In the treatment of closed diaphyseal fractures of the tibia surgical intervention is not always necessary; therefore, the decision to do surgery should be made with a firm belief that this treatment method is superior to nonsurgical treatment. Intramedullary nailing of proximal third diaphyseal fractures of the tibia, currently a popular treatment modality, frequently is associated with technical difficulties, an increased incidence of reoperation, and residual malalignment. 4,15
Closed treatment of these fractures is not a panacea; however, multiple reports of successful treatment of fractures of the tibial diaphysis have been published using functional braces and other nonsurgical methods. 3,16,17,21,22 After publishing experiences with tibial fractures at all levels, 21 a separate review of the fractures located in its proximal third was done.
MATERIALS AND METHODS
Patients with proximal third diaphyseal fractures of the tibia selected to have treatment with a functional brace met the following criteria: the initial shortening was less than 15 mm in axially unstable oblique, spiral, comminuted, and segmental fractures, and there was an angular deformity less than 5° after manual alignment of the fragments. In transverse fractures, initial shortening greater than 15 mm was accepted if a closed reduction restored axial stability. Patients included in this study first were evaluated on their referral to a special fracture clinic, where patients considered by the residents to have met the aforementioned criteria were seen. This method of selection excluded an unknown number of patients with closed fractures who were treated either surgically or nonsurgically for reasons such as associated polytrauma, patients with a surgically treated compartment syndrome that developed during the initial period of observation, patients with an unacceptable initial shortening, uncorrectable angular deformity, or an intact fibula with presence of an angulated tibial fracture.
A group of 1650 closed diaphyseal fractures of the tibia in skeletally mature patients, were treated with a functional brace from June 1983 to June 1992. 21 One thousand two (61%) patients were followed up to completion of healing. One hundred eight (10%) of the fractures were proximal third fractures of the tibia. The age of the patients ranged from 16 to 82 years with a mean of 36.3 years. Associated injuries occurred in 18 patients. Nineteen patients (17.6%) were females and 89 patients (82.4%) were males.
The left side was involved in 45 patients (42%) and the right side was involved in 63 patients (58%). The energy of trauma was defined as high in 71 patients (65.7%) and low in 37 patients (34.3%). The mechanism of injury was defined as low-energy trauma when a direct blow or fall caused the fracture. High-energy trauma was defined when the injury resulted from a motor vehicle accident, motorcycle accident, or pedestrian-vehicle accident.
Fractures were classified according to the Orthopaedic Trauma Association (OTA) classification. 1 Forty-five fractures (42%) were comminuted (OTA Group 42-B); 35 fractures (31%) were oblique (OTA Group 42-A2); 20 fractures (19%) were transverse (OTA Group 42-A3); four fractures (4%) were segmental (OTA Group 42-C) and four fractures (4%) were spiral (OTA Group 42-A1). An associated fibular fracture was present in 73 patients (68%); 66 (91%) fibular fractures were located in the proximal third, four (5%) were located in the middle third, and three (4%) were located in the distal third.
Patients initially were treated in the emergency room where the decision was made to align the fracture, with the patient under sedation, or reduce the fracture, with the patient under anesthesia. Information is not available regarding the number of patients in the two categories. In both instances, the limb was stabilized in an above-the-knee cast or splint. Patients were discharged when their neurovascular status was stabilized, which required overnight hospitalization in some instances. Information regarding the number of patients who required overnight hospital admission is not available. After discharge, progressive weightbearing as tolerated was encouraged.
Patients were seen in the outpatient clinic 1 week after discharge. At followup, the limb was fit with a prefabricated functional brace if swelling and acute pain had subsided. If swelling or pain or both still were significant at that time, the application of the brace was postponed for an additional week. The brace consisted of two prefabricated components of molded polyethylene (PE), spanning the region from below the knee to the ankle, with femoral condylar extensions (Fig 1). Both components were connected by a series of velcro straps. An articulated plastic foot insert connected the insert with the brace. 18 The time of application of the brace ranged from 4 days to 19.1 weeks, with an average of 31.3 days and a mean of 11 days. Eighty-four patients (80%) had the brace applied during the first 6 weeks. Radiographs were taken after application of the brace to confirm maintenance of acceptable alignment. Instructions given to the patients included progressive weightbearing as tolerated with the aid of crutches, active range of motion (ROM) of the knee and ankle, and frequent tightening of the straps on the brace. General principles of fracture care as elevation of the extremity and avoidance of periods of prolonged standing also were recommended while the limb still was swollen.
Patients generally were followed up initially at 1 week after application of the brace, at which time new radiographs were taken. Subsequently, they were followed up in most instances, clinically and radiologically at 1-month intervals. If radiographs showed a mild loss of alignment or reduction, tightening of the straps of the brace was done. If acceptable alignment could not be obtained in this manner, the fracture was remanipulated and if this failed, treatment was discontinued and a different treatment modality was adopted. Patients in this latter group were defined as having failure of the functional brace.
Radiographs obtained during the initial visit were measured in the fracture clinic by orthopaedic residents who were supervised by the senior author (AS). Shortening was measured on the initial radiographs in millimeters between the edges of clearly defined major fragments. Measurements were made at every followup and recorded. Displacement was defined as a percentage of the amount of translation of the fragments with respect to the width of the bone at the level of the fracture, and was measured on anteroposterior (AP) and lateral radiographs. Angulation was measured in degrees following the longitudinal axis of the main fragments.
The initial displacement of the fragments ranged from 0% to 100% with an average of 20.08%. Seventeen fractures (16%) had an initial displacement of 50% or greater. Initial shortening ranged from 0 to 15 mm with an average of 3.86 mm. Nineteen fractures (18%) had initial shortening of 10 mm or greater. Thirty-two fractures (30%) had an initial valgus angulation that ranged from 0° to 25 ° with an average of 5.58°; 24 fractures (22%) had an initial varus angulation that ranged from 0° to 10° with an average of 5.61°. There were 21 fractures (19%) with a posterior angulation that ranged from 0° to 12° with an average of 4.4°; 30 fractures (28 %) had an initial anterior angulation that ranged from 0° to 15° with an average of 5.63°.
Fracture union was determined when the patient was pain-free while bearing full weight and peripheral callus bridging was confirmed on radiographs; nonetheless patients were encouraged to continue wearing the brace until callus maturation was evidenced. Time of union was defined as the number of weeks from the day of the fracture to the day of discontinuation of the brace.
Patient and Fracture Variables
Three patients (3%) required change from the brace to a cast temporarily to ensure maintenance of adequate reduction, and subsequently wore the brace to completion of the treatment. Two additional patients (2%) had skin breakdown as a complication caused by pressure from the brace; they were treated by padding the brace at the level of the injured skin and with dressing changes until the skin defects healed.
Compartment syndromes did not develop in any patient after application of the brace. Patient and fracture variables included in the study were age, side of injury, type of fracture (transverse, oblique, comminuted, spiral, or segmental), initial displacement, initial shortening, initial angulation (frontal and sagittal planes), and associated injuries.
Energy of trauma was correlated with the type of fracture. In high-energy trauma, the most frequent type of fracture was comminuted in 34 patients (48%), whereas in low-energy trauma the most frequent type was an oblique fracture in 17 patients (48 %).
Outcome variables were defined as healing time, final shortening, and final angulation and displacement in the frontal and sagittal planes. The differences among mean values of outcome variables were compared using analysis of variance (ANOVA) followed by the least significant difference method to determine which groups had differences in outcome. Associations between variables were assessed with chi square and Student’s t test. Statistical significance was determined through the probability value.
A delayed union was defined as a fracture that had not shown evidence of osseous bridging between the two major fragments by the thirtieth week after injury; and a nonunion if bridging was absent 9 months after the initial insult.
Healing of the fractures took place from 6.6 weeks to 40.5 weeks, with an average of 17.1 weeks, and a mode of 11.6 weeks. Three patients (2.8%) required discontinuation of the conservative treatment and had some form of surgery. One of these patients was treated with a fibular ostectomy because of severe varus at the fracture site, which occurred on initiation of weightbearing ambulation. The patient then was treated with a functional brace. Another patient was treated with interfragmentary screw fixation because of loss of reduction that did not respond to manipulation; the third patient was treated with bone grafting followed by application of a functional brace. All three fractures in these patients healed after these interventions.
Final shortening of the fractures ranged from 0 to 20 mm with an average of 3.6 mm. Ninety-seven fractures (92%) healed with shortening of 10 mm or less. Twenty-eight fractures (26.7%) healed with no displacement in any plane. The average final displacement was 19.4% (Fig 2).
No angulation in the frontal plane was seen in 46 fractures (43.8%). Twenty-two patients (21%) had a valgus angulation with a range from 0° to 11° and an average of 3.9°. Thirty-seven fractures (35.2%) healed with varus angulation averaging 5.61°, ranging from 0° to 15°. Thirteen fractures (12.3%) had angulation greater than 6° in the coronal plane; most of these were in varus (10 fractures).
In the sagittal plane, no final angulation was seen in 58 (55.2%) fractures. Final posterior angulation (recurvatum) was seen in 17 (16.2%) fractures with a range from 0° to 10° and an average of 4.6°. Final anterior angulation (antecurvatum) was encountered in 30 (28.6%) fractures with a range from 0° to 11° and an average of 4.5°.
Results of Multivariate Analysis
Fracture type was associated with time to union. Comminuted and transverse fractures had an increased time to healing of 18.1 and 17.1 weeks, respectively. The shortest average time to union was seen in oblique fractures with 15.6 weeks.
Energy of trauma also was associated to the time to union. Low-energy trauma resulted in union of the fracture in an average of 15.2 weeks whereas high-energy trauma led to union at an average of 18.1 weeks. This difference was statistically significant (p = 0.03).
The presence of a fibular fracture was associated with delayed healing in an average of 18.3 weeks whereas fractures in patients without a fibular fracture healed in an average of 14.9 weeks (p = 0.02).
The age of the patients did not affect the time to union. When 30 years was used as a defining point, the average time of healing in patients younger than 30 years was 16.6 weeks compared with 17.4 weeks for patients older than 30 years (p = 0.755).
The time from the initial injury to the day the brace was applied had a direct association to the number of weeks to union. This difference was statistically significant with a p value less than 0.0005.
Shortening was considered unacceptable if it exceeded 15 mm, based on the fact that such shortening rarely produces a limp. Final shortening was associated with the energy of trauma, fracture type, initial displacement, and association to a fibular fracture. The average final shortening in patients with high-energy trauma was 4.8 mm whereas that of patients with low-energy trauma was 1 mm (p < 0.001). Patients with an associated fibular fracture had a mean final shortening of 4.9 mm whereas patients without an associated fibular fracture it had an average final shortening of 0.9 mm (p < 0.001). The three patients with a distal fibular fracture did not have any shortening on the final radiologic evaluation; there was no statistically significant difference in shortening between patients who had a concomitant fibular fracture in the middle or upper thirds of the bone. Patients who had an initial displacement less than 50% had an average final shortening of 2.5 mm, whereas patients with greater than 50% initial displacement had an average shortening of 9.1 mm (p < 0.001). Finally, shortening was greater with segmental fractures, followed by spiral and comminuted fractures.
An angular deformity was considered acceptable if it was less than 6°. There were no associations between final anterior and posterior angulation and final varus and valgus angulation, displacement, or shortening.
Final angulation varied according to the presence of fibular fracture. On average, final angulation with a concomitant fibular fracture was 1° varus; without a fibular fracture it was 1.4°. In the 26 patients who had a neutral initial alignment, nine progressed to having a varus angulation of an average of 6.1°, whereas the others maintained their neutral alignment.
There was no association between presence of a fibular fracture and a greater incidence of anterior or posterior angulation (Fig 3).
Associations were found between the amount of initial displacement and final displacement and between energy of trauma and final displacement. Fractures in patients with an initial displacement less than 50% healed with a final displacement of 14.4% on average, whereas fractures with an initial displacement greater than 50% healed with an average displacement of 45.2%; this difference was statistically significant (p < 0.001). Patients with a high-energy trauma had a final displacement of 23.6% and patients with low-energy trauma had a final displacement of 10.9% (p < 0.001).
Numerous closed tibial fractures currently are being treated by surgical means, closed intramedullary nailing being the most popular surgical modality. Nailing, however, as it is true with all treatment approaches, may be accompanied by complications. Reports of infection after intramedullary nailing of closed diaphyseal fractures show an incidence ranging from 0% in 29 patients reported by Hooper et al 9 to 4.4% in 68 patients reported by Bone and Johnson. 2 Knee pain after intramedullary nailing has been a frequent complication. 5,7,10,24 Court-Brown et al 5,7 reported knee pain in 60% of 25 patients. Removal of the nail was necessary to treat the pain and often did not resolve the pain. In another study, Keating et al 10 reported the need for removal of the nail because of knee pain after treating tibial fractures in as many as 80% of 61 patients, and after 16 months, the pain had not resolved in 22 (36%) of these patients.
As reported by Court-Brown et al, 7 malunion was present in as many as 16% of 25 fractures treated with unreamed nails in their comparative study between reamed and unreamed intramedullary nailing. Compartment syndromes also have been described after intramedullary nailing of the tibia 7,12,24 and in one study the incidence was as high as 12%. 24 These patients may have a compartment syndrome develop even if treated by nonsurgical means. However, it is logical to assume that in instances when before surgery the compartment pressures may have been elevated, but not to a dangerous degree, the required traction on the leg applied at the time of surgery, which decreases the volume of the deformed compartment, and the emptying of bone marrow and blood into the fracture site may have contributed to the development of the full-blown syndrome. 19 In addition, specific problems encountered with proximal diaphyseal fractures treated by intramedullary nailing include a high incidence of reoperation, undesirable residual angulation, and technical difficulties. 4,11,15
In the current study, complications were limited to superficial skin breakdown, which was treated readily by padding of the brace and observation, and a 2.8% rate of nonunion. The patients who had nonunion develop then were treated surgically and subsequently achieved union. No compartment syndrome developed after application of the brace.
Twelve percent of the fractures healed with an angulation greater than 6° in the frontal plane. In the report by Lang et al 11 of intramedullary nailing of proximal third diaphyseal fractures, 13 of 32 fractures (41%) healed with an angulation greater than 6° in the frontal plane.
Even though in the current study the absence of a fibular fracture did not produce a significantly higher varus angulation, the likelihood of this occurrence cannot be overlooked, especially for more proximal fractures. As stated previously, the angulation occurs when the patient starts to bear weight with an intact fibula, resulting in varus and recurvatum. 17 It now is thought that in patients with an intact fibula who are to be treated nonsurgically, the introduction of weightbearing ambulation should be delayed until intrinsic stability develops. They also should be monitored closely with followup radiographs.
The current study indicated shortening greater than 10 mm in eight patients (8.4%). In other studies with intramedullary nailing, shortening greater than 10 mm was present in five (5%) of 100 patients according to Bone and Johnson, 2 and in one (3.4%) of 29 patients in the study by Hooper et al. 9 Ninety-eight percent of the patients in the current study had less than 13 mm final shortening. We think that this amount of shortening is clinically irrelevant. Axially unstable fractures that are subjected to manual traction to regain length, experience a return to the initial shortening on introduction of weightbearing. 14,21
Downing et al 8 concluded that there is no difference in the overall cost to the community between patients treated with a cast and patients treated with intramedullary nailing. The study is based on a small series of patients and even though it includes calculations on the impact of lost workdays with the primary treatment, it does not include the costs that could be raised by complications such as infection, need for hardware removal, or nail exchange. In a previous study it was indicated that there is a significant difference in the cost of care of surgically and nonsurgically treated tibial fractures. 21 This study did not discuss economic factors, but it is suspected that the cost of intramedullary nailing is greater than functional bracing.
The current study confirms some of the facts that are true for other closed diaphyseal fractures treated with functional bracing. The type of fracture has a definite impact on healing time, with a longer period required to heal comminuted and transverse fractures. Energy of trauma has a direct association to healing time and this is readily explained by the amount of soft tissue trauma, which increases as the energy of trauma increases. The role of soft tissues in stabilizing tibial fractures is well known. 20 One of the most popular classifications for fractures in Europe is based on the concept of soft tissue injury around fractures. 23
Because fractures with a concomitant fibular fracture healed more slowly than those that did not have a concomitant fibular fracture, probably can be explained as a result of a lower magnitude of soft tissue damage in the latter. This behavior had been reported previously. 18 The main difference found in this selective group of fractures was that even though fractures with an intact fibula angulated with a greater frequency toward varus, the strength of this association was not significant. The average final varus angulation in these fractures was 6.1°, an amount which is cosmetically acceptable and unlikely to result in degenerative joint disease of the knee. 13 However, on the basis of a long-held interest in the behavior and treatment of tibial fractures, patients with proximal or distal tibial fractures with an intact fibula should be denied early application of the brace and weightbearing ambulation until intrinsic stability is documented. Oblique fractures located in the proximal third of the tibia that run from proximal to distal and from lateral to medial do not angulate or angulate minimally. 18
Age did not affect the healing time. As far as the tibia is concerned, age does not seem to affect the healing time in a population of adults. 21
The time of application of the brace also affected the time to healing. This emphasizes the importance of early motion and graduated weightbearing in the healing of fractures. 21
There has been a report of unsatisfactory results with conservative treatment of tibial fractures compared with intramedullary nailing. 9 In that study Hooper et al concluded that conservative treatment had to be discontinued because of ethical concerns knowing they had better results with intramedullary nailing. We think that their inadequate results came from inappropriate selection of patients and inadequate treatment in the conservative group. They included seven (21%) open fractures of the tibia, corrected shortening with traction and manipulation, and postponed weightbearing for 4 weeks and in some instances it was not started until after union of the fracture. The use of functional braces should be restricted to closed fractures with acceptable initial shortening. 21 The likelihood of reappearance of initial shortening after manipulation is high in axially unstable fractures. The initial shortening in these fractures ideally should be no greater than 15 mm. Gradually increased weightbearing with a functional brace should be encouraged from the beginning of the treatment, starting with weightbearing as tolerated in the above-the-knee cast, and then continued until union is obtained. The importance of physiologically induced motion at the fracture site has been shown in animal and human studies. 14,20 Efforts to maintain alignment of the fracture frequently are made throughout the treatment by tightening the Velcro straps. If these considerations are followed, the possibilities of a good result are enhanced. 16,18,19,21
The final ROM of the joints adjacent to the tibial fracture was not reported in this study. After long experience with functional bracing of tibial fractures it was concluded that the relatively short period of immobilization of adjacent joints in long-leg casts does not produce long-lasting limitation of motion of the knee and ankle. 21
The patients included in the current study were not consecutive patients. An unknown number of patients with closed fractures of the proximal third of the tibia might not have received nonsurgical treatment because of initial shortening greater than 15 mm or had what seemed to be severe angular deformities not likely to be correctable by manipulation. However, these instances are rare, because severe deformities mainly are associated with open fractures. Patients with polytrauma frequently were treated by surgical means. Functional bracing in this selective group of fractures results in high union rate and low morbidity.
The authors thank the residents in the Department of Orthopaedics at the University of Southern California/ Los Angeles County Medical Center, particularly Frances Sharpe, MD, James Shankwiler, MD, Larry Gersten, MD, and Phillip Sobol, MD for contributions. They also thank Patricia Normand, MS and Edward Ebramzadeh PhD, from the Vernon Luck Research Laboratories at Orthopaedic Hospital of Los Angeles, for statistical preparation of the data provided.
1. Anonymous: Fracture and dislocation compendium: Orthopaedic Trauma Association Committee for Coding and Classification. J Orthop Trauma 10(Supp 1): 1–153, 1996.
2. Bone LB, Johnson KD: Treatment of tibial fractures by reaming and intramedullary nailing. J Bone Joint Surg 68A: 877–887, 1986.
3. Brown PW, Urban JG: Early weight-bearing treatment of open fractures of the tibia: An end-result study of sixty-three cases. J Bone Joint Surg 51A: 59–75, 1969.
4. Buehler KC, Green J, Woll TS, Duwelius PJ: A technique for intramedullary nailing of proximal third tibia fractures. J Orthop Trauma 11: 218–223, 1997.
5. Court-Brown CM, Gustilo T, Shaw AD: Knee pain after intramedullary nailing: Its incidence, etiology, and outcome. J Orthop Trauma 11: 103–105, 1997.
Court-Brown CM, Keating JF, McQueen MM: Infection after intramedullary nailing of the tibia. J Bone Joint Surg 74B: 770–774, 1992.
7. Court-Brown CM, Will E, Christie J, McQueen M: Reamed or unreamed nailing for closed tibial fractures. J Bone Joint Surg 78B: 580–583, 1996.
8. Downing ND, Griffin DR, Davis TR: A Comparison of the relative costs of cast treatment and intramedullary nailing for tibial diaphyseal fractures in the UK. Injury 28: 373–375, 1997.
9. Hooper GJ, Keddell RG, Penny ID: Conservative management or closed nailing for tibial shaft fractures: A randomized prospective trial. J Bone Joint Surg 73B:83–85, 1991.
10. Keating JF, Orfaly R, O’Brien PJ: Knee pain after tibial nailing. J Orthop Trauma 11:10–13, 1997.
11. Lang GJ, Cohen BE, Bosse MJ, Kellam JF: Proximal third tibial shaft fractures: Should they be nailed? Clin Orthop 315: 64–74, 1995.
12. Mawhinney IN, Maginn P, McCoy GF: Tibial compartment syndromes after tibial nailing. J Orthop Trauma 8:212–214, 1994.
13. Merchant TC, Dietz FR: Long-term follow-up after fractures of the tibial and fibular shafts. J Bone Joint Surg 71A:599–606, 1989.
14. Park SH, O’Connor K, McKellop H, Sarmiento A: The influence of active shear or compressive motion on fracture healing. J Bone Joint Surg 80A: 868–878, 1998.
15. Ricci WM, O’Boyle M, Borreli J, Bellabarba C, Sanders R: Fractures of the proximal third of the tibial shaft treated with intramedullary nails and blocking screws. J Orthop Trauma 15: 264–270, 2001.
16. Sarmiento A, Gerstein LM, Sobol PA, Shankwiler JA, Vangsness CT: Tibial shaft fractures treated with functional braces. J Bone Joint Surg 71B:602–609, 1989.
17. Sarmiento A, Kinman PB, Latta LL: Fractures of the proximal tibia and tibial condyles: A clinical and laboratory comparative study. Clin Orthop 145:136–145, 1979.
18. Sarmiento A, Latta L: Functional Bracing in Management of Tibial Fractures: The Intact Fibula. In Moore TM (ed). American Academy of Orthopaedic Surgeons Symposium on Trauma to the Leg and its Sequelae. St Louis, MO, CV Mosby 278–298, 1981.
19. Sarmiento A, Latta L: Functional Fracture Bracing: A Manual. Baltimore, Lippincott, Williams & Wilkins 133–137, 2002.
20. Sarmiento A, Latta L, Zilioli A, Sinclair W: The role of soft tissues in stabilization of tibial fractures. Clin Orthop 105:116–129, 1974.
21. Sarmiento A, Sharpe FE, Ebramzadeh E, Normand P, Shankwiler J: Factors influencing outcome of closed tibial fractures treated with functional bracing. Clin Orthop 315: 8–24, 1995.
22. Suman RK: The management of tibial shaft fractures by early weight in a patella tendon bearing cast: A comparative study. J Trauma 17: 97–107, 1977.
23. Tscherne H, Oestern HJ: A new classification of soft tissue damage in open and closed fractures. Unfallheilkunde 85: 111–115, 1992.
24. Williams J, Gibbons M, Trundle H, Murray D, Worlock P: Complications of nailing in closed tibial fractures. J Orthop Trauma 9:476–481, 1995.