Trauma is a growing global public health concern, and it is a major cause of death and disability worldwide. Trauma disproportionately affects low and middle-income countries1-4 and accounts for more disability than either HIV (human immunodeficiency virus), tuberculosis, malaria, or cancer do. Trauma is also the leading cause of death globally among people between the ages of five and forty-five years2,3. Accidents involving motor vehicles are the main cause of nonfatal injuries, with musculoskeletal trauma accounting for the majority of these injuries3. Femoral fractures are the most common musculoskeletal injury, accounting for 9% of all nonfatal injuries3.
In most resource-poor countries, skeletal traction is still the mainstay of treatment for femoral shaft fractures and is often accompanied by the known complications of pin-track infection, malunion, nonunion, thromboembolic disease, and decubitus ulcers5,6.
With the mission of improving fracture care worldwide, the Surgical Implant Generation Network (SIGN) nailing system was designed to be used in resource-poor areas at low cost and without real-time imaging, a fracture table, or power reaming. The SIGN nail is now regularly used in more than 200 hospitals located in forty-eight developing countries as well as in relief efforts following disasters such as the 2010 earthquake in Haiti, the 2007 earthquake in Pakistan, and the 2004 tsunami in Banda Aceh, Indonesia1. Unfortunately, there is a paucity of literature examining the outcomes and effectiveness of the SIGN intramedullary nailing system in the treatment of femoral shaft fractures6-8. Thus, we conducted a retrospective analysis of prospectively collected data on femoral shaft fractures treated with the SIGN nail in Mulago National Hospital in Uganda.
Materials and Methods
Mulago National Hospital is a 1500-bed facility that represents the national teaching and referral hospital of Uganda, an East African country with a gross domestic product (GDP) per capita of $1200, which ranks 208th out of 229 nations worldwide9. As the only government hospital in Kampala and the only facility offering twenty-four-hour emergency care, Mulago National Hospital receives the majority of trauma patients injured in the capital and surrounding areas. The Department of Orthopaedic Surgery, which has two hospital wards and three operating rooms, admits nearly 2000 patients with fractures annually. The SIGN intramedullary nailing system was introduced in the hospital in 2005.
Skeletally mature patients who presented to Mulago National Hospital with a closed fracture of the femoral diaphysis (AO/OTA10 type 32-A, 32-B, or 32-C) and underwent treatment with the SIGN intramedullary nail between February 1, 2007, and March 1, 2008, were eligible for inclusion in the study. Patients who had a concomitant fracture that limited mobilization or who had sustained a serious head injury, a spinal cord injury, or polytrauma were excluded. Patients whose fracture represented a nonunion or a pathologic fracture were also excluded. Patients from a remote place of residence were excluded in order to maximize the likelihood of follow-up.
All patients were treated with the SIGN intramedullary nailing system by a single surgeon (P.S.), and all relevant data were collected prospectively and entered into a database. Data were collected on patient demographics, injury characteristics, associated injuries, the fracture classification, and details of the preoperative treatment. Data were also collected on details of the surgical procedure, including the timing of the surgical intervention, the surgical approach (antegrade or retrograde), the reduction technique (open or closed), the implant size, and the number of interlocking screws used. Data collected at each follow-up visit included the development of infection, range of knee and hip motion, leg length discrepancy, and radiographic appearance of the fracture. All patients were followed for a minimum of six months or until fracture-healing. The prospectively collected data were then retrospectively reviewed by the senior author (A.J.) with use of an institutional review board-approved protocol.
The SIGN intramedullary nail is designed to be placed without use of fluoroscopy, power reaming, or a fracture table, and may be inserted in either an antegrade or a retrograde direction. We have generally preferred to use the antegrade technique for fractures of the proximal or middle third of the femur and the retrograde technique for fractures of the distal third, as the centered starting point helps to ensure that the nail will be centered within the distal fragment. However, the choice of antegrade or retrograde nailing was also influenced by a wide variety of other factors including the patient's body habitus, the condition of the soft tissues, and concerns regarding positioning of the patient during anesthesia. All patients received perioperative antibiotic treatment with ceftriaxone.
Antegrade nailing was performed with the patient in the lateral position. Open reduction of the fracture (Fig. 1) was usually necessary because of a lack of intraoperative fluoroscopy as well as a lengthy delay between the time of injury and the surgery. The length of the incision depended primarily on the elapsed time since the injury. The ends of the proximal and distal fragments were freed, and the fracture was brought out to length with use of manual maneuvers. One or more assistants were typically required for this task. Hand reaming was performed through the fracture site until the full length of the isthmus had been reamed. Any bone retrieved from the flutes of the reamer was saved for placement in the fracture site at the end of the procedure.
Next, the proximal incision was made and the proximal aspect of the femur was entered with use of a curved awl positioned at the tip of the greater trochanter between the posterior and middle thirds of the femur. (The SIGN intramedullary nail is designed with 9° of proximal curvature to accommodate this trochanteric entry site.) Hand reamers were then used to ream the segment bone between the osseous entrance and the diaphysis. The reaming was also used to determine the proper nail diameter, which is 2 mm less than the diameter of the reamer that elicits “chatter” for a distance of approximately 6 cm.
The target arm was attached to the nail, and the longitudinal positions of the proximal and distal interlocking screws were noted. The target arm was then removed and the nail was inserted, with rotation allowed during placement of the nail. After the nail had been inserted, the target arm was reattached to guide the placement of the interlocking screws (Fig. 2). We have typically preferred to use three interlocking screws (one in the larger fragment and two in the smaller fragment), but we frequently modified this number if the fracture appeared to be relatively stable (requiring only two screws) or particularly unstable (requiring four screws). The distal interlocking screws were usually placed first. A window was drilled in the near cortex at the position indicated by the target arm, and a specially designed cannulated slot finder was placed through this cortical window into the slot of the nail. The hole in the far cortex was then drilled through the cannulated slot finder, the length of the screw was measured, and the interlocking screw was placed. The proximal interlocking screws were then placed from lateral to medial or from anterior to posterior, with the direction depending on how the nail rotated during insertion, as the two orientations provide equal stability.
Retrograde nailing was performed with the patient in the supine position. As in the antegrade technique, the fracture site was opened to allow for reduction as well as reaming through the fracture site. After the knee incision had been created, the femoral canal was opened distally with use of the awl, and hand reaming was performed. The nail was inserted in a retrograde direction, and the target arm was used to place the interlocking screws with use of a technique similar to that described above. A more detailed description of the SIGN nailing technique has been published previously11.
Patients were encouraged to begin mobilization on the day following surgery. Weight-bearing status was individualized on the basis of the fracture pattern, the quality of the fixation, and patient factors (such as weight), but was typically weight-bearing as tolerated. Because fluoroscopy was not routinely available in the operating room and portable radiography was not routinely available in the recovery room, the first postoperative radiographs were typically made in the radiology suite prior to discharge from the hospital. Patients were asked to return to the treating surgeon for follow-up at six weeks and at three, six, and twelve months postoperatively, and radiographs were made at each visit (Figs. 3-A, 3-B, and 3-C). The time to union was recorded as the number of weeks until the first visit at which fracture union was observed to have occurred. The fracture was considered to have united once bridging callus was visible on both the anteroposterior and the lateral radiograph and the patient did not report pain at the fracture site.
Source of Funding
There was no external funding for this investigation.
The initial study population consisted of seventy consecutive patients with a closed femoral shaft fracture treated with use of the SIGN intramedullary nail by a single surgeon (P.S.) between February 1, 2007, and March 1, 2008. Twenty (29%) of these patients were excluded because of inadequate follow-up; follow-up is often difficult for patients in developing countries because of a lack of resources and transportation. The twenty patients who were excluded because of inadequate follow-up were similar to the included patients with regard to demographic data and injury characteristics.
The characteristics of the remaining fifty patients who formed the final study cohort are detailed in Table I. The patients were generally relatively young (mean age, thirty-one years; range, fifteen to seventy years), and 74% (thirty-seven) were men. The majority of the patients (forty-four, 88%) were injured in an accident involving a motor vehicle, including eighteen motorcyclists, seventeen pedestrians struck by a vehicle, seven motor vehicle occupants, and two bicyclists struck by a vehicle. Twenty patients had a concomitant injury, including eight minor head injuries, nine soft-tissue injuries (including lacerations and contusions), and three other fractures (one anteroposterior compression type 1 (APC-1) pelvic fracture, one iliac wing fracture, and one nondisplaced medial malleolar ankle fracture). Twenty-one of the fifty femoral fractures were classified as AO/OTA type 32-A, twenty-one were type 32-B, and eight were type 32-C. Initial management of the femoral fracture consisted of skeletal traction in four patients, skin traction in thirty-three patients, splinting in ten patients, and no immobilization in three patients. Two patients were diagnosed clinically with a pulmonary embolism preoperatively and were treated with low-molecular-weight heparin.
The mean time to surgery was 13.2 days (range, zero to thirty-three days). Twenty-seven fractures were treated with antegrade SIGN nailing, and twenty-three with retrograde nailing. Open reduction was used in forty-eight of the fifty patients. The mean nail diameter was 9.7 mm (range, 8 to 11 mm). The mean time from surgery to discharge was 6.9 days (range, two to eleven days).
All fifty fractures healed, but two required dynamization for the treatment of delayed union. Both of the dynamizations were performed three months after the nailing procedure, and both fractures healed within three months of the dynamization. Forty-six of the fifty fractures healed within 10° of anatomic alignment. Two proximal shaft fractures, one treated with an antegrade nail and the other with a retrograde nail, healed with >10° of varus angulation. Two distal shaft fractures treated with a retrograde nail healed with >10° of valgus alignment. No malalignment in the sagittal plane occurred. An interlocking screw located distant from the insertion point of the nail missed the nail in two patients, but both of the fractures healed uneventfully. One of these screws was located in a fracture fragment that had a second (appropriately placed) interlocking screw, and the other was in a length-stable fracture.
No patient had a leg length discrepancy of >2 cm. Forty-seven (94%) of the patients regained full range of motion at the knee. The three patients with a decrease of >20° in knee motion had all undergone retrograde nailing, and two of these patients had sustained a nondisplaced patellar fracture intraoperatively during knee manipulation performed because of stiffness resulting from delayed treatment. Both patellar fractures were detected intraoperatively and were treated with Ethibond (Ethicon, Somerville, New Jersey) suture fixation followed by immobilization in a long leg splint postoperatively. The final range of knee motion was 0° to 90° and 0° to 20° in the two patients with a patellar fracture and 0° to 90° in the third patient with knee stiffness. All fifty patients regained full hip motion.
Complications and Subsequent Procedures
Seven (14%) of the patients had complications that required further treatment (Table II). As previously noted, two patients required dynamization for the treatment of delayed union and two patients required treatment of an iatrogenic patellar fracture. In addition, one patient with HIV developed a superficial infection that resolved with antibiotics. Another patient developed a deep infection that was treated with suppressive antibiotics until fracture union occurred, and the infection then resolved after implant removal. One patient treated with retrograde nailing had prominence of the nail at the knee and required implant removal after fracture-healing. Seven additional patients had complications (such as the previously noted malalignment on radiographs and joint stiffness) that did not require further treatment. Thus, the overall complication rate was 28%. Ten of the fourteen patients who developed a complication had been treated with the retrograde nailing technique.
This retrospective analysis of prospectively collected data demonstrated that the SIGN intramedullary nailing system can be used effectively to treat femoral shaft fractures in a resource-poor setting that lacks intraoperative real-time imaging, power reaming, or a fracture table. Certain complications secondary to resource limitations may be expected.
All fractures healed after intramedullary fixation with the SIGN nail, although two fractures required nail dynamization for the treatment of delayed union. No implant failures occurred. Two interlocking screws located distant from the insertion point missed the nail, but this did not lead to any further complications. Four fractures healed in malalignment. Two proximal fractures healed in varus angulation of >10°; in one case this was because a retrograde nail that was too short was used in a fracture that did not have axial stability. Two fractures healed in valgus angulation because the fracture was at the diaphyseal-metaphyseal junction, which allowed angulation at the fracture site. One patient had a nail that was prominent at the knee and required removal after fracture-healing. These complications would not be acceptable in a setting where intraoperative imaging was available but may be expected in its absence. In the absence of fluoroscopy to evaluate the final position of the nail, a larger arthrotomy may be required to ensure appropriate placement of the nail before closure.
The use of the retrograde nailing technique may be associated with a higher complication rate. Two patients treated with retrograde nailing in our series sustained an iatrogenic patellar fracture during manipulation for knee stiffness resulting from delayed treatment (>17 days after the injury).
Comparison of our complication rates with those of other series involving the SIGN nail is limited by the paucity of previously published research. Ikem et al. prospectively analyzed forty patients who underwent SIGN nailing of the femur, tibia, or humerus in Nigeria and reported complications involving delayed union (two fractures, 5%), superficial wound infection (two, 5%), and screw loosening (one, 2%)7. However, interpretation of their results is hindered by the fact that their report was not stratified by the type of long bone treated or by the indication for surgery (acute fracture, nonunion, or malunion). Ikpeme et al., also in Nigeria, prospectively studied the outcomes of thirty-seven SIGN nailing procedures performed for the treatment of acute fracture, nonunion, or malunion of the femur or tibia; they reported complications involving osteomyelitis (four fractures, 11%), superficial infection (three, 8%), and nonunion with hardware failure (one, 3%)8. Although interpretation of their report is also hindered by lack of stratification by long bone or by indication, knee stiffness was documented in two of the six patients treated with the retrograde technique.
In spite of the complications associated with SIGN nailing of femoral shaft fracture, the procedure still compares quite favorably with traction treatment. Gosselin and Lavaly5 recently reported on a series of fifty-four femoral shaft fractures managed with Perkins traction in Sierra Leone, a West African country with an economic profile similar to that of Uganda (including a GDP per capita that ranks 220th out of 229 worldwide)9. Although their series included a small number of open fractures (four, 8%), the group was otherwise similar to our cohort with respect to age (mean, 34.2 years), sex (81% male), and mechanism (involving a motor vehicle in nearly two-thirds of the cases). Traction was applied with use of the Perkins method, a technique that involves straight traction through a Steinmann pin placed in the proximal aspect of the tibia, and active knee range-of-motion exercises were started within three to four days. Gosselin and Lavaly documented complications including nonunion (four, 7%), malunion (five, 9%), shortening of >2.5 cm (three, 6%), refracture after mobilization (two, 4%), and knee stiffness (eleven of thirty-four, 32%). Nearly half of the patients (twenty-three, 43%) developed an infection, including seventeen (31%) who required pin removal and antibiotics, four (7%) who required insertion of a second pin distally, and two (4%) who required surgical debridement to remove a sequestrum. The mean duration of traction was forty-five days (range, twenty-three to sixty-six days), and the mean length of the hospital stay was fifty-two days (range, twenty-five to 108 days)5. The results of our current study of SIGN nailing, including the rates of nonunion (0%), malunion (8%), shortening (0%), refracture (0%), knee stiffness (6%), and infection (4%) and the mean time from surgery to discharge (6.9 days), compare favorably with the values reported by Gosselin et al. following traction treatment.
Although we did not include a cost-effectiveness analysis in our study, prior research has shown that the shorter hospital stay and lower complication rate associated with SIGN nailing can actually render it more cost-effective than traction in the developing world6. Given that the majority of our patients were young men—individuals who often represent the most economically productive segment of the population—the benefits of SIGN nailing are likely to be even more substantial if lost wages are taken into account.
Angelini et al. recently reported on a technique involving percutaneous wave plating as a more affordable alternative to intramedullary nailing for the treatment of femoral shaft fractures12. Their study, performed in Brazil, differed from ours in that intraoperative fluoroscopy was utilized, AO/OTA type-B and C femoral shaft fractures were excluded, and patients were restricted to toe-touch weight-bearing postoperatively. In spite of these differences, Angelini et al. still reported a higher rate of nonunion (three of fifty-seven, 5%) and a higher rate of implant failure (two, 4%) than those in our study of intramedullary nailing (in which both rates were 0%). Although plate fixation may represent a viable alternative in settings where intramedullary nails are not available, reamed intramedullary nailing remains the standard of care for the treatment of femoral shaft fractures.
One other finding of our study deserves emphasis. As has been documented in a number of prior reports1-4,7,13, accidents involving motor vehicles represented the predominant mechanism of injury in our study (88%, forty-four of fifty). Although automobiles account for the majority of traffic in Kampala, automobile drivers and passengers accounted for only a minority of the injuries that involved motor vehicles (seven of the forty-four, 16%); the majority of those injured in such accidents were vulnerable road users such as motorcyclists (eighteen, 41%), pedestrians (seventeen, 39%), and bicyclists (two, 5%). These results emphasize the need for improved road traffic safety—in Uganda and in the rest of the developing world—and the need to pay special attention to the safety of pedestrians, motorcyclists, and other vulnerable road users.
Our study was inherently limited by the absence of a comparison group and by the lack of adequate follow-up of twenty (29%) of the seventy consecutive patients. However, follow-up in resource-poor areas is often limited because of financial issues and transportation factors.
In conclusion, the SIGN intramedullary nailing system promotes predictable healing of closed femoral shaft fractures in a resource-poor setting. Although complications may be expected in the absence of real-time imaging, especially when the retrograde nailing technique is used, the complication rate remains low compared with that associated with traction treatment of femoral diaphyseal fractures.
NOTE: The authors thank the members of the Mulago Department of Orthopaedic Surgery for their support of this study and Faridah Nalubwama for her assistance with data management.
Investigation performed at Mulago National Hospital, Kampala, Uganda