Outcome of unstable pertrochanteric fractures in high-risk geriatric treated with external fixators

To review the outcome of high-risk geriatrics with unstable pertrochanteric fractures treated with external fixator. Eighteen consecutive patients with pertrochanteric fractures were operated with external fixator by using closed reduction technique. AO type, accompanying diseases, operating time, transfused blood units, duration of hospitalization, complications, healing time and mortality were recorded. Patients were followed clinically and radiologically for two years at least. The Harris Hip Score was used to document hip function at each regular follow-up. Fifteen patients with average age of 71.5 years were available for final evaluation. The mean operating time was 41.1 min and union was achieved in all cases with an average time of 16 weeks. The mean Harris Hip Score at one, three, six, twelve and twenty-four months post-operatively was 42.07 ± 8.55; 55.07 ± 11.62; 70.07 ± 10.32; 86.27 ± 9.06 and 89.27 ± 8.81 respectively. Complications included all the patients undergoing loss of some motion in the knee and four of them had to experience revision surgery because of knee stiffness, ten cases of pin-tract infections, seven cases of deep venous thrombosis, two cases of migration of the screws and two cases limb shortening. The use of external fixator in high-risk geriatric provided a satisfactory long-term efficacy, but the short-term functional results were quite unsatisfactory owing to high complication rate. Therefore, the advantages and disadvantages should be fully weighed when using external fixators, which was especially suitable for those patients who could not tolerate prolonged operative time and open surgery. Therapeutic study, level 4.


Introduction
Pertrochanteric fractures are one of the most common fracture types in the elderly and are associated with a high risk of mortality and functional disability. Surgery is usually indicated, bringing about pain relief, early weight bearing, and an early return to the preinjury functional level [1][2][3]. However, over time, in patients 65 years or older associated with hip fracture have had an increase in comorbidities [4]; moreover, the increase of preexisting medical conditions was highly significant factors correlated to patient mortality [5,6]. Therefore, the management of hip fracture for serious medical conditions remains challenging.
Some authors put forward that external fixation techniques, especially for high-risk elderly patients, which fully adheres to the concept of minimally invasive surgery, its biomechanical features are attractive, fixation is rapid, surgical reduction is maintained and post-operative complications have been minimal [7,8]. Nevertheless, compared to subcapital fractures the pertrochanteric ones are observed in older ages whose life expectancy is decreased and even lower or no walking expectancy [1,9], the long-term functional outcome of unstable pertrochanteric fractures for geriatric treated with external fixators has been rarely reported, the purpose of this study is to review our experience with the use of external fixation for unstable pertrochanteric fractures and to assess patient satisfaction using generic outcome instruments.

Materials and methods
This study had been approved by the Institutional Review Board of the authors' affiliated institutions and informed consent was obtained from participants to be included in the study. We retrospectively reviewed medical records for geriatric patients undergoing pertrochanteric fracture and treated with external fixation during January 2013 to December 2017. Inclusion criteria included age above 60 years, unstable pertrochanteric fracture, and American Society of Anesthesiologists (ASA) grades III or IV [10]. In a word, the indications for using the external fixator were severe fracture involved with the proximal femur where traditional methods of open reduction and internal fixation were deemed unsuitable.  [11], in which two part fractures with comminution of the posteromedial cortex, subtrochanteric extension or a reverse obliquity pattern were considered unstable [12].
All operations were implemented under epidural anaesthesia and performed by the same surgeon who had been majoring in hip fractures in our trauma centre for ten years, the patients were positioned on a fracture table in the supine position with the contralateral limb widely abducted. The fracture was reduced by applying traction and internal rotation to the injured limb in a position of slight abduction. When fracture reduction is achieved, no further traction is applied and care must be taken not to change the limb position throughout the remainder of the procedure. Fracture reduction must be checked with fluoroscopy-both anteroposterior and lateral views. Two guidewires were placed on the front of hip joint, which were, respectively, proved to locate in top one-third and bottom one-third of the femoral neck and head under anteroposterior fluoroscopy. Furthermore, two long, self-tapping and hydroxyapatite coated 6.5-mm Schanz screws were inserted first in a parallel to either one of the guidewires under lateral fluoroscopy and advanced to within 10 mm from the subchondral bone of the head. Attention was paid to ensure that the tip of the Schanz pin was at least 5 mm far from the articular surface to prevent penetration. Then, a unilateral, external fixator frame with uncoated stainless steel pins, which attached to the proximal pins, acted as a template for the introduction of the two distal pins, the two distal screws were inserted perpendicular to the long axis of the proximal femoral shaft and were implanted to a depth of two screw threads beyond the opposite bone cortex. The duration of surgery was determined as starting at the time of skin incision and ending when the final pin was secured to the fixator frame.
Post-operatively, all patients underwent a progressive physiotherapy measures, mobilization of the knee and ankle was encouraged 6 to 12 h after surgery, as tolerated, sitting on bed or on a chair on the first post-operative day, and active hip movements were also started under the supervision and guidance of physiotherapists. Pin sites were cleaned daily with povidone iodine and a family member was instructed how to take care of the surgical wound after discharge.
The patients were scheduled to be followed up clinically and radiologically at 1, 3, 6, 12, 24 months. Radiologic examination included anteroposterior and oblique views of the operated hip. Union was defined by the presence of trabeculae bridging the fracture site or obvious periosteal callus within the fracture line, then, the patients were allowed to fully weight bearing without any facility, if the fracture site was not displaced after continuous walking for two weeks, the external fixator was totally removed in our outpatient clinic. A fracture was determined to have a nonunion when no bridging callus formation was visible on both frontal and lateral hip radiographs after a minimum period of 15 weeks.
Besides, the quality of fracture reduction, as shown by the immediate post-operative radiographs, was evaluated according to the classification described by Baumgaertner et al. [13]. The reduction was considered good when normal or slight valgus alignment was evident on the anteroposterior radiograph, less than 20°of angulation on the lateral radiograph, and no more than 4 mm of displacement of any fragment. An acceptable reduction was characterized by a reduction that met the criterion of a good reduction with respect to either alignment or displacement, but not both. Finally, a poor reduction did not meet any of the above criteria.
In a word, accompanying diseases, operating time, transfused blood units, duration of hospitalization, complications, healing time and mortality were recorded at final follow-up. The Harris hip score was adopted to assess patient-activity status, any residual pain, claudication and hip mobility [14],the explanation of outcome using the Harris hip score was as follows: < 70 (poor result), 70-79 (fair result), 80-89 (good result) and > 90 (excellent result), Paired t-test was used to evaluate for statistically significant difference in the change of HHS from one to three months post-operatively, from three to six months post-operatively, from six to twelve months post-operatively and from twelve to twenty-four months post-operatively. The level of significance was set at 5%.

Clinical results
This cohort study consisted of 18 consecutive patients who were treated with the Orthofix pertrochanteric external fixator from January 2013 to December 2017, three patients were lost to follow up due to mortality and remote rural areas and finally functional outcome data at one, three, six, twelve and twenty-four months after the surgical intervention was available for 15 patients.
Eight were men and seven were women with average age 71.5 (60-88) years (Table 1). All the patients were suffering from one or more accompanying diseases ( Table 1). The distribution of the fractures according to the A.O. Classification is presented in Table 1. The first post-operative day, 80% of the patients sat on a chair or the edge of a bed, and 33.3% walked on the second day with the help of the medical staff. Mean duration of operation was 41 (30-60) min. Haemoglobin less than 9 g/dL was observed, at least once during the hospitalization period, in 86.7% of the patients (13 patients), the treatment of 11 of the patients did not necessitate any blood transfusion, for the remaining 4 patients were needed. Finally, all patients healed well, the average time to union was 16 weeks (range, 14-20 weeks), Surprisingly, loss of some motion in the knee occurred in fifteen patients as pain in the entrance of the screws when the mobilization of the knee joint was encouraged, and its extent depended mostly on the severity of the pain, which led to many patients could not wear shoes and socks by themselves due to squatting on toes could not be accomplished in their daily life. An average of 60 degrees of range of motion of the knee was lost in these fifteen patients at one month after surgery. At one-year follow-up, four patients had to be operated again for knee stiffness while the remaining still lost some motion of knee joint and further surgical interventions were not required.
Ten pin-tract infections, most of which resolved with local care and antibiotics, only one patient pin-tract problems necessitated removal of the apparatus sooner than desired, and a brace was substituted. Finally, the fracture healed with an acceptable position after the removal of the external fixator at three months, manipulation was carried out and then a walking hip-spica brace was worn for an additional four weeks.
The most serious complications were deep venous thrombosis that affected 7 patients. Four patients were transferred to orthopaedics from neurosurgery at least five days after vehicle accident, who were all associated with closed head trauma at the time of accident, when pre-operative preparation was administrated, deep vein thrombosis were founded in these patients and three of them presented with pulmonary embolism. Another one with a sever pneumonia was admitted three days after fall from stand because consensus was not achieved among family member at that moment, similarly, deep vein thrombosis was discovered on admission. The others were detected as complaining continuous pain of the calf after discharge from hospital. In two cases, migration of the screws further into the femoral head was observed at the follow-up X-rays and in one of them the screws penetrated into the joint space. These cases were treated with withdrawal of the screws and controlled loading of the affected limb until the fracture union and the subsequent removal of the screws.

Radiological results
The quality of the final reduction was rated as good in 60% of patients (9 patients), acceptable in 27% (four patients) and poor in 13% (two patients). This result was mainly attributed to an increased incidence of varus position of the fracture (greater than 10º) in these unstable fractures of patients (Fig. 1). Besides, slight convergence of the proximal screws was noticed in 33% of patients.

Functional outcomes
The mean HHS at one, three, six, twelve and twenty-four months post-operatively was 42.07 ± 8.55; 55.07 ± 11.62; 70.07 ± 10.32; 86.27 ± 9.06 and 89.27 ± 8.81 respectively. Paired t-test showed that the progressive improvement in HHS from one to twenty-four months post-operatively Fig. 1 A 77-year-old male, AO type 31 A3.3. a.Preoperative anteroposterior X-ray; b.Pre-operative 3-D CT scan; c.Immediate postoperative anteroposterior X-ray; d.6 months post-operative anteroposterior and oblique X-ray, patient full weight bearing; e.12 months post-operative anteroposterior and lateral X-ray, the quality of the final reduction was rated as poor was significant (p = 0.000, 95% CI: 8.3 to 17.7; p = 0.000, 95% CI: 10.7 to 19.3; p = 0.000, 95% CI: 12.9 to 19.5; p = 0.002,95% CI: 1.3 to 4.7; respectively). At twentyfour months post-operatively, excellent, good, fair and poor outcome was seen in 9 patients, 4 patients, 1 patient and 1 patient, respectively. Figure 2 shows the trend in the improvement of HHS and Fig. 3 displayed the variation tendency of each item of HHS post-operatively during the follow-up period. Of the fifteen pertrochanteric fractures, one patient (Case 7) got the worst HHS 68 and another one patients gained the best HHS 100(Case 12) at final follow-up.

Discussion
It is widely accepted that the treatment of pertrochanteric fractures in high-risk geriatric is to offer rigid fixation at fracture site allowing early mobilization as soon as possible so as to reduce morbidity and mortality. As pertrochanteric fractures are mainly emerged in elderly patients where the rate of the accompanying diseases is added [1,9]. Therefore, the surgeon has to choose a method that its application will be fast, with a simple pre-operative planning, which should be as minimally invasive as possible, with minimal blood loss and one that ensures the rapid mobilization of the patient and his/her discharge from hospital. The final result of this decision will be the decrease of post-operative complications and deaths and the assurance of an as early as possible return of the elderly patients to the walking condition they had before the fracture with the least medical expense. The external fixation meets these requirements [1,15].
Though intramedullary nailing has also been the most commonly used procedure to treat unstable pertrochanteric fractures as its biomechanically and biologically property are superior to extramedullary implants [3,16], but it has its  [17][18][19]. First, anatomic reduction was the primary goal for pertrochanteric fractures since nonanatomical reduction inevitably leads to intramedullary nailing implant malpositioning, and which was closely related to mechanical failure of internal fixation of extracapsular proximal femoral fracture in this fragile population [20]; hence, prolonged operative time was needed in order to achieve anatomical reduction, on the contrary, to some extent malreduction was acceptable when external fixator was applied, especially for these elderly patients have associated medical problems of hypertension, diabetes mellitus, ischaemic heart disease and poor general condition labelling them high-risk surgical candidates for open reduction and internal fixation of their fractures [17]. Second, blood loss could aggravate functional outcomes and increase mortality in patients with hip fracture by lowering haemoglobin levels. Carmelo et al. reported that the mean visible blood loss was 226 ml in the Proximal Femoral Nail Antirotation group versus 285 ml in the Trochanteric Gamma3 Nail group observed intraoperatively in the treatment of intertrochanteric fractures [21], which did not contain hidden blood loss after surgery. Similarly, Foss et al. found that the overall blood loss was 1473 mL greater than that observed intraoperatively in patients undergoing hip surgery [22],and Yu et al. [23] further reported 277.2 ± 7.6 mL hidden blood loss in patients undergoing proximal femoral nail anti-rotation (PFNA) for intertrochanteric fractures. Oppositely, there was hardly any bleeding occurred in these patients treated with external fixator intraand post-operatively.
Contrary to our findings, however, Zhang et al. [[24]] reported that the long-term functional outcome studies on HHS value in pertrochanteric fractures at 36 months postoperatively was lower than the HHS at one month post-operatively, compared to their mean HHS was 80.51 ± 2.91 at one month post-operatively whilst our mean HHS was only 42.07 ± 8.55, It was probable that better explanation in our study was owing to the patients associated with higher risk accompanying disease pre-operatively. With the cooperation between the physiotherapist and the patient gradually strengthened over time, the patients HHS tended to improve in the first twelve months, tended to plateau between twelve to twenty-four months post-operatively even though revision surgery were performed for four patients who were dissatisfied with functional outcome because of knee stiffness, which was in accordance with the results from previously published studies [25,26]. Of course, the longer-term follow-up in these patients with proximal femoral fracture was challenging due to loss to follow-up bias and high mortality [24,27], but it could be speculated that HHS tend to decrease with age in the geriatric population that sustained intertrochanteric fracture and treated with exterior fixator [28] Despite the mentioned benefits above, the complication rates of the technique-and patient-related could not be considered negligible and knee stiffness, pin-tract infection, deep venous thrombosis, varus angulation of the fracture (greater than 10°) and limb shortening were too frequently encountered. In the current study, 15 cases knee joint stiffness or joint mobility declined, 9 cases pin-tract infection, these results were approximately consistent with the findings of the similar bibliography in regard to femur fracture treated with external fixator [29,30], knee stiffness and pin-tract infection were problems inherent with the use of this method, Babar [29] had reported that fourteen patients with close comminuted fractures of the shaft of the femur were treated with external fixation, as a result, 11 lost an average of 34 (0-75) degrees of range-of-motion in their knees, likewise, Zhang et al. [30] claimed that 14 cases knee joint stiffness were observed in the treatment for femoral nonunion with segmental bone defect with monolateral external fixation. External fixator had an adverse impact on the activity and function of knee joint and some factors might be attributed to this phenomenon. Above all, the extension and contraction mobilization of the thigh muscles were greatly limited by screws inserted into the femur, which lead to the activity of knee joints worsen below the entrance of the screws due to it impede the movement of the iliotibial band, vastus lateralis and vastus intermedius when knee is flexed, namely screws block the knee motion. Second, adhesions and tissue degeneration were usually occurred between muscles or between muscles and aponeurosis as a result of injury caused by screws when implanting, especially deteriorated if knee motion was repeatedly happened since it would inevitably lead to the inflammatory reaction because these organizations and screws that were rubbed against each other. It was a living proof that most patients complained of irritation and pain in the opening site of entrance of the screws when knee exercising, worse still, some of them undergone local pyogenic infection. In turn, the adhesions and tissue degeneration would further affect knee mobilization. Unfortunately, four patients had to be hospitalized again and be treated for knee stiffness owing to the inflammatory reaction of the soft tissue that finally lead to these organization fast degeneration, lower elasticity, and even fibrosis. As expected, which were totally consistent with our intraoperative finding. From the reasons mentioned above, it might be speculated that soft tissue adhesions and degeneration was the main cause that affect knee return to the preinjury level even though the external fixator was removed after fracture healed.
In accordance to Kazemian study [19], of the twenty pertrochanteric fractures, ten patients had pin track infection, similarity, there were nine pin-tract infection in our study, no chronic osteomyelitis occurred and all were treated with frequent pin care together with oral antibiotics.
It was reported that the pre-operative prevalence rate of deep vein thrombosis associated with fractures about the hip was 2.6% ~ 29.8% [31][32][33], what's more, Aging and Delayed surgery have been proved to be risk factors of deep venous thrombosis for patients with hip fractures, to be more specific, the incidence increased each day from 14.5% if surgery was delayed > 1 day to 33.3% if surgery was delayed > 7 days despite prophylactic anticoagulation [34,35]. In our group, four patients who for head injury and pulmonary infection experienced a delay of > 5 days from the time of injury until time of surgery and developed pulmonary embolus preoperatively for an overall incidence of thromboembolic disease of 26.7%. Therefore, on one hand, it was of paramount important that mechanical measures were immediately performed to minimize deep venous thrombosis or clinically detected pulmonary embolism when patients upon admission if pharmacological prophylaxis was contraindication, on the other hand, to shorten delayed surgery as soon as possible, a comprehensive, multidisciplinary cooperation was indeed needed to ensure that established protocols were consistently implemented, which would ensure a uniform and high standard of care for these high-risk elderly patients with intertrochanteric fractures.
In our study, varus angulation of the fractures (greater than 10º) with subsequent limb shortening were two cases and this might be on account of significant convergence of the proximal pins with respect to the middle of the femoral neck, besides, another explanation could be the applied device configuration might influence the rate of malreduction and malunion, particularly in these unstable fractures cohort, as a result of increased working length of the spanning external fixator, since a shorter working length should probably increase the rigidity and strength of the construct and the overall fracture stability.
There are several limitations to our study. First, the small sample size may have introduced bias. However, the focus of our study was to assess the 2-year follow-up functional and radiographic outcomes of the external fixator in the treatment of unstable pertrochanteric fractures in elderly patients as open reduction and internal fixation was relative contraindications, an area that has not been studied extensively in the literature. Despite the small sample size, it was still possible to provide useful information with regard to predict the long-term functional outcome of these patients. Second, this study had problems inherent to its retrospective nature. Finally, because of the observational nature of this study, we may not have addressed all potential confounding variables in our analyses. Nevertheless, this study provides preliminary objective evidence about the effect of unstable proximal femur fractures treated with external fixation. Despite these limitations, this study presents long-term follow-up results and evaluates the causes that impact the clinical outcome of patients undergoing fixation.
In conclusion, the use of external fixator in high-risk geriatric provided a satisfactory long-term efficacy, but the short-term functional results were quite unsatisfactory owing to high complication rate. Therefore, the advantages and disadvantages should be fully weighed when using external fixators, which was especially suitable for those patients who could not tolerate prolonged operative time and open surgery.