, what is the risk of the agent used versus the risk of DVT and its complications? After DVT develops, if no treatment is undertaken, what is the risk of PE compared with the complications of therapy?

The concept of venous thromboembolism (VTE) prophylaxis with traumatic injuries of the musculoskeletal system is a misnomer. The process of clot formation has likely already begun at the time of injury. Prophylaxis is really ex post facto . The concept of DVT “protection” rather than prophylaxis is probably more accurate for the orthopaedic trauma patient.

There are four types of patients with orthopaedic trauma who should be considered for VTE (venous thrombosis and PE) protection: polytrauma patients, elderly hip or pelvis fracture patients, isolated extremity injury patients, and spinal cord injury patients. It is challenging to generalize about VTE prophylaxis for all four groups together as one. VTE prophylaxis or protection for each type of orthopaedic patient will be discussed separately.

Polytrauma Patient.

Without prophylaxis, patients with multisystem or major trauma have a DVT rate that exceeds 50% with a fatal PE rate of 0.4% to 2.0%. PE is a common cause of death in trauma patients. VTE accounts for about 9% of readmissions to the hospital after trauma. Polytrauma patients represent a heterogeneous group and present many challenges. These patients are often cared for by multiple services (e.g., general surgery, critical care, orthopaedic trauma surgery) as a team. Many of these polytrauma patients with acidosis, coagulopathy, and hypothermia are initially treated with a damage control orthopaedics (DCO) approach (e.g., temporary spanning external fixation of long bone fractures).

Although VTE prophylaxis in the treatment of these patients needs to be individualized, several concepts are now becoming fairly standard. The recommendations published by the American College of Chest Physicians (ACCP) every 2 to 3 years as a Chest journal supplement now in its 9th edition are often considered to set the standard. Note, however, that ACCP panelists may have vested economic interests in the agents used for VTE prevention. Indeed, though less chic, Coumadin, when it can be monitored, is an acceptable and economic pharmacologic alternative with a long history. For prophylaxis, an international normalized ratio (INR) of 1.4 to 1.6 is adequate, and for treatment of thrombosis with embolism, an INR of 2.0 to 2.5. These can be administered without a loading dose.

The recommendation for major trauma is for the use of low-dose unfractionated heparin (UFH), low-molecular-weight heparin (LMWH), or mechanical prophylaxis, preferably intermittent pneumatic compression. In addition, the recommendations add the caveat that for “major trauma patients at high risk for VTE (including those with acute spinal cord injury, traumatic brain injury, and spinal surgery for trauma)” that mechanical prophylaxis be added to pharmacologic prophylaxis when it is not contraindicated by lower extremity injury. Furthermore, the recommendations of the 9th edition, similar to the prior several editions, recommend against the use of an inferior vena cava (IVC) filter as primary VTE prevention. The use of periodic surveillance with venous compression ultrasound in major trauma patients was not recommended.

Hip Fracture Patient.

Without prophylaxis, these patients have DVT rates of 50% with a proximal DVT rate of 25%. Fatal PE is more common in hip fracture patients than in total hip and knee arthroplasty patients. Recommendations from the 9th edition Chest supplement are for the use of one of the following antithrombotic approaches (rather than no antithrombotic prophylaxis) for a minimum of 10 to 14 days: LMWH, fondaparinux, low-dose UFH, adjusted-dose vitamin-K antagonist, aspirin, or an intermittent pneumatic compression device. Interestingly, the Supplement noted a preference for LMWH compared with the other agents that they recommended. In addition, they recommended that if hip fracture surgery would be delayed that LMWH would be started. From an ethical standpoint, one might question if overvigorous anticoagulation is appropriate in every case. Consider, for example, an elderly, demented nursing home patient who falls and has a hip fracture. Indeed, there is potentially a wide range of mischief from the use of anticoagulation in elderly patients, who may experience bleeding, stroke, and diagnostic misadventures as a result of overzealous prophylaxis!

Isolated Lower Leg Injury Patient Distal to The Knee.

Isolated extremity injuries are probably the most common injuries seen by orthopaedic physicians. The recommendation was for “no prophylaxis rather than pharmacologic thromboprophylaxis in patients with isolated lower leg injuries requiring leg immobilization.” On the other hand, surveillance of patients for VTE and protection or prophylaxis seems prudent, particularly for patients with risk factors such as underconditioning middle age or old, multiparity, and so on. At a minimum, one could consider simple measures (early mobilization, ankle pump exercises, graduated compression stockings, intermittent pneumatic compression (IPC) with or without graduated compression stockings) or more intensive measures (preoperative and immediate postoperative graduated compression stockings followed by a short course of LMWH, synthetic pentasaccharides, or adjusted vitamin K antagonists). Patient and office staff education are wise steps.

Spinal Cord Injury Patient.

Acute spinal cord injury was the risk factor most strongly associated with the development of DVT in major trauma. Rogers and colleagues in their meta-analysis noted that spinal cord injuries or spinal fractures are high risk for VTE. It is recommended that “thromboprophylaxis be provided for all patients with acute spinal cord injuries.” The recommendations for patients with acute spinal cord injury were for pharmacologic prophylaxis (low-dose UFH or LMWH) together with mechanical prophylaxis (e.g., intermittent pneumatic compression).

Treatment of Existing Deep Venous Thrombosis and Pulmonary Embolism.

After DVT or PE is suspected, the clinical impression should be confirmed by diagnostic testing. Parenteral anticoagulants can be started unless contraindicated while the diagnostic testing is pending; this is the current recommendation for patients in whom there is a high clinical suspicion of acute VTE. If there is an “intermediate” or “low” clinical suspicion of DVT, the recommendation is not to start parenteral anticoagulants. The options for anticoagulation are parenteral anticoagulation with LMWH, fondaparinux, IV UFH, or subcutaneous UFH. Furthermore, there is more focus on the specific location of the DVT (i.e., proximal or distal in the lower extremities). When there are no severe symptoms or risk factors for extension, the recommendation is for serial imagery of the deep veins for 2 weeks over initial anticoagulation. If there are severe symptoms or risk factors for extension, the recommendation is for anticoagulation. For patients with an acute proximal DVT, the recommendation is for LMWH or fondaparinux over IV UFH and over subcutaneous heparin.

There are also now changes in the recommendation for how, when, and where the anticoagulation is administered. When patients with an acute DVT of the leg are treated with LMWH, the suggestion is for once-a-day over twice-a-day administration with the caveat that the once-a-day administration is twice the dosage of the once-a-day dose. In terms of where treatment is given, the recommendation is for initial treatment at home in patients with acute DVT provided “home circumstances are adequate.” It has been suggested that the optimal duration of thromboprophylaxis after multiple trauma be “largely based on rational, clinical decision-making on a case-by-case basis.”

Although options such as thrombolytic therapy for the treatment of acute DVT may be considered, the latest recommendations are for anticoagulant therapy alone for catheter-directed thrombolysis. Even if catheter-directed thrombolysis is not available, the latest recommendations are for anticoagulant therapy alone over systemic thrombolysis. Operative venous thrombectomy is also deemphasized because anticoagulant therapy alone is suggested over operative venous thrombectomy. Furthermore, the recommendations are for anticoagulation of the same intensity and duration in patients who undergo thrombosis removal as in patients who are comparable who did not undergo thrombosis removal.

Inferior vena cava filters are used less frequently and are no longer recommended as primary prophylaxis against VTE. IVC filters have associated complications such as venous stasis leading to edema, pain, varicose veins, and skin ulcers in a condition known as the postphlebitic syndrome. Other complications include bleeding or thrombus formation at the site of insertion, migration of the filter, and perforation of the vena cava. Martin and coworkers described a case report of phlegmasia cerulea dolens as a complication of an IVC filter for prophylaxis against PE in a man with a fracture of the acetabulum. In addition, filters are not 100% effective.

Vena cava interruption is performed when heparinization is contraindicated, as in patients with a preexisting bleeding disorder; severe hypertension; neurologic injury; or bleeding problems of pulmonary, gastrointestinal (GI), neurologic, or urologic etiology. If anticoagulation fails to stop pulmonary emboli, vena cava interruption is indicated. Also, if patients develop complications with anticoagulation, they can be switched to vena cava interruption. An additional approach is the preoperative use of vena cava interruption in patients who are at extremely high risk for PE.


The current literature clearly indicates that certain trauma patients benefit from some form of surveillance or prophylaxis. DVT and PE are common causes of morbidity, mortality, and litigation associated with the care of orthopaedic trauma patients. The complete prevention of thromboembolism in orthopaedic trauma is impossible because trauma cannot be anticipated. One or more components of Virchow’s triad are usually present from the time of injury, so the concept of “DVT prophylaxis” is a misnomer for trauma patients.

Questions remain such as, “Is there a genetic predisposition to VTE?” Risk stratification is being used in other areas of medicine and is only beginning to be understood in orthopaedic trauma. In addition, combinations of injuries, multiple lower extremity fractures with a spinal cord injury, or a pelvic fracture together with a femur fracture likely exponentially increases the risk of VTE. Although current prophylactic regimens in trauma patients significantly reduce the relative risk for DVT and PE, no method provides 100% protection. Further randomized controlled trials of DVT prophylaxis in trauma patients are needed. Nonetheless, our ability to diagnose VTE and protect patients from it is constrained by acute hemorrhage and an inability to tolerate anticoagulation, soft tissue contusion, and extremity injuries that prevent the placement of IPC and GCS. Prophylaxis is also impossible, and the best we can do is to try for VTE protection. Nonetheless, there are many methods of DVT surveillance and protection at our disposal, and they should be considered. Although the ideal method of documentation for hospital or outpatient examinations is unknown, we have found that a note such as “no signs or symptoms of PE/DVT” along with the documentation of VTE protection or prophylaxis to be prudent and reasonable. Clinicians are also advised to stay informed of the consensus recommendations that are published every 2 to 3 years in the Chest supplement.

Multiple Organ System Dysfunction and Failure

Multiple organ failure (MOF) is defined as the sequential failure of two or more organ systems remote from the site of the original insult after injury, operation, or sepsis. The organ failure can be pulmonary, renal, hepatic, GI, central nervous, or hematologic. These systems can be monitored for objective criteria for failure, but criteria vary from series to series ( Tables 23-2 and 23-3 ). The risk of developing MOF and the severity of the MOF can also be graded by measuring the effects on specific organ systems.

TABLE 23-2


Organ or System Dysfunction Advanced Failure
Pulmonary Hypoxia requiring intubation for 3–5 days ARDS requiring PEEP >10 cm H 2 O and Fio 2 >0.5
Hepatic Serum total bilirubin ≥2–3 mg/dL or liver function tests ≥ twice normal Clinical jaundice with total bilirubin ≥8–10 mg/dL
Renal Oliguria ≤479 mL/day or creatinine ≥2–3 mg/dL Dialysis
Gastrointestinal Ileus with intolerance of enteral feeds >5 days Stress ulcers, acalculous cholecystitis
Hematologic PT/PTT >125% normal, platelets <50,000–80,000 DIC
Central nervous system Confusion, mild disorientation Progressive coma
Cardiovascular Decreased ejection fraction or capillary leak syndrome Refractory cardiogenic shock

ARDS, Adult respiratory distress syndrome; DIC, disseminated intravascular coagulation; Fio 2 , fraction of inspired air in oxygen; PEEP, positive end-expiratory pressure; PT, prothrombin time; PTT, partial thromboplastin time.

Source: Deitch EA: Pathophysiology and potential future therapy, Ann Surg 216:117–134, 1992, with permission.

TABLE 23-3


Pulmonary Need of ventilator support at Fio 2 ≥0.4 for 5 consecutive days
Hepatic Hyperbilirubinemia >2.0 g/dL and an increase of serum glutamic-oxaloacetic transaminase
Gastrointestinal Hemorrhage from documented or presumed stress-induced acute gastric ulceration. This can be documented by endoscopy; if endoscopy is not performed, then the hemorrhage must be sufficient to require 2 units of blood transfusion.
Renal Serum creatinine level >2.0 mg/dL. If a patient has preexisting renal disease with elevated serum creatinine level, then doubling of the admission level is defined as failure.

Multiple organ failure is the end result of a transition from the normal metabolic response to injury to persistent hypermetabolism and eventual failure of organs to maintain their physiologic function. A 1991 consensus conference used the term multiple organ dysfunction syndrome (MODS) to describe this spectrum of changes. Organ dysfunction is the result of either a direct insult or a systemic inflammatory response, known clinically as the systemic inflammatory response syndrome (SIRS), which can be reversible or progress to MODS or MOF. SIRS can be caused by a variety of infectious and noninfectious stimuli ( Fig. 23-5 ). Treatment of the offending source must be undertaken early because when organ failure has begun, treatment modalities become progressively ineffective. Fry identified the mortality rate for failure of two or more organ systems as about 75%. If two organ systems fail and renal failure occurs, then the mortality rate is 98%. MOF has been described as the number one cause of death in surgical intensive care units (ICUs).

Figure 23-5

The interrelationship among systemic inflammatory response syndrome (SIRS), sepsis, and infection.

(From Bone RC, Balk RA, Cerra FC, et al: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis, Chest 101:1644–1655, 1992.)

The basic theory behind the development of MOF and the closely related ARDS has undergone modification since the 1970s and mid 1980s. Moore and Moore described the earlier models, which promoted an infectious basis for ARDS and MOF, with two possible scenarios: (1) insult → ARDS → pulmonary sepsis → MOF or (2) insult → sepsis → ARDS and MOF. Current thinking promotes an inflammatory model of MOF with an inflammatory response from a number of infectious and noninfectious stimuli. Two patterns exist: the one-hit model (massive insult → severe SIRS → early MOF) and the more common two-hit model (moderate insult → moderate SIRS → second insult → late MOF). Research into the pathogenesis of MOF has focused on how the inflammatory response is propagated independent of infection. Moore and Moore have the global hypothesis that postinjury MOF occurs as the result of a dysfunctional inflammatory response. Deitch created an integrated paradigm of the mechanisms of MOF. In general, three broad overlapping hypotheses have been proposed in the pathogenesis of MOF: (1) macrophage cytokine hypothesis, (2) microcirculatory hypothesis, and (3) gut hypothesis. Further understanding of MOF must extend to the cellular and molecular levels.

Organ injury in MOF is largely caused by the host’s own endogenously produced mediators and less caused by exogenous factors such as bacteria or endotoxins ( Table 23-4 ). There is increasing evidence that biologic markers for the risk of development of MOF may be more useful than anatomic descriptions of injuries. Nast-Kolb and associates measured various inflammatory markers in a prospective study of 66 patients with multiple injuries (injury severity score [ISS] >18) and found that the degree of inflammatory response corresponded with the development of posttraumatic organ failure. Specifically, lactate, neutrophil elastase, interleukin-6 (IL-6), and IL-8 were found to correlate with organ dysfunction. Strecker and coworkers studied 107 patients prospectively and found that the amount of fracture and soft tissue damage can be estimated early by analysis of serum IL-6 and creatine kinase and is of great importance with regard to long-term outcome after trauma. These investigators found significant correlations between fracture and soft tissue trauma and ICU stay; hospital stay; infections; SIRS; MOF score; and serum concentrations or activities of serum IL-6, IL-8, and creatine kinase during the first 24 hours after trauma.

TABLE 23-4


Humoral Mediators

  • Complement

  • Products of arachidonic acid metabolism: lipoxygenase products, cyclooxygenase products

  • Tumor necrosis factor

  • Interleukins (1–13)

  • Growth factors

  • Adhesion molecules

  • Platelet activating factor

  • Procalcitonin

  • Procoagulants

  • Fibronectin and opsonins

  • Toxic oxygen free radicals

  • Endogenous opioids-endorphins

  • Vasoactive polypeptides and amines

  • Bradykinin and other kinins

  • Neuroendocrine factors

  • Myocardial depressant factor

  • Coagulation factors and their degradation products

Cellular Inflammatory Mediators

  • Polymorphonuclear leukocytes

  • Monocytes and macrophages

  • Platelets

  • Endothelial cells

Exogenous Mediators

  • Endotoxin

  • Exotoxin and other toxins

Source: Adapted with permission from Balk RA: Pathogenesis and management of multiple organ dysfunction or failure in severe sepsis and septic shock, Crit Care Clin 16(2):337–352, 2000.

Blood transfusions are a frequent part of polytrauma treatment and an independent risk factor for MOF. Zallen and associates have identified the age of packed red blood cells (PRBCs) to be a risk factor, with the number of units over 14 days and 21 days as independent risk factors for MOF. Old but not outdated PRBCs prime the neutrophils for superoxide production and activate the endothelial cells, which are pathogenic mediators for MOF.

Multiple organ failure is a syndrome distinct from respiratory failure that can complicate airway injury, resuscitation, or anesthesia after an accident. With the development of improved patient categorization, transport, and emergency care, it has become recognized that there is a threshold beyond which the survival from injury is problematic. With simple injuries (e.g., an ankle fracture and laceration from a fall), the physiologic effects are not additive. However, in high-energy blunt trauma, the systemic effects—for example, of a pulmonary contusion, ruptured spleen, and fractured pelvis—become more than additive.

The ISS, used to quantify the extent of trauma, was derived from the abbreviated injury score (AIS) of the American Medical Association Committee on Medical Aspects of Automotive Safety, which was updated in 1985 as AIS-85. Injuries to six body regions (head and neck, face, chest, abdomen and pelvic viscera, extremities and bony pelvis, and integument) are graded as (1) mild, (2) moderate, (3) severe, (4) critical—outcome usually favorable, and (5) critical—outcome usually lethal. The ISS equals the sum of the squares of the three highest AIS grades. The ISS score has a maximal value of 75.

When the ISS is 25 or greater, the patient is at risk for MOF and will benefit from specialized trauma center care. The median lethal ISS scores have been determined by age group (in years): ages 15 to 44 years, an ISS of 40; ages 45 to 64 years, an ISS of 29; and ages 65 years and older, an ISS of 20. Moore and Moore identified the following variables to be predictive of MOF: age older than 55 years, ISS of 25 or greater, more than 6 units of blood in the first 24 hours after admission, high base deficit, and high lactate level. These investigators stratified patients at risk for MOF ( Table 23-5 ).

TABLE 23-5


Category Risk Factors MSOF Probability (%)
I ISS 15–24 4
II ISS ≥25 14
III ISS ≥25 plus >6 U RBCs/first 12 hr 54
IV ISS ≥25 plus >6 U RBCs/first 12 hr plus lactate ≥2.5 mmol at 12–24 hr 75

ISS, injury severity score; MSOF, multiple system organ failure; RBC, red blood cell.

Source: Moore FA, Moore EE: Evolving concepts in the pathogenesis of postinjury multiple organ failure, Surg Clin North Am 75:257–277, 1995.

One of the consequences of MOF is the depletion of body protein reserves. Amino acids are essential components of the energy systems that maintain the body’s homeostasis; this deficit cannot be replenished by IV glucose or lipids. As MOF progresses, the peripheral metabolic energy source switches from the conventional energy fuels of glucose, fatty acids, and triglycerides to the catabolism of essential branched-chain amino acids. The multiple-injury patient is like a diesel submarine on the bottom of the ocean with a limited air supply. When the air supply is exhausted, damage control systems can no longer be maintained. Amino acids are lost as muscles are oxidized for energy, and the supply is not replenished.

Tscherne emphasized the role of necrotic tissue in the pathogenesis of MOF. It is well known that a gangrenous limb, for example, can provoke a systemic catabolic response with the dramatic reversal of alarming symptoms when an urgent amputation is undertaken for gangrene. Pape and colleagues noted the importance of soft tissue injuries (extremities, lung, abdomen, and pelvis), which create a pathophysiologic cascade after blunt trauma.

Dead tissue (e.g., muscle, bone marrow, and skin) provokes an inflammatory autophagocytic response. In this setting, consumption of complement and plasma opsonins has been measured. The complement system is activated with depletion of factors C3 and C5 with elevated levels of C3a and increased metabolism of C5a. C3a and C5a are anaphylatoxins and may cause the pulmonary edema in ARDS by affecting the smooth muscle contraction and vascular permeability. Plasma opsonin activity is decreased with the consumption of the complement system. The opsonins are critical for antibacterial defense, and their consumption may lead to an increased susceptibility to infection. Several investigators identified serum factors that stimulate muscle destruction. Multiple mediators and effectors have been implicated in the pathogenesis of MOF, but exactly which mediator or combination of mediators is responsible for the hypermetabolic response is not known. This response consumes the individual’s energy reserve and leads to MOF. When MOF is established, the sequence of organ failure apparently follows a consistent pattern, with involvement first of the lung and then the liver, gastric mucosa, and kidney.

Positive blood culture results have been documented in 75% of patients with MOF, but it is not clear whether infection is the cause or simply accompanies MOF. Goris and associates were able to induce MOF in rats by injecting a material that causes an inflammatory response. Sepsis causes tissue destruction and, therefore, similar to broken bones, releases activators of autophagic systems into the blood stream.

The immune system’s response in polytrauma can be measured. Polk and colleagues defined a scoring system for predicting outcome by combining points for ISS and contamination with a measurement of monocyte function (the surface expression of D-related antigen). This method appears promising in predicting survival.

Because patients with identical injuries can have vastly different inflammatory responses after severe trauma, there may be a genetic predisposition for a compromised immune system after trauma. Hildebrand and colleagues performed a prospective cohort study of patients with an ISS greater than 16 and noted that the IL-6-174G/C polymorphism was associated with the severity of the posttraumatic systemic inflammatory response. These authors concluded that there may be a genetic predisposition to an enhanced inflammatory response after polytrauma that may be associated with an adverse outcome.

A multiple system approach to the multiple-injury patient has proved valuable in preventing the development of MOF ( Table 23-6 ). Avoidance of pulmonary failure, prevention of sepsis, and nutritional support are the keys. Mechanical ventilation is regulated in a special care unit under the supervision of anesthesiologists or traumatologists with experience and training in this area of intensive care. Immediate wound débridement and constant attention to the details of wound management, pulmonary toilet, cleanliness of access lines, and urinary tract sterility are required to prevent sepsis. With open fractures, parenteral or local wound antibiotics are therapeutic. An assessment of nutritional reserve, including measurement of triceps skin fold, total lymphocytes, and serum transferrin, is helpful in determining the need for nutritional support. If possible, the GI tract should be used, but in patients with extensive intraabdominal injury and poor nutritional reserves, early total parenteral nutrition with amino acid supplementation is essential. Nutrition has an important role in preventing the translocation of bacteria and toxins from the gut into the splanchnic circulation.

TABLE 23-6


Resuscitative Phase

  • Aggressive volume resuscitation in early stages of treatment

  • Appropriate monitoring of volume resuscitation with measurement of arterial base deficit and serum lactate level, use of pulmonary artery catheters, calculation of oxygen delivery and consumption, use of gastric tonometry

Operative Phase

  • Timely operative management of soft tissue injuries with débridement of nonviable and infected tissue

  • Early fixation of all possible long bone and pelvic fractures

  • Vigilance in preventing the missed injury

Intensive Care Unit Phase

  • Early nutritional support

  • Appropriate use of antibiotics

  • Specific organ support

  • Timely reoperative surgery for missed injuries and complications of trauma

Orthopaedic Management

Early total care of significant pelvic, spinal, and femoral fractures can have a powerful role in avoiding the cascade of events leading to pulmonary failure, sepsis, and death. Increased understanding of the metabolic consequences of fracture surgery further clarifies the timing of orthopaedic surgery in polytrauma patients; specifically, when is early total care of fractures safe, and when is DCO useful? There is optimism that the DCO approach to polytrauma patients will decrease the incidence of multisystem organic dysfunction and failure. IM nailing of the femur has been shown to be a “second hit” to the patient. Harwood and colleagues reported that DCO was associated with a lesser systemic inflammatory response. Even though DCO is used in many centers for patients with the lethal triad of acidosis, hypercoagulability, and hyperthermia, there is limited scientific evidence (e.g., randomized, prospective studies) that it is effective. There are many studies that support the early fixation of fractures. Demling stressed control of the inflammatory process to prevent further stimulus to MOF by early rapid removal of injured tissue and prevention of further tissue damage by early fracture fixation.

Overall, early fixation has been shown to decrease rates of respiratory, renal, and liver failure. Seibel and associates showed that in the blunt multiple-trauma patient with an ISS ranging from 22 to 57, immediate internal fixation followed by ventilatory respiratory support greatly reduces the incidence of respiratory failure, positive blood culture results, complications of fracture treatment, and MOF. When patients were treated with the same ventilatory support but with 10 days of traction before fracture fixation, pulmonary failure lasted twice as long, positive blood culture results increased 10-fold, and fracture complications increased by a factor of 3.5. If no ventilatory support was used and traction was used for 30 days, pulmonary failure lasted three to five times as long, positive blood culture results increased by a factor of 74, and fracture complications increased by a factor of 17. Carlson and coworkers demonstrated that fixation in less than 24 hours after injury versus nonoperative fracture management decreased the late septic mortality from 13.5% to less than 1%. In a series of 56 multiple-injury patients, Goris and colleagues showed that the advantage of controlled ventilation combined with early fracture fixation was greater than that of either ventilation or fracture fixation alone. The greatest advantage was observed in patients with an ISS of more than 50.

Meek and coworkers retrospectively studied 71 multiple-trauma patients with similar age and ISS with respect to timing of fracture stabilization. The group with long bone fractures stabilized within 24 hours had a markedly lower mortality rate than the group treated with traction and cast methods. In a prospective study, Bone and associates compared the incidence of pulmonary dysfunction in 178 patients with acute femoral fractures who underwent either early (in the first 24 hours after injury) or late (>48 hours after injury) stabilization. The patients were further divided into those who had multiple injuries and those with isolated fracture of the femur. In none of the patients with isolated femoral fractures, whether treated with early or late stabilization, did respiratory insufficiency, required intubation, or needed placement in the ICU occur. In the patients with multiple injuries, those who had delayed stabilization of fractures had a significantly higher incidence of pulmonary dysfunction.

Early femoral fixation may not play as critical a role in the outcome, however, in an aggressively managed surgical ICU. Reynolds and associates studied 424 consecutive patients with femur fractures treated with IM rods, and half of these were done in the first 24 hours. Of these 424 patients, 105 had an ISS of 18 or greater; these patients were studied for the relationship of fracture, fixation, timing, and outcomes. IM fixation was done in the first 24 hours in 35 of 105, between 24 and 48 hours in 12 of 105, and after more than 48 hours in 58 of 105. A few days’ delay in fracture fixation did not adversely affect outcome, and pulmonary complications were related to the severity of injury rather than to timing of fracture fixation. Indeed, “fracture fixation” is a generic term for everything from open surgical intervention with extensive blood loss on a subtrochanteric fracture to the placement of a few screws through percutaneous incisions or the rapid assembly of a fixator. Only prospective research can identify which variables are truly important—anesthesia, blood loss, necrosis, ventilation, and micromotion, to name a few. The physiologic consequences of medullary nailing of the femur are the best known. Large trials in injured patients are needed to compare femoral nailing with other fixation methods.

The type of femoral fixation may play a role in risk consideration. With IM nailing, there is the risk of additional bone marrow emboli and potential associated lung dysfunction. Pape and associates found ARDS in a higher percentage of patients treated with a reamed IM femoral nail acutely performed (eight of 24, 33%) versus delayed nailing (two of 26, 8%) in patients with femur fractures and severe thoracic injuries. Charash and coworkers repeated the Pape study design and reported contradictory findings with favorable results in acute reamed IM nailing versus delayed nailing: pneumonia (14% vs. 48%) and pulmonary complications (16% vs. 56%). Bosse and coworkers studied severe chest-injured patients with femur fracture treated within 24 hours with reamed IM nail or plating. The retrospective study was controlled for group A, femur fracture with thoracic injury; group B, femur fracture with no thoracic injury; and group C, thoracic injury with no femur fracture. The overall ARDS rate in patients with femur fractures was 10 of 453 (2%). There was no significant difference in ARDS or pulmonary complications or MOF whether the femur fracture was treated with rodding or plating. Bosse and associates found no contraindication for reamed femoral nailing in the first 24 hours even if a thoracic injury was present. Pape and associates assessed lung function in two groups of patients undergoing early (≤24 hours) IM femoral nailing. One group had femoral nailing after reaming of the medullary canal (RFN), and the other group had a small-diameter solid nail inserted without reaming (UFN). These investigators found that lung function was stable in UFN patients but deteriorated in RFN patients. They concluded that IM nailing after reaming might potentiate lung dysfunction, particularly in patients with preexisting pulmonary damage such as lung contusion. In contrast, Heim and associates, in a rabbit model, compared reamed versus unreamed nailing of femoral shaft fractures and showed that both techniques resulted in bone marrow intravasation and resulting pulmonary dysfunction. In sum, the pathomechanics of FES with respect to what causes the lung injury, what mode of femur fixation is best, and when to do the fracture repair has yet to be precisely unraveled. Really good evidence-based investigation in humans is required to solve this problem. Recent research regarding central mechanisms of bone regulation suggests that fracture repair depends not only on local but also on central mediators. The response of the organism as a whole to the noxious effects of injury may be needed to optimize repair systems.

When treating the patient with overt MOF, recognize that a potentially lethal condition is present and that the usual methods to control specific complications (e.g., pneumonia, renal failure, GI bleeding) will be ineffective. The patient must be assessed as a whole. Focused intervention is required to turn the situation around. In such conditions, a delay in performing operative procedures may be unwise. Significant unstable long bone, pelvic, and spinal fractures can be stabilized. However, long bone stabilization may need to be performed by means of DCO (temporary spanning external fixator) to avoid creating a “second hit” to the patient that would worsen recovery from MOF. There is a role for the use of DCO with patients who already have MODS or who are thought to be at risk for its development. With patients with diagnosed MSOS, DCO allows the possibility of skeletal stabilization of femoral shaft and tibial shaft fractures and unstable knee and ankle joints without contributing an additional fatal insult or “hit” to the patient’s physiology from an invasive orthopaedic procedure. Just simply placing a fixator does not solve the physiologic problem. External fixation that provides adequate stability at the fracture site is important. Until we control for quality, we may not be able to assess the value of external fixation for immediate fracture care. In patients who are at risk of developing MOD, DCO provides a way to provide skeletal stabilization of the above injuries without causing an additional physiologic “insult” or “hit” that might cause the patient to develop MODS. DCO treatment has been reported to be associated with lesser systemic inflammatory response than early total care for femur fractures. Patients often require blood, calories (preferably enteral when feasible), effective antibiotics, controlled ventilation, and dialysis. All of these must be continuously monitored. Keel and Trentz stated that the development of immunomonitoring will help in the selection of the most appropriate treatment for polytrauma patients. Treatment measures for MOF often require careful balancing of risks and benefits. If anticoagulation prevents the propagation of thrombi, it can also be a cause of bleeding.

The orthopaedist’s role is to assess fractures that are causing continued recumbency and to locate sources of devitalized tissue and sepsis in the musculoskeletal system. Sacrifice of a crushed but viable limb, loss of fracture reduction, or performance of a quick but not optimal limb stabilization are examples of the difficult choices or procedures that have to be made to save a life. The use of spanning external fixation (traveling traction), so-called DCO, is a good option because of its minimal additional tissue trauma, provisional bony stability, and improved ability to mobilize the patient.

In summary, in the presence of major thoracic and head injuries, there are potential risks of worsening a brain injury or precipitating ARDS from early orthopaedic procedures. Carlson and Velmahos and associates have shown no added morbidity for early fixation when chest or head injuries are present. However, Townsend and colleagues and Pape report increased risk for secondary brain injury and ARDS associated with early fixation. Reynolds and associates showed that a modest delay did not affect the outcome. Dunham and colleagues noted no difference between early and late fracture fixation. Bhandari and colleagues stated that head injury does not seem to be a contraindication to reamed IM nailing. Giannoudis and colleagues noted that the literature does not provide clear-cut guidelines for the management of orthopaedic injuries in head-injured patients. These authors stated that it was best to individualize treatment. Finally, as Deitch and Goodman have noted, the best way to treat MOF is the prevention of MOF in the first place.

Local Complications of Fractures

Local complications or unwanted therapeutic outcomes are simply part of taking care of fractures. Local failures of fracture treatment can manifest as immediate, delayed, or long-term adverse outcomes. Delayed complications include CRPS and disuse atrophy—the “fracture disease.” Arthrosis and malunion are examples of the long-term adverse results with permanent impairment and economic importance. Any treatment program, no matter how thoughtfully conceived and carefully performed, has a failure rate that cannot be entirely eliminated. The patient, physician, and system variables inherent in each given clinical situation mean that, in practice, complication rates are usually in excess of those rates published in the literature. With multiple injuries, the rates become more than additive. This is expressed in the ISS by adding the squares of injury components. The purpose of this section is to provide a framework for understanding local complications of fractures.

Soft Tissue and Vascular Problems

An accident, unlike an elective operation, causes the transmission of force of an undetermined magnitude to human tissue. However, through accident reconstruction, it is possible to estimate the magnitude of energy transfer. For example, a fall from 30 ft is equivalent to being struck by a car going 30 mph. In the immediate hours, days, and weeks after an injury, it should not be surprising that areas of skin demarcation, skin sloughing, ecchymosis, or thrombosed vessels appear. These areas, if operative interventions are appropriate, are the consequences of injury and not of its treatment. In addition, after an osteosynthesis, there is additional opportunity for the slow accumulation of hematoma from bleeding from bone surfaces. Today’s shortened hospitalizations with early mobilization and less control over the postoperative activity have contributed to an increased incidence of postoperative hematoma. Postoperative hematoma manifests as swelling, pain; loss of function; and frequently, serous drainage either from a wound or from the drain tract. Hematomas may not resorb but instead continue to increase in size and cause wound separation, skin sloughing, and infection. Collections of blood or fluid can be detected with ultrasonography. It is usually best to reexplore the wound under adequate anesthesia, evacuate the hematoma, irrigate the fracture site, drain the field, and apply a compression dressing.

Arterial injuries may manifest acutely with signs of hemorrhage and ischemia, or the presentation may be delayed, as in an arteriovenous fistula or a pseudoaneurysm. In civilian injuries with associated fractures or dislocations, the arterial injury rate is from 2% to 6%. For isolated fractures or dislocations, the arterial injury rate is less than 1%. War-related extremity injuries, a high proportion of which result from high-velocity gunshot wounds, consist of a long bone fracture with associated vascular injury in about one third of cases. Certain injury patterns such as a knee dislocation, especially in a posterior direction, have a 30% incidence of associated popliteal artery damage. Although standard angiography was the preferred diagnostic test for vascular injury, computed tomography angiography (CTA) is rapidly supplanting angiography. CTA avoids the hazards of arterial puncture contrast reactions and provides excellent visualization of the vascular tree. Traditional angiography at most centers has been replaced by CTA.

Rieger and colleagues retrospectively assessed the accuracy of multidetector computed tomography angiography (MDCT) as the initial diagnostic technique to depict arterial injury in patients with extremity trauma. Prospective sensitivity and specificity were 95% and 87%, respectively, and retrospective sensitivity and specificity were 99% and 98%, respectively. Inaba and colleagues also studied the ability of multislice helical computed tomography angiography (MCTA) to detect arterial injury in the traumatized. MCTA achieved 100% sensitivity and 100% specificity in detecting clinically significant arterial injury. No missed injuries were identified during the follow-up period, which was a mean of 48.2 days. Despite these reported accuracies, there are reported concerns about the limitations of MDCT such as that reported by Portugaller and colleagues, who noted lower sensitivity in the infrapopliteal area caused by small vessel diameter.

Computed tomography angiography is clearly the wave of the future for imaging the vascular system. Improvements in technique and accuracy of interpretation are likely to strengthen the available evidence to support its widespread application.

When femur fractures are associated with femoral artery injury, the results of vascular repair and limb function are characteristically good, although delayed diagnosis of pseudoaneurysm and claudication can be a problem. Cases requiring amputation because of a delay in diagnosis can occur. Popliteal artery injury, even when diagnosed early, may not be amenable to vascular reconstruction. Vascular injuries distal to the popliteal trifurcation are basically not fixable and carry a much worse prognosis. Revascularization is usually not needed if one vessel is patent on angiography and the distal pressure is 50% of the brachial artery pressure. When revascularization is required, a good functional result can be expected in only about 25% of cases. In Flint and Richardson’s experience, six of 16 patients undergoing revascularization distal to the trifurcation required early amputation, and six more required late amputation (total, 12 of 16) for osteomyelitis, nonunion, and persistent neuropathy and its associated complications. Early amputation without revascularization is often appropriate in patients with loss of vascular inflow, long bone fracture, neurotmesis, or extensive soft tissue damage. The high rate of infection and complications can result in a delayed amputation when revascularization is performed. In general, amputation after the onset of infection is at a higher anatomic level than would have been selected had the amputation been performed initially.

Posttraumatic Arthrosis

Posttraumatic arthrosis is considered a complication of fractures ( Fig. 23-6 ). Wright, a retired judge, used a questionnaire to determine a consensus view of the factors related to the development of posttraumatic arthrosis after fracture. He found the following: that lower limb joints are more likely to develop arthritis than upper extremity joints, that older patients are at higher risk for the development of posttraumatic arthrosis (although younger patients have a longer time frame to develop posttraumatic arthrosis), and that occupation is a risk factor. Kern and associates also reported the association of osteoarthritis and certain occupations. Specific components of the pathomechanics of posttraumatic arthrosis include (1) incongruity of the articular surface, (2) cartilage damage from the load transfer, (3) malalignment, (4) malorientation of the joint, and (5) repetitive loading injury.

Jun 11, 2019 | Posted by in ORTHOPEDIC | Comments Off on Diagnosis and Treatment of Complications

Full access? Get Clinical Tree

Get Clinical Tree app for offline access