Secondary Logo

Journal Logo

External Fixation as a Bridge to Intramedullary Nailing for Patients with Multiple Injuries and with Femur Fractures: Damage Control Orthopedics

Scalea, Thomas M. MD; Boswell, Sharon A. RN, CEN; Scott, Jane D. ScD, MSN; Mitchell, Kimberly A. MS; Kramer, Mary E. RN; Pollak, Andrew N. MD

The Journal of Trauma: Injury, Infection, and Critical Care: April 2000 - Volume 48 - Issue 4 - p 613-623
ANNUAL MEETING ARTICLES
Free

Background: The advantages of early fracture fixation in patients with multiple injuries have been challenged recently, particularly in patients with head injury. External fixation (EF) has been used to stabilize pelvic fractures after multiple injury. It potentially offers similar benefits to intramedullary nail (IMN) in long-bone fractures and may obviate some of the risks. We report on the use of EF as a temporary fracture fixation in a group of patients with multiple injuries and with femoral shaft fractures.

Methods: Retrospective review of charts and registry data of patients admitted to our Level 1 trauma center July of 1995 to June of 1998. Forty-three patients initially treated with EF of the femur were compared to 284 patients treated with primary IMN of the femur.

Results: Patients treated with EF had more severe injuries with significantly higher Injury Severity Scores (26.8 vs. 16.8) and required significantly more fluid (11.9 vs. 6.2 liters) and blood (1.5 vs. 1.0 liters) in the initial 24 hours. Glasgow Coma Scale score was lower (p < 0.01) in those treated with EF (11 vs. 14.2). Twelve patients (28%) had head injuries severe enough to require intracranial pressure monitoring. All 12 required therapy for intracranial pressure control with mannitol (100%), barbiturates (75%), and/or hyperventilation (75%). Most patients had more than one contraindication to IMN, including head injury in 46% of cases, hemodynamic instability in 65%, thoracoabdominal injuries in 51%, and/or other serious injuries in 46%, most often multiple orthopedic injuries. Median operating room time for EF was 35 minutes with estimated blood loss of 90 mL. IMN was performed in 35 of 43 patients at a mean of 4.8 days after EF. Median operating room time for IMN was 135 minutes with an estimated blood loss of 400 mL. One patient died before IMN. One other patient with a mangled extremity was treated with amputation after EF. There was one complication of EF, i.e., bleeding around a pin site, which was self-limited. Four patients in the EF group died, three from head injuries and one from acute organ failure. No death was secondary to the fracture treatment selected. One patient who had EF followed by IMN had bone infection and another had acute hardware failure.

Conclusion: EF is a viable alternative to attain temporary rigid stabilization in patients with multiple injuries. It is rapid, causes negligible blood loss, and can be followed by IMN when the patient is stabilized. There were minimal orthopedic complications.

From the R. Adams Cowley Shock Trauma Center (T.M.S., S.A.B.), University of Maryland Medical System, Program in Trauma (T.M.S.), Charles McC. Mathias Jr. National Study Center for Trauma and EMS (J.D.S., K.A.M., M.E.K.), and Division of Orthopedics and Program in Trauma (A.N.P.), University of Maryland School of Medicine, Baltimore, Maryland.

Address for reprints: Thomas M. Scalea, MD, R. Adams Cowley Shock Trauma Center, 22 South Greene Street, Room T3R35, Baltimore, MD 21201-1595.

Submitted for publication September 24, 1999.

Accepted for publication December 31, 1999.

This study was not funded from corporate or other sources. The authors contributed their time to the study.

Presented at the 59th Annual Meeting of the American Association of the Surgery of Trauma, September 16–18, 1999, Boston, Massachusetts.

Although the benefits of early fracture fixation are well documented, questions remain concerning the optimal timing of fracture fixation in adult trauma patients with multiple injuries. 1–6 Early intramedullary nail (IMN) fixation of long bone fractures in patients with multiple injuries has been associated with a reduced risk of pulmonary complications. 2,4,6 However, some recent reports have implicated fat embolization associated with IMN worsening pulmonary complications for patients with certain lung injury profiles. 7 Additional studies suggest that patients with severe traumatic head injury may experience poorer outcomes if treatment of musculoskeletal injuries includes early surgical intervention. 8,9 Substantial blood loss associated with major operative procedures certainly complicates optimal fluid resuscitation in the patient with a brain injury, and general anesthesia interferes with the ability to serially assess neurologic function. Additional patient populations at particular risk for complications from major operative procedures include those who are hypothermic, coagulopathic, or hemodynamically unstable.

External fixation (EF) plays an important role in primary management of pelvic fractures in patients with multiple injuries and with competing injuries in many centers. 10,11 In the past 3 years , we have largely adopted the use of EF for femur fixation in unstable trauma patients with multiple injuries. We use EF as a “bridge” or “temporizing device” to achieve the benefits of early fracture stabilization during early resuscitation, and postpone the additional stresses posed by IMN until the patient is stabilized.

The purpose of this study was to investigate the clinical course and outcomes of all adult trauma patients admitted to our center with femur fracture who were treated primarily with EF versus IMN. We were specifically interested in determining the characteristics of the EF and IMN populations, and outcomes of mortality, length of stay, and discharge disposition. We were additionally interested in determining the reasons provided in the medical record for selecting EF as the primary repair procedure. Our central premise is that, although patients with primary EF of the femur should be more severely injured than patients with primary IMN, survival would be comparable for EF and IMN groups.

Back to Top | Article Outline

MATERIALS AND METHODS

The study was a retrospective analysis of prospectively collected trauma registry data of patients admitted to the R. Adams Cowley Shock Trauma Center (STC), the primary adult resource center for trauma in the State of Maryland, from July 1, 1995, to June 30, 1998. Patients were identified through query of the STC trauma repository searching for all cases of acute femur fracture. Trauma repository variables concerning patient characteristics (i.e., age, sex), clinical course, and outcomes (i.e., mortality, discharge destination) were obtained for all identified cases. If primary repair of the femur was designated as EF, medical records were then abstracted to determine the indications for EF and subsequent clinical course. For patients treated with EF, trauma repository data and abstracted medical record data were linked to create a comprehensive picture of clinical course and outcomes.

Study inclusion criteria were acutely injured patients admitted to the STC with a femur fracture, who were treated primarily with either EF or IMN. Cases were defined in terms of femur fracture and repair only. Mechanism of injury included blunt as well as penetrating trauma.

During the study period, a total of 324 patients met our criteria (i.e., femur fracture and primary femur repair). Cases were divided into EF (n = 43 [13%]) and IMN (n = 281 [87%]) groups. The decision regarding which fixation procedure to use was based on the combined judgment of the attending trauma surgeon, orthopedic surgeon, and any other physician specialties consulting on the case.

The entire study population was characterized in terms of patient attributes (i.e., age, sex, mechanism of injury), clinical course, and discharge disposition. Multiple comparisons were then made between EF and IMN groups to determine what systematic differences existed between groups. Groups were compared by age, sex, race, mechanism of injury, Injury Severity Score (ISS), and Glasgow Coma Scale (GCS) score. The groups were also compared in terms of the proportion of patients: admitted “in shock”, i.e., admission systolic blood pressure [SBP] ≤ 90 mm Hg; with Abbreviated Injury Scale (AIS) score of the head of ≥3; and undergoing major nonorthopedic operative procedures in the first 24 hours after admission, i.e., thoracotomy, laparotomy, and craniotomy).

Groups were compared for the type and volume of blood products used in the first 24 hours (i.e., packed red cells, fresh frozen plasma, and platelets). The proportion of patients requiring each product, and the median and interquartile range (IQ range) of volume per infused product were explored.

Additional data were collected through structured medical record abstraction for all EF patients to identify use of invasive monitoring upon admission, i.e., pulmonary artery catheter, intracranial pressure [ICP] monitor; physiologic parameters at admission, i.e., admission serum lactate, central venous pressure, pulmonary wedge pressure, and ICP; and rationale for use of EF as the primary method of fixation. In addition, characteristics of the hospital course and surgical procedures, i.e., operating room [OR] time, OR estimated blood loss, and postoperative complications, were investigated.

Outcome variables were obtained from the STC trauma repository for all patients. Outcomes of interest included intensive care unit length of stay (ICU-LOS), hospital LOS, mortality, and discharge destination. Characteristics of the clinical course and overall hospitalization were compared for EF and IMN groups.

Data were analyzed by using the SAS statistical package. Univariate and bivariate analyses were performed. The Shapiro-Wilk statistic was computed to determine whether data were normally distributed. For normally distributed continuous variables, the Student’s t test was used to determine differences in means between EF and IMN groups. For continuous data that were not normally distributed within each of the two treatment groups, and were unbounded in a positive direction, e.g., intravenous fluids, length of hospital stay, a nonparametric test (Wilcoxon rank-sum test) was used to determine statistically significant differences between the EF and IMN groups. Nonparametric tables include sample medians and IQ range, i.e., the 25th and 75th percentiles of the distribution of data. IQ range provides a measure of data dispersion similar to standard data “range.” However, for small populations that may include “outliers”, i.e., widely divergent values due to rare or unusual cases, IQ is a more stable and reliable estimate as to how data are distributed. Pearson’s χ2 square and Fisher’s exact tests were used to assess levels of association for categorical data. The two-tailed alpha was set at the 0.05 level. 12

Back to Top | Article Outline

RESULTS

Between July of 1995 and June of 1998, a total of 324 trauma patients, ages 11 to 96 years (mean age, 30.5 years) with femur fracture, underwent EF or IMN as the primary fracture fixation procedure (Figs. 1–3). The patients were predominantly male (71%) and white (63%), and the mechanism of injury included motor vehicle crashes (59%), pedestrians struck (6%), motorcycle crashes (11%), and other mechanisms (24%; fall, assault, other). Patients were moderately injured as evidenced by mean admission ISS of 18.1, mean admission GCS score of 13.8, and 20% had AIS-Head scores of ≥3. Five percent of patients presented in shock (defined as admission systolic blood pressure of <90 mm Hg) and 40% required an ICU stay. For patients requiring ICU care, the median ICU-LOS was 9 days (Table 1).

Table 1

Table 1

FIG 1

FIG 1

FIG 2

FIG 2

FIG 3

FIG 3

Back to Top | Article Outline

EF versus IMN

Forty-three patients (13%) were treated with EF. The other 281 patients (87%) underwent IMN as the primary procedure for treatment of their femur fracture. No differences were found between EF and IMN groups in terms of patient age, sex, race, and mechanism of injury. However, the EF group was substantially more seriously injured. The EF group had a mean admission ISS of 26.8 (p = 0.001) and a mean admission GCS score of 11 (p = 0.001). Seventeen percent presented in shock (p = 0.007), and 56% had an admission AIS-Head ≥ 3 (p = 0.001) (Table 2). The proportion of patients requiring major nonorthopedic operative procedures within the first 24 hours was 21% for EF patients versus 3.6% for IMN (p = 0.001) (Table 3).

Table 2

Table 2

Table 3

Table 3

Eighty-three percent of the EF group required an ICU stay (p = 0.001). The proportion of EF patients requiring blood products in the first 24 hours was also substantially higher than the IMN group (p = 0.001). This finding was true for packed red cells, fresh frozen plasma, and platelets. The volume of packed red blood cells administered to the EF group was significantly more than for IMN (p = 0.007). However, among patients receiving blood products, there were no differences between groups in the volume of fresh frozen plasma and platelets (p = NS) (Table 4).

Table 4

Table 4

Back to Top | Article Outline

EF Clinical Course

Medical records were reviewed to determine the indications for the use of EF as the primary method of femur fixation. As many as three reasons were provided per case and included shock/physiologically unstable (65%), major thoracoabdominal injuries requiring emergent therapy (51%), significant head injury (46%), and other injuries severe enough to preclude safe IMN (46%) (Table 5).

Table 5

Table 5

Admission serum lactate levels were obtained for all EF patients. The median serum lactate level was 4.3, with an IQ range of 3.3 to 7. A total of 91% of lactate levels “normalized.” The median number of hours for lactate normalization was 28, with IQ range of 20 to 39 hours (Table 6).

Table 6

Table 6

Twenty-five percent of those treated with EF had a pulmonary artery catheter placed. Although the median initial cardiac index was 4.6 L/min per m 2 and filling pressures were seemingly adequate, the elevated serum lactates suggest inadequate perfusion. Mean cardiac index was resuscitated to 5.9 L/min per m 2 to clear lactate to normal.

ICP monitors were placed in 28% of EF cases. The median opening ICP pressure was 22 mm Hg, with an IQ range was 12.5 to 29.5 mm Hg. Treatments used in treating elevated ICP included mannitol (100%), barbiturates (75%), and/or hyperventilation (75%). ICP remained elevated during the first day, increasing to a median of 27 mm Hg (IQ range, 21–40 mm Hg). None of the patients required craniotomy within the first 24 hours.

The median time between primary EF and secondary IMN was 4 days (IQ range, 2.5–6 days). The median time in the OR for primary EF was 35 minutes. The median OR time for the secondary IMN was 135 minutes (Table 7). No patient had intraoperative hypotension or hypoxia during EF. Postoperative complications occurred slightly more often in patients with EF versus IMN. The only direct complication of EF was bleeding around the pin sites, which was self-limited. One other patient underwent amputation for a mangled extremity after initial EF. He survived. One patient developed hardware failure, and another developed acute osteomyelitis after initial EF. No complications occurred in 86% of patients treated with EF and 91% of patients treated with IMN (Table 8).

Table 7

Table 7

Table 8

Table 8

Seven patients treated with EF never had IMN. One of these patients had a severe closed head injury. The other patient had a severe pelvic crush injury. He died early in his hospital course. Five patients were young adults (ages, 11–16 years), and after further consideration, it was determined that they were skeletally immature. EF was used as definitive fixation in them.

Back to Top | Article Outline

OUTCOMES

We compared clinical outcomes of LOS and discharge disposition for primary EF (n = 43) and primary IMN (n = 281) groups. For patients admitted to the ICU, median ICU-LOS was 11.0 days for EF patients (IQ range, 6–19 days) compared with 8.0 days (IQ range, 4–16 days) for IMN patients (p = 0.06). Overall hospital LOS was 17.5 days for the EF group (IQ range, 8.8–26.5 days), whereas the median hospital LOS was 5.7 for IMN (IQ range, 3.0–10.1days) (p = 0.001) (Table 9).

Table 9

Table 9

There were a total of five deaths among all study patients (n = 324) with four deaths in the EF group (9% vs. < 1%, p = 0.001). One patient died very early. He had a pelvic crush injury and bilateral femur fractures. After application of external fixation to his pelvis and both femurs, he died from organ failure. The three remaining deaths were due to severe and irreversible brain injury and these deaths each occurred at least 10 days after admission. One of these patients remained in external fixation. The other two had an uneventful conversion from EF to IMN, but succumbed later to their head injury. No deaths in the EF group seemed to be secondary to the fracture management selected.

For all patients discharged alive, 48% were discharged directly home, whereas 52% were discharged either to a “rehabilitation hospital” or to another “acute care” setting. However, substantial differences were noted between groups. Patients with primary EF were discharged to a “rehabilitation hospital” or other “acute care” facility 72% of the time, while this was required in only 49% of patients treated with IMN (p = 0.008) (Table 9).

Back to Top | Article Outline

DISCUSSION

The optimal timing of fracture fixation continues to generate considerable controversy. Decisions about the timing of fixation requires balancing the risks of operative stress versus any benefit that can be gained by early fixation. As early as the mid-1970s, reports demonstrating the advantage of early stabilization of long-bone fractures began to surface. 13 In 1985, Seibel et al. defined the important role that bony injuries play in the overall outcome of patients with multiple injuries. 2 In that study, injury severity did not correlate with the magnitude nor the duration of pulmonary failure. Instead, the authors found an association between respiratory failure and the number of days patients spent in skeletal traction. They proposed that the “pulmonary failure septic state” was caused by a combination of factors, including fat embolism syndrome, enforced supine positioning leading to atelectasis with subsequent pneumonia, and an increased need for narcotics to treat pain caused by continual motion at the fracture site.

Seibel et al. also proposed that the fracture hematoma itself served as a metabolic organ stimulating mediator release, ultimately leading to multiple organ failure. 2 They concluded that early fracture fixation decreased many of the pulmonary complications and resulted in better patient outcome. They believed operative fixation performed on the night of admission was technically easier and exposed patients to the operative stress when nutritional and immunologic reserves were at their highest level. Patients in the early fixation group had a lower risk of infection as well as systemic and orthopedic complication rate when compared with the group of patients undergoing delayed fixation.

Other authors have demonstrated very similar results. Bone et al. found increased pulmonary complications, including acute respiratory failure, fat embolism, and pneumonia in patients who had delayed stabilization of fractures. 3 In addition, these patients had longer hospital and ICU stays and increased costs. Early fracture fixation has been shown to decrease the incidence of respiratory failure in patients with major fractures and patients with multiple orthopedic injuries. 6 The salutory effect of early fracture fixation was seen primarily in the most severely injured (ISS > 40). Early fracture fixation may do more than simply prevent acute respiratory failure. Lozman et al. demonstrated better cardiac function in patients undergoing immediate fracture fixation, and Goldstein et al. demonstrated a better overall pulmonary performance in patients undergoing early fixation of pelvic fractures. 14,15 Early fracture fixation may be especially important in femur fractures. They are high-energy injuries and patients often have multiple associated injuries often in the thorax. Charash et al. described a 50% incidence of pulmonary contusions in their study on patients with femur fractures. 4 That group as well as Behrman et al. demonstrated improved outcome with early femoral fixation. 5

More recently, however, the absolute necessity of early fracture fixation has been called into question. In 1995, Reynolds et al. demonstrated that delaying femoral fixation for several days did not seem to effect the patient’s outcome. 16 They postulated that this delay allowed the patients to stabilize. Any deleterious affects that may result from delaying the fixation were offset by increased hemodynamic stability. Similarly, Rogers et al. have shown that stabilization of isolated femur fractures can be delayed up to 72 hours without an increase in complications. 17 In their rural trauma center, emergent fracture fixation increased the number of emergency operations, increased time of surgery, and did not seem to be a wise use of resources.

Patients with long-bone fractures and significant head injuries may require a different treatment algorithm. In 1992, Poole et al. demonstrated no reduction in pulmonary complications when early fracture fixation was used in patients with closed head injuries. 18 Instead, pulmonary complications seemed more closely related to the severity of the head injury and chest injury (if present). In 1997, Jaicks et al. demonstrated poorer central nervous system outcomes in patients with head injuries when the concomitant long-bone fractures were stabilized early. 9 Early fracture fixation was associated with significantly higher intraoperative fluid requirements, and a higher rate of both intraoperative hypotension and hypoxia. They postulated that this led to lower GCS score at the time of discharge, relative to the patients whose fracture fixation was delayed. However, recently Scalea et al. found no difference in discharge GCS score when patients with closed head injuries underwent early fracture fixation. 1 In addition, they found no difference in ICU or hospital LOS or the need for vasopressors, inotropes, or fluid resuscitation when fractures stabilized early.

It would seem reasonable that the risks of early fracture fixation may be most pronounced in patients who are the most severely injured, particularly if they require other life-saving measures during their course. Damage control is a term coined by Rotondo et al. in 1993 and has been most often used for devastating abdominal injury. 19 By using this philosophy, only major injuries resulting in significant blood loss are addressed at the time of initial laparotomy. Intestinal injuries are stapled, and packing is often used as an adjunct to hemostasis. The patient is then transferred as quickly as possible to the ICU for rewarming, monitoring, and ongoing resuscitation. Generally 24 to 48 hours later, when the patient is adequately resuscitated, warm, and has a normal coagulation profile, he or she is taken back to the OR and unpacked. Gastrointestinal reconstruction can be performed at that time.

This pattern of care can be used for injuries outside of the abdomen as well. These techniques can be adapted for use in the thorax, central vascular system, and even in the extremities for peripheral vascular injuries. 20–22 Applying this technique to the badly injured patient with bony injury would involve stabilizing fractures soon after injury, minimizing the operative time, and preventing heat and blood loss. For patients with closed head injuries, this protocol would potentially prevent secondary brain injury. Patients should then be taken to the ICU. Once resuscitated, they are returned to the OR for more elective definitive fracture fixation. Presumably, patients would better tolerate definitive fracture fixation when they were more stable, particularly if it required prolonged operative time, significant soft-tissue dissection, and/or blood loss.

EF is a technique that has been well established in the care of patients with multiple injuries. Early use of EF can reduce pelvic volume and limit blood loss in patients with significant pelvic fractures. 10,11 In addition, the use of EF on the upper extremities and tibia is commonplace. There are little data about the use of EF for femur fractures. The treatment of femoral shaft fractures with EF has been reported in pediatric patients with good results. 23 In adults, EF of femoral shaft fractures has been shown to have a high rate of complication. Nonunion and deep pin tract infection approach 20%. 24–26 In addition, 45% of patients treated with EF develop knee stiffness. However, in these series’ EF was used only for complex and/or open femoral shaft fractures. External fixation as an initial treatment with later conversion to IMN has been studied extensively in tibial fractures. Two early reports condemned this treatment with infection rates as high as 44%. 27,28 However, both these series used an extended period of EF with high rates of pin tract infection before IMN. By limiting the duration of EF and associated pin tract infection, other authors achieved a high union rate and infection rate of under 6%. 29

In our institution, we use EF as a method of initial stabilization for fractured femurs in patients with multiple injuries. We include all types of femoral shaft fractures and converted these to IMN relatively early. The mean time to definitive fracture fixation after EF was approximately 4 days. Although the current study addresses the use of EF for stabilization of only femoral fractures, we routinely use temporary EF for stabilization of fractures of the humerus, forearm, and tibia in victims of multi-system trauma who demonstrate relative or absolute contraindications to early open reduction and internal fixation or IMN. The choice and timing of definitive fixation procedures in these instances varies, depending on the specific bone and soft-tissue injuries. Further study is necessary to define better the results and feasibility of this staged fixation technique for long bones other than the femur, but many similarities between these patients and those included in the present study are evident.

The patients treated with external fixation were more seriously injured and less physiologically stable than those treated with standard intramedullary nailing. Patients treated with EF had a significantly higher ISS and were much more likely to present in shock. In addition, they were more likely to have serious closed head injuries. They had a significantly lower admitting GCS score and were more likely to have an AIS head score ≥ 3. Not surprisingly, resuscitation needs were much higher for patients treated with EF.

Our practice has been to aggressively fix femur fractures as soon as possible after admission. We routinely perform IMN of closed femoral shaft fractures as soon as initial work-up is completed. External fixation was selected only when the care of other injuries precluded definitive fixation or if the patient was thought not to be physiologically suitable for early IMN. The reasons for this were numerous and they included anatomic as well as physiologic parameters. Patients with contraindications to primary IMN are paradoxically the same patients least likely to tolerate the complications of skeletal traction. These relatively sophisticated decisions were made in concert by the trauma attending surgeon, orthopedic attending surgeon, and the other consulting services.

It would seem that patients with multiple injuries and with a closed head injury are ideally suited for this type of therapy. This was the case in 40% of our patients. The short operative time of 30 minutes allows patients to be followed up with serial neurologic examinations. The minimal blood loss with EF and short anesthetic time limits the incidence of hypoxia and hypotension and minimizes secondary brain injury. Jaicks et al. reported a 62% incidence of intraoperative hypotension and an 11% incidence of intraoperative hypoxia during early definitive fracture fixation in patients with head injuries. 9 They postulated that these intraoperative problems explained the lower discharge GCS score in patients who underwent early fracture fixation. Townsend et al. have echoed this concern. 8 In a recent report, they demonstrated an inverse relationship between timing of fracture fixation and intraoperative hypotension. Patients with femur fractures and moderate or severe head injuries had a 68% incidence of hypotension when definitive fracture fixation was performed within 2 hours of admission. This rate fell to 8% if fracture fixation was delayed more than 24 hours. We believe that the use of external fixation as a temporizing measure should allow for the advantages of rigid fracture fixation without the aforementioned intraoperative complications. None of our patients developed intraoperative hypotension or hypoxia. We were able to accomplish external fixation quickly. Mean OR time was 35 minutes for external fixation, and blood loss was under l00 mL. This finding was in contradistinction to a time for IMN of over 2 hours and a blood loss over four times larger than for EF.

Our approach to “damage control orthopedics” mirrors that of torso injury. After EF is completed, patients are taken to the ICU for ongoing resuscitation. When they are deemed physiologically stable, they return to the OR for conversion to IMN. At the very least, intracranial hypertension must be controlled. In general, we repeat the head computed tomographic scan and use those results in addition to serial neurologic examinations to determine neurologic stability. Hypothermia must be corrected, and the patient must have a normal coagulation profile. We use volume, red blood cells, and inotropes to support cardiovascular function to clear lactate to normal. We attempt to optimize ventilatory mechanics, weaning FIO2 to the minimum level necessary. When all of these factors have been controlled, we deem the patient a candidate for IMN. In general, this is performed in the next available elective time slot.

The complication rate in patients treated with EF compares very favorably to those treated with primary IMN. Complications were relatively rare and were for the most part of little physiologic consequences. We only followed patients to the time of hospital discharge. However, Nowotarski et al. reported on 54 patients treated earlier at our institution where initial femoral EF was converted to IMN. 30 They had 11-month follow-up. Major fracture-related complications occurred only in two patients.

Nine percent of the patients treated with EF died. Although this rate is statistically significantly higher than those treated with IMN, the small cell size precludes conclusions. Three of the deaths were from closed head injury. One patient was treated with EF alone. The other two had an uneventful conversion to IMN. It seems difficult to implicate the choice to use EF as causative. The last death was a 17-year-old boy who sustained a high-energy crush injury. He underwent external fixation of his pelvis and both femurs after bilateral hypogastric embolization. He could not be resuscitated and died within 24 hours of admission.

There is substantial literature demonstrating the salutory effects of early operative fixation in patients with long-bone fractures. Although this has been questioned, it seems unwise to abandon this practice. There are some clear advantages to early stabilization of fractures. It is important to remember that fracture fixation is an operation of magnitude and has physiologic consequences. Blood loss can be substantial, and this potential, combined with prolonged anesthesia and the soft-tissue injury that can accompany fracture fixation, may be the difference between compensated and uncompensated shock in patients with multiple injuries. Certainly, patients who have their immune system primed by injury may have precipitous release of mediators by a second insult such as early fracture fixation, especially if it is combined with significant blood loss. 31 Over the past few years, our practice has gravitated to an increased use of temporary EF followed by IMN. In the earlier study at the STC, EF was used in 4% of the total femoral shaft fractures seen. In the current study, EF was used in 13% of patients.

EF has the potential to deliver all of the beneficial effects of early fracture fixation with virtually none of the potential complications. Immediate EF followed by early closed medullary nailing is a safe method for treating femoral shaft fractures in badly injured patients. It is a valuable addition to the armamentarium of those caring for severe injury. This technique follows all of the principles of damage control, a treatment philosophy that has been adopted by virtually every trauma center in this country.

Back to Top | Article Outline

REFERENCES

1. Scalea TM, Scott JD, Brumback RJ, et al. Early fracture fixation may be “just fine” after head injury: no difference in central nervous system outcomes. J Trauma. 1999; 46:839–846.
2. Seibel R, LaDuca J, Hassett JM, et al. Blunt multiple trauma (ISS 36), femur traction, and the pulmonary failure-septic state. Ann Surg. 1985; 202:283–295.
3. Bone LB, Johnson KD, Weigelt J, Scheinberg R. Early versus delayed stabilization of femoral fractures: a randomized prospective study. J Bone Joint Surg Am. 1989; 71:336–340.
4. Charash WE, Fabian TC, Croce MA. Delayed surgical fixation of femur fractures is a risk factor for pulmonary failure independent of thoracic trauma. J Trauma. 1994; 37:667–672.
5. Behrman SW, Fabian TC, Kudsk KA, et al. Improved outcome with femur fractures: early versus delayed fixation. J Trauma. 1990; 30:792–798.
6. Johnson KD, Cadambi A, Seibert B. Incidence of adult respiratory distress syndrome in patients with multiple musculoskeletal injuries: effect of early operative stabilization of fractures. J Trauma. 1985; 25:375–384.
7. Pape HC, Auf’m’Kolk M, Paffrath T, Regel G, Sturm JA, Tscherne H. Primary intramedullary femur fixation in multiple trauma patients with associated lung contusion: a cause of posttraumatic ARDS? J Trauma. 1993; 34:540–548.
8. Townsend RN, Lheureau T, Protetch J, et al. Timing fracture repair in patients with severe brain injuries. J Trauma. 1998; 44:977–981.
9. Jaicks RR, Cohn SM, Moller BA. Early fracture fixation may be deleterious after head injury. J Trauma. 1997; 42:1–6.
10. Burgess AR. The management of hemorrhage associated with pelvic fractures. Int J Orthop Trauma. 1992; 2:101.
11. Henry SM, Tornetta P, Scalea TM. Damage control for devastating pelvic and extremity injuries. Surg Clin North Am. 1997; 77:879–895.
12. Seigel S, Castellan NJ Jr, eds. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill; 1998.
13. Riska E, Von Bonsdorf H, Hakkinen S. Prevention of fat embolism by early internal fixation of fractures in patient with multiple injuries. Injury. 1976; 8:110–116.
14. Lozman J, Deno DC, Feustel PJ, et al. Pulmonary and cardiovascular consequences of immediate fixation or conservative management of long-bone fractures. Arch Surg. 1986; 121:992–999.
15. Goldstein A, Phillips T, Sclafani SJA, et al. Early open reduction and internal fixation of the disrupted pelvic ring. J Trauma. 1986; 26:325–333.
16. Reynolds MA, Richardson JD, Spain DA, et al. Is the timing of fracture fixation important for the patient with multiple trauma? Ann Surg. 1995; 222:470–481.
17. Rogers FB, Shackford SR, Vane DW, et al. Prompt fixation of isolated femur fractures in a rural trauma center: a study examining the timing of fixation and resource allocation. J Trauma. 1994; 36:774–777.
18. Poole GV, Miller JD, Agnew SG, et al. Lower extremity fracture fixation in head-injured patients. J Trauma. 1992; 32:654–659.
19. Rotondo M, Schwab CW, McGonigal M, et al. Damage control: an approach for improved survival in exsanguinating penetrating abdominal injury. J Trauma. 1993; 35:375–382.
20. Wall MJ, Soltero E. Damage control for thoracic injury. Surg Clin North Am. 1997; 77:863–878.
21. Reilly PM, Rotondo MR, Carpenter JP, et al. Temporary vascular continuity during damage control: intraluminal shunting for proximal superior mesenteric artery injury. J Trauma. 1995; 37:757–759.
22. Scalea TM, Mann R, Austin R, et al. Staged operation for exsanguinating lower extremity trauma: an extension of a technique. J Trauma. 1994; 36:291–293.
23. Schneider J, Crosby LA. Treatment of pediatric femur fracture with external fixation. Nebr Med J. 1996; 8:432–435.
24. Mohr VD, Eickhoff U, Haaker R, et al. External fixation of open femoral shaft fractures. J Trauma. 1995; 39:648–652.
25. Murphy CP, D’Abrosia RD, Dabezies EJ, et al. Complete femur fractures: treatment with Wagner external fixation device or the Grosse-Kempf interlocking nail. J Trauma. 1988; 32:1553–1561.
26. Alonso J, Geissler W, Hughes JL. External fixation of femoral fractures. Clin Orthop. 1989; 241:83–88.
27. Maurer DJ, Merkow RL, Gustilo RB. Infection after intramedullary nailing of severe open tibial fractures initially treated with external fixation. J Bone Joint Surg Am. 1989; 71:835–838.
28. McGraw JM, Lim EV. Treatment of open tibial shaft fractures: external fixation and secondary intramedullary nailing. J Bone Joint Surg Am. 1988; 70:900–911.
29. Antich-Adrover P, Marti-Garin D, Murias-Alvarez J, Puente-Alonso C. External fixation and secondary intramedullary nailing: a randomized prospective trial. J Bone Joint Surg Br. 1997; 79:433–437.
30. Nowotarski PJ, Turen CH, Brumback RJ, et al. Conversion of external fixation to intramedullary nailing for fractures of the femoral shaft in polytrauma patients. J Bone Joint Surg Am. In press.
31. Moore FA, Moore EE. Evolving concepts in the pathogenesis of postinjury multiple organ failure. Surg Clin North Am. 1995; 75:257–277.
Back to Top | Article Outline

DISCUSSION

Dr. John A. Morris Jr. (Nashville, Tennessee): Dr. Scalea and his colleagues have described the concept of orthopedic damage control. They have looked at 321 femur fractures admitted to Maryland Shock Trauma over a 3-year period, and 280 of those received early fixation using intermedullary nails.

Forty-three patients received initial orthopedic damage control, which consisted of placement of external fixation followed about 4 days later by standard placement of an IM nail. It is clear that the 43 patients who received orthopedic damage control were sicker based on ISS, GCS, lactate, units of blood transfused, and associated injuries.

The long-term outcome between the two groups favored the patients who had relatively isolated femur fractures as opposed to the patients who had orthopedic damage control, as would be expected. Complications attributed directly to external fixation were minimal.

To put this study in perspective, Dr. Scalea and his colleagues have provided us with a Phase 2 study which says orthopedic damage control can be done safely, safely in a population of very sick patients. What they have not done, nor have they intended to do, is to demonstrate the efficacy of external fixation as a bridge to intermedullary nailing of femur fractures.

The efficacy study would require a multi-institutional trial with a study group all requiring orthopedic damage control. The patients would then be randomized to our current therapy or to orthopedic damage control. It is only in this fashion that we can answer the question, Is the additional time and expense spent in the process of external fixation on a temporary basis a benefit to both the patient and the patient’s payer?

I have two questions for the authors. First, how would you define the entry criteria for such a study on orthopedic damage control? Would it be based purely on the presence of head injury or would there be a series of hemodynamic ventilatory and metabolic parameters such as exist for patients undergoing thoracoabdominal damage control?

I encourage the authors to continue this exciting work. I especially encourage them to form a multi-institutional trial to validate the efficacy of this procedure.

And that brings me to my second question. How do I sign up for that study?

I thank the Association and the members for the privilege of the floor.

Dr. Hans-Christoph Pape (Hannover, Germany): I first would like to congratulate Dr. Scalea for his data that are very interesting and challenging.

The bottom line seems to be that you precluded patients in a critical condition from primary nailing, and you found a comparably good outcome. This matches precisely the experience that we have had in Hannover. We have previously identified similarly subgroups of patients that we named borderline in our 1993 publication. Probably, the consideration of an additional subgroup of patients with severe head trauma may make sense.

Also, your paper is in line with three presentations which were given last year at the meeting of the Orthopedic Trauma Association in fall and several papers are also due this fall. They advocate the conversion of early fixation into IM nailing, and all of them say it is safe.

And now, in contrast, there is a paper from 1997 which is also from the Shock Trauma Center, from some authors from the Shock Trauma Center, Michael Bussey and Andrew Burgess. They also looked at severely injured patients they had in one of their groups with thoracic trauma and ISS of 30. Yet, the ARDS incidence was only 2%. They concluded that primary fracture fixation is not harmful. One of their conclusions was this.

Now, when I first read your abstract, I figured that maybe there is a contradiction between the findings published by the Orthopedic Department and your department, but then I looked a little more closely and it turned out that in the Bussey paper, patients were included from 1983 until 1994, whereas you included patients from July 1995 to June 1998.

So I have two questions. First of all, did you change your management protocol within the recent years? If so, based on which findings?

Second, in your paper, you indicated that there was a difference in blood loss, 90 mL for external fixation group and 400 for the IM nailing group. How did you assess that the 400 mL in the IM nailing was only due to IM nailing but not due to other orthopedic surgery that might have been done simultaneously?

Dr. Stephen M. Cohn (Miami, Florida): It seems that there is a little bit of irony here. Last year you published a paper saying that it was okay to treat all head-injured patients with internal fixation early. But at the same time, 15% of your population was being excluded from internal fixation by the external fixation group.

I had the opportunity to talk to Andy Burgess about 3 years ago, and he told me you were doing this, and I do think it is an innovative way to stabilize the fracture and minimize bleeding, etcetera. But, on the other hand, it seems to support the view that not every patient, particularly those with head injury, can safely undergo—at least, it seems like you have changed your philosophy and now you feel that early fracture fixation in head injury is potentially harmful. Maybe you could comment on that.

Dr. Lawrence H. Pitts (San Francisco, California): The idea of a quick procedure with little blood loss sounds very attractive to the neurosurgeon treating a critically ill patient.

Alex Valadka—I am sorry he is not here—I think just out of interest in the topic reviewed—and I think he said he presented at EAST; I may have remembered that wrong—anyway, reviewed a hundred just IM nailings and found on the anesthetic record hypertension to below 90 for two successive tick marks in a third of the patients and blood loss routinely about 500 cc, which is not far off from the 400 you reported here. So anything other than that sounds attractive to a neurosurgeon.

A question that I have is if you do external fixation in one of these critically ill patients and they remain critically ill for a period of time, intracranial hypertension and so forth, how long can you delay before you feel it necessary to go on to a more definitive management? How much time can you buy with this technique? Thanks.

Dr. Clayton H. Shatney (San Jose, California): Dr. Scalea, one question. If your orthopods are like mine, they would like probably antibiotics in these people prior to going ahead and doing the definitive procedure. Is that the case? Do you treat these folks with antibiotics prior to the ORIF, and if so, what is the current antibiotic de jour?

Dr. Thomas M. Scalea (closing): Thank you. I am gratified to see this degree of interest at this late hour. Instead of answering individual questions, let me make some summary comments. This is a process that is continuing to evolve. Ten years ago we thought we had answered the question. Everybody had early IM nailing for long-bone fractures. Now we are not so sure. It is not yet clear which patients ought to have early IM nailing and which patients should not. I absolutely agree with Dr. Morris this is going to require a significant amount of work and it almost certainly will need to be multi-institutional in nature in order to answer the question.

Rather than say that we are changing, I prefer to think that we are increasing the sophistication of our decision-making. Decisions about the timing of fracture fixation should be based on physiologic principles. This should involve an assessment of the cardiovascular, pulmonary, and neurologic stability in concert with associated injuries. The nature of the fracture, the magnitude of operation needed to definitively fix this fracture are also important components of the decision-making. It is only a global assessment like this that allows for the development of a rational plan.

In general, this is the way we decide. We gauge the initial depth of shock by a combination of vital signs, but more importantly, the initial base deficit. Ongoing cardiovascular stability is best gauged in our opinion by supporting oxygen delivery to clear lactate to normal. We examine pulmonary mechanics to identify rapidly worsening respiratory failure from conditions such as contusion and/or aspiration and if so, what we can do to optimize pulmonary performance. We examine the nature of their head injury. In general, we repeat CT scans about 6 hours after admission. Neurologic performance is gauged by clinical exam plus the evolution of the head CT. Lastly, we take factors such as age and chronic medical conditions in an attempt to gauge physiologic reserve.

As many of you know, we are quite liberal with invasive monitoring such as intracranial pressuring monitoring and invasive hemodynamic monitoring. We place these monitors in the resuscitation unit to guide our resuscitation. Some patients are best served by us being able to watch them closely and address minute-to-minute neurologic and cardiovascular performance. If so, we are going to opt for the 30-minute 90-cc blood loss procedure. On the other hand, if the patient is stable and safe for a prolonged period of anesthesia, we would just as soon go ahead and perform IM nailing on the night of admission.

The Shock Trauma Center reported its early experience with this technique in the early 1990s. In that series, 4% of patients with femur fractures had external fixation performed as a bridge to more definitive fixation. In our current series, it is 15% of these patients. Thus, it seems quite clear that more and more of the time we believe that the 30-minute, 90-cc blood loss procedure without physiologic insult is a better idea. This is particularly in contradistinction to our practice 10 years ago.

Though not in this series, we have treated some patients definitively with external fixation alone. Some of them simply never achieved adequate physiologic parameters to undergo more definitive fixation. Their fractures started to heal, and we deemed that this would be acceptable long-term therapy. It is clearly not our first choice. In general, we can stabilize almost everybody within 6 or 7 days. They then go back to the operating room. We do not generally keep people on antibiotics for any period of time and the presence of an external fixator by itself is not an indication for antibiotics in our institution. When they return to the operating room for their IM nailing, they get a pre-op dose of antibiotics to cover whatever bugs with which they have become colonized in the ICU. They then get their surgery. In general, we would treat with one additional day of antibiotics post-op and then they are stopped.

I would close by saying that this is the type of project that really defines the role of the general surgeon in caring for patients with multiple injuries. It seems odd even to me that I am up here talking about external fixation and IM nails. However, it is only by practicing in a truly collaborative, multidisciplinary way that we would be able to get at some of these fundamental issues. We need to define the question and then put together a multi-institutional trial to at least begin to scratch the surface.

I would like to thank all of my co-authors, the Shock Trauma Center, and the Association for the privilege of the floor. Thank you.

© 2000 Lippincott Williams & Wilkins, Inc.