Key points

Introduction

The approach to treating major fractures in patients who have sustained severe trauma has evolved steadily over the past 40 years. Initially, patients who had sustained substantial trauma were viewed as too sick to undergo fracture surgery, but as increasing evidence was published linking prolonged recumbency with morbidity, these same patients became viewed as too sick not to undergo surgery. However, as the enthusiasm for early fracture fixation in these patients increased, it became clear that some patients paid a different physiologic price for too aggressive of a surgical approach after major trauma: multiorgan failure, particularly adult respiratory distress syndrome (ARDS). This paradoxic situation illuminated the need for a way to stabilize fractures early while minimizing physiologic insult. From this need was born the concept of damage control orthopedics (DCO).

Introduction

The approach to treating major fractures in patients who have sustained severe trauma has evolved steadily over the past 40 years. Initially, patients who had sustained substantial trauma were viewed as too sick to undergo fracture surgery, but as increasing evidence was published linking prolonged recumbency with morbidity, these same patients became viewed as too sick not to undergo surgery. However, as the enthusiasm for early fracture fixation in these patients increased, it became clear that some patients paid a different physiologic price for too aggressive of a surgical approach after major trauma: multiorgan failure, particularly adult respiratory distress syndrome (ARDS). This paradoxic situation illuminated the need for a way to stabilize fractures early while minimizing physiologic insult. From this need was born the concept of damage control orthopedics (DCO).

Changing the nature of stabilization

Before the 1980s, patients who sustained severe multiple traumatic injuries were viewed as being too ill to tolerate the physiologic stressors associated with internal fixation of major fractures. Classically, these patients were treated in a staged fashion, as described by Wolff and colleagues. Generally, patients were first resuscitated and emergency surgery performed for immediately life-threatening conditions. Patients then went through a stabilization phase for several days or weeks in preparation for definitive surgery. Accordingly, patients were placed in traction for a period, until they seemed well enough to undergo fracture fixation. Unfortunately, a patient in prolonged traction is compelled to stay relatively immobilized in a recumbent position, which makes pulmonary toilet challenging, and the lack of fixation is thought to increase the likelihood of pulmonary complications, including thrombotic and fat embolism. With this method of treatment, the final rehabilitative phase is delayed until the fractures have been surgically stabilized or healed in traction, thus leaving patients bedridden and hospital-bound for long periods.

In the early 1970s, some groups began to take a more proactive approach to fracture stabilization. Seibel and colleagues began performing immediate fracture fixation in patients who sustained blunt multitrauma during this period. Anecdotally, the investigators observed a surprising reduction in intensive care unit (ICU) days, pulmonary emboli, and fracture-related complications. Emboldened by these observations, they began to prospectively study the effects of timing of fixation on the clinical course of these patients. Patients with femoral or acetabular fractures and a minimum injury severity score (ISS) of 22 were treated with fixation either within the first 24 hours or after 10 or more days of traction. Patients who underwent delayed fixation had more ICU and ventilator days, had a higher number of positive blood cultures, and required a greater number of antibiotics. The authors concluded that delaying fracture stabilization increased the risk of pulmonary complications from prolonged recumbency and nutritional depletion, narcotic use, and extrapulmonary sources of infection, such as open wounds and necrotic debris within fracture hematomas.

Additional studies were published in the mid to late 1980s, espousing the virtues of early fixation in polytrauma patients in conjunction with early aggressive ventilatory support. However, these studies were frequently heterogeneous with regard to the involved fractures or the associated injury patterns, and the single prospective series during this period had too small of a cohort ( N = 18) for definitive conclusions to be drawn.

The era of early total care

To better evaluate the effects of the timing of fixation on the complications, clinical course, and cost associated with treating femoral fractures, Bone and colleagues conducted a seminal prospective randomized study with 178 patients. Patients were randomized to femoral fracture internal fixation in less than 24 hours or greater than 48 hours. The authors found that in patients with isolated femoral fractures and an ISS less than 18, the timing of fixation had no appreciable effect on outcomes. However, with polytrauma patients (ISS>18), the timing of femoral fracture fixation did seem to play a major role in patients’ clinical courses. Late fixation was associated with a higher incidence of ARDS, fat embolism syndrome, pulmonary emboli, and pneumonia. These patients also had a mean ICU stay and total length of stay that were 5 and 10 days longer, respectively, and an average cost of hospitalization that was 50% higher than that of the early fixation cohort. The study has been criticized for having unequal numbers of lung injuries in the 2 groups, no details of resuscitation, and limited statistical comparisons, but it clearly remains one of the most influential studies on this topic.

Following this paper, additional authors published studies supporting early fixation of all fractures (known as early total care [ETC]) in polytrauma patients. However, some authors reported no benefit with early fixation. In addition, some evidence suggested that although early fixation might be beneficial, fixation in the first 24 hours might be too early and might actually be detrimental to patients. It began to become apparent that in certain patients, particularly those with pulmonary or traumatic brain injuries, a less-aggressive approach to fracture fixation might be needed.

Emergence of the 2-hit model

Over time, the concept of a 2-hit model of posttraumatic physiologic response began to emerge. In the 1-hit model, the initial trauma is believed to generate a major inflammatory cascade with the potential for subsequent ARDS, independent of the timing of fracture fixation. The development of multiple organ failure is dependent on the extent of initial injury and the timing and quality of resuscitation. In contrast, the 2-hit model suggests that the initial trauma generates a less-severe systemic inflammatory state. The immune system then becomes primed for an exaggerated inflammatory response after a second physiologic insult. This second hit was initially thought to be from an infectious process, such as pneumonia or other sources of infection, but some consideration was given to a noninfectious source of this second hit.

Anderson and Harken described a mechanism of lung injury that begins with the priming of neutrophils and macrophages through inflammatory mediators generated by multisystem trauma. Once these cells are primed, a noninjurious or mildly injurious second stimulus causes an excessive inflammatory reaction that leads to cell-mediated lung injury. The authors suggested that inhibiting this cellular priming could be an effective prophylaxis against secondary lung injury via this 2-hit phenomenon.

The work of Bone and colleagues and of Waydhas and colleagues supports this notion that inflammatory mediators play an important role in the response to polytrauma. In their series of patients, they prospectively followed 3 markers for inflammation: C-reactive protein, neutrophil elastase, and platelet count. They found that with 2 abnormal parameters, the probability of postoperative major organ failure was 73%, and if only 1 or no parameters were abnormal, that probability decreased to 17%. On further review of their data, the investigators found that nearly 80% of patients with postoperative organ failure despite low inflammatory markers had undergone extensive surgeries, such as fixation of pelvic or femoral fractures. Conversely, most patients (71%) who had a benign clinical course in the face of elevated inflammatory indicators had undergone mild procedures, such as fixation of facial fractures. Although no difference was seen in the timing of fixation between patients with and without organ failure, the authors concluded that surgeries more than 72 hours after trauma may induce a second-hit phenomenon and organ failure.

Eventually, it became apparent that the 2-hit model was not an all-or-none phenomenon, and that some patients could undergo early extensive fracture fixation without developing postoperative organ failure. Pape and colleagues reported on a retrospective series of 106 patients with an ISS greater than 18 who underwent reamed nailing of femoral shaft fractures. Patients who had an Abbreviated Injury Scale (AIS) thorax score of 2 or greater and underwent nailing in the first 24 hours after trauma had a significantly higher incidence of ARDS, had significantly longer ventilator and ICU times, and trended toward a higher incidence of pneumonia compared with a cohort that had an AIS-thorax score of 1 or less. The authors posited that in the setting of a severe pulmonary contusion, the normal compensatory mechanisms of the lung may become taxed to their limits. The authors termed this a borderline situation , wherein the patient can no longer accommodate the added physiologic insult of prolonged surgery and intramedullary instrumentation. The authors then stated that the extent of pulmonary injury can be difficult to gauge in the early postinjury period. Accordingly, they recommended using external fixation or perhaps unreamed nailing in severely injured patients with associated thoracic trauma, despite having no unreamed nails in their series.

Reamed femoral nailing has been shown to increase lung capillary permeability in animal models and to increase pulmonary arterial pressures in humans when compared with unreamed femoral nailing. Unreamed femoral nailing has been shown to be safe when performed in the first 24 hours after trauma, despite severe thoracic injury (AIS≥3). However, early (<24 hours) reamed femoral nailing has not been shown to increase the incidence of ARDS in vivo, even when performed in patients with an AIS-thorax score of 4 or greater. With that in mind, reamed nailing has been shown to increase certain inflammatory markers to nearly significant levels (s-ICAM-1, P = .052; CD11b, P = .08), which implies that the mode of initial fixation in borderline patients, not only the timing thereof, warrants some consideration.

Damage control orthopedics

It became increasingly evident that early fracture fixation needed to be performed while minimizing the secondary physiologic insult to the patient. After Townsend and colleagues reported on hypotension observed in patients with brain injuries who underwent early femoral nailing, a question to the authors referenced the concept of “resuscitative orthopedics,” specifically “minimal stabilization rather than definitive fixation.” Two years later, Scalea and colleagues reported on a 3-year series of patients who underwent a similar limited initial approach to femoral shaft fractures in polytrauma patients, terming it damage control orthopedics (DCO). The authors credited Rotondo and colleagues with applying the naval war term damage control to the limited initial treatment of penetrating abdominal trauma.

Scalea and colleagues described damage control in 43 patients with femoral fractures and a median ISS of 27. These patients underwent early temporary external fixation for a median of 4 days before staged nailing. This cohort was compared with 281 patients who underwent primary femoral nailing with a median ISS of 17 ( P = .001). The median operative time for external fixation was 35 minutes, compared with 135 minutes for primary intramedullary nailing. Despite a significantly higher incidence of laparotomy, AIS-head score of 3 or greater, shock on presentation, and days in the ICU, the damage control cohort had only 4 deaths (9% vs <1%; P = .001). The authors believed that femoral fractures should be fixed as early as possible but that some patients were not physiologically replete enough to withstand early nailing. Paradoxically, this subset of patients would be poor candidates for prolonged traction, and thus external fixation was a bridge to definitive treatment, affording the advantages of fixation with minimal operative time and physiologic insult. Before this era, external fixation of adult femoral fractures was uncommon and typically used only for definitive cases. Temporary external fixation of femoral fractures had been described, but mainly for soft tissue management, not for resuscitative reasons.

Clearly, minimizing the interval between provisional and definitive fixation has a direct effect on the length of hospitalization. However, to mitigate the effects of a hyperstimulated immune response, delaying major fracture surgery in the first few days after injury seems to be advantageous in avoiding the second-hit phenomenon in certain polytrauma patients. Specifically, delaying secondary major fracture fixation until the fifth day after injury seems to have a significantly protective effect in terms of the inflammatory response and pulmonary and hepatic dysfunction. The clinical ramifications of this are debated, but some clinicians delay definitive fixation after DCO until the fifth day based on this finding.

After its initial description in the literature, several papers on DCO were published, extolling its virtues in the multitrauma patient. Before the published description of DCO, the Department of Orthopaedics and Trauma Surgery at Hannover Medical School had been performing staged fixation of femoral shaft fractures in polytrauma patients at risk for pulmonary complications for almost 10 years. In 2002, Pape and colleagues published a retrospective comparison of complication rates in this patient population before and after the implementation of this protocol. They found that the incidence of ARDS with primary nailing decreased from 55% to 27% after they began practicing DCO in selected patients. Moreover, the ARDS rate associated with primary external fixation decreased from 97% to 22%, reflecting the use of this technique in borderline patients, as opposed to solely in patients in extremis.

The following year, Pape and colleagues published a prospective, randomized, multicenter study comparing DCO with primary nailing of femoral shaft fractures in polytrauma patients. They found a significant increase in serum proinflammatory markers when primary nailing was performed, although they were unable to demonstrate an association with postoperative complications. External fixation was not associated with an increase in these markers and, interestingly, neither was the secondary nailing, performed a mean of 3 days later. This lack of a significant inflammatory response may have been because of a relatively quiescent inflammatory milieu or a type II error from a small sample size.

Early appropriate care

Nahm and colleagues recently coined the term early appropriate care (EAC) to describe the preferential fixation of femoral fractures in the first 24 hours in contrast to other extremity fractures that could be splinted and fixed at a later date. This procedure would provide the benefits of early treatment while minimizing the length of the original surgery—a compromise between the staged treatment of DCO and the fixation of all fractures in ETC—and could be used in most patients, provided an aggressive approach to resuscitation was used.

The staged method of DCO has some potential negative aspects, mainly the possibility of infection after external fixation and the need for additional surgery. The risk of secondary infection with staged nailing of femoral fractures has been estimated to be less than 3%, even with many weeks of delay between external fixation and definitive fixation. The infection risk for the tibia initially treated in external fixation is believed to be higher, at approximately 9%, but this is less of an issue because closed tibias can be more easily splinted in borderline patients, and little of the current association with pulmonary dysfunction of nailing of femoral fractures is thought to apply to tibial fractures. As for the need for additional surgery, a staged nailing can be expected to increase the cost of hospitalization as a result of the cost of the external fixator and the potential increase in hospital length of stay (LOS). Although some evidence shows that DCO may not increase the LOS despite the additional surgical procedure, evidence also exists to the contrary.

Although use of DCO has been reported in up to 35% of femoral fractures in polytrauma patients (ISS≥18), DCO has been sparingly used in other series describing similar patient cohorts. Despite using DCO in only 12% of femoral fractures with an ISS of 18 or greater, O’Toole and colleagues reported rates of postnailing ARDS and mortality of only 1.5% and 2.0%, respectively. When comparing their findings with the ARDS rate of 26% described by Pape and colleagues, the authors noted that their patients underwent primary nailing or DCO an average of 13 hours after admission, whereas virtually all of the patients in the series by Pape and colleagues underwent operative treatment in less than 8 hours.

Similarly, Nahm and colleagues observed an ARDS rate of 1.7% in 492 patients with an ISS of 18 or greater who underwent femoral nailing in the first 24 hours. Patients with an AIS-thorax of 3 or greater who underwent early nailing experienced a 5.7% incidence of ARDS. The authors emphasized an aggressive resuscitation protocol, including the serial monitoring of blood pH and base deficit or lactate, throughout the preoperative period and during surgery.

The reason for the lower rate of ARDS observed after femoral nailing in polytrauma patients in multiple North American series compared with the German series is unknown. One possibility is that the German patients had worse lung injuries. O’Toole and colleagues posited several other potential differences, such as genetic predisposition for ARDS, smoking rates, and differences in resuscitation or ICU care.

Patients with severe head injuries and concomitant extremity fractures constitute a particular subset of polytrauma patients, in that the possible hypotension or hypoxia associated with prolonged surgery may be especially deleterious. The principles of EAC nonetheless seem to also apply to them. Jaicks and colleagues found that early fixation of orthopedic injuries with an AIS-head of 2 or greater was associated with increased fluid requirements intraoperatively and in the immediate postoperative period, but they found no difference in neurologic complications during the initial hospitalization. Townsend and colleagues observed a 2- to 8-fold increase in the incidence of intraoperative hypotension when femoral internal fixation was performed in the first 24 hours, although the Glasgow Coma Score (GCS) on discharge seemed to be related to the primary injury rather than the timing of fixation. Additional authors also found no effect of femoral internal fixation timing on neurologic outcomes in patients with severe head injuries, although resuscitation must be tailored to maximize cerebral perfusion pressures and oxygenation.