Outcome of bipolar hemiarthroplasty following displaced femoral neck fractures: A multicenter study

INTRODUCTION

There is an increase in the prevalence of hip fractures worldwide due to an increase in the aging population, with a corresponding increase in the occurrence of osteoporosis.^[1] An estimated 1.5 million hip fractures are reported worldwide each year, and by 2050, this is expected to be 3.9 million, with a high predilection for the Caucasian population and an increase as we move away from the equator.^[2] It is not common in the black population. An incidence of 19.3/100,000 population was reported in a study done in South Africa, with the least occurrence in black people, unlike in the Caucasian/white population in South Africa, with a significantly higher incidence of over 60/100000.^[3,4]

Hip fractures comprise fractures of the femoral head, femoral neck, intertrochanteric, and subtrochanteric fractures, with 53–60% accounted for by femoral neck fractures.^[5] Of the femoral neck fractures, more than two-thirds are displaced, making it a surgical emergency with increased morbidity associated with delayed treatment.^[4] Femoral neck fractures occur most commonly in elderly female patients, usually above 60 years of age, often following trivial falls, although a few cases occur in the middle-aged group as a result of high-energy trauma.^[6,7]

In general, the aim of surgical treatment is to achieve an immediate return to premorbid activity levels as early as possible, thereby reducing morbidity and mortality. The choice of treatment is based on the premorbid functional level, age, risk profiles such as cognitive function, and degree of physical fitness.^[8] Operative alternatives for displaced femoral neck fractures include internal fixation, hemiarthroplasty (HA), and total hip arthroplasty (THA). There seems to be a consensus that internal fixation should be the treatment of choice for younger patients with high demand and prolonged life expectancy. There is little choice of the optimal treatment option between bipolar HA (BHA) and THA in elderly patients. However, most surgeons perform BHA for frail elderly patients with low functional demand, while total hip replacement (THR) is reserved for elderly patients with higher functional demand.^[9,10] BHA following femoral displaced femoral neck fractures has a reported advantage of reduced dislocation rates, less complex surgery, shorter operation times, less blood loss, and fewer initial costs compared with THR. However, those who had THR had better hip function with a persistent increase in Harris Hip scores (HHSs) at 24 months, which extended up to 4 years.^[11] The study also demonstrated no stastically significant difference between THR and BHA in those with or without acetabular erosion. However, with regard to pain, function, age, and sex, the health-related quality of life, as measured by the European quality of life index score, was higher in the THA group at each follow-up, with the difference most marked at 48 months.^[12] Recent studies of THR and BHA in hip fractures found no significant difference in hip function, quality of life, and reoperation rates over a 24-month period among patients who were independent ambulators before the operation.^[13] For surgeons practicing in developing or low-resource centers, BHA remained the most versatile option available for the surgical treatment of displaced femoral neck fractures due to its ease, availability, low cost, and readily accessible surgical expertise. Therefore, this study aimed to evaluate the outcome of surgical treatment of displaced femoral neck fractures with BHA in Nigeria using surgical site infection (SSI), length of hospital stay (LHS), and functional outcome as the defined outcomes.

MATERIALS AND METHODS

This is a retrospective multicenter study involving three tertiary hospitals in the south-western and north-western parts of the country. The study period extended from January ^1st, 2010, to December 2024. The case files of patients who had BHA were retrieved from the medical records of the participating institutions, and relevant data for the study were collected using a predesigned pro forma.

The inclusion criteria were patients aged >60 years who had undergone BHA and had radiologically confirmed displaced intracapsular femoral neck fractures; all patients must have had at least 2 years of follow-up after discharge.

The exclusion criteria were all pathological fractures of the neck of the femur, those with unipolar HA, and femoral neck fractures treated with internal fixation. The outcome variables were SSI, LHS, and functional outcomes, which were assessed using the HHSs for pain and ambulatory status at the patients’ last clinic attendance.

Outcome variables are defined as follows: SSI is an infection that occurs within 1 year of the index operations when an implant is used;^[14] LHS is spent from admission till discharge, while functional outcome was assessed using modified HHS for pain and ambulatory status,^[15] as the last clinic attendance assessed as walking unaided, ambulation with a walking stick, ambulation with a walking frame, and wheelchair bound.^[16]

Data collected using the pro forma designed for this study were entered into IBM Statistical Package for the Social Sciences version 25 for analysis. Categorical variables were expressed as percentages, while continuous variables were expressed as means and standard deviations (SDs). The association between categorical variables was assessed using the Chi-square test and that between continuous variables was assessed using a paired t-test to compare means; the level of significance was set at p < 0.05.

RESULTS

A total of 84 BHAs were performed during the study period. Eleven (13.1%) were not included for not meeting the inclusion criteria, leaving 73 (86.9%) who underwent BHA for analysis.

Their mean age was 75 ± 10.7 SD years, with a range of 60–110 years. There were 36 males (49.3%) and 37 females (50.7%), resulting in a male-to-female ratio of 1:1.03. The majority of patients were aged 70–79 years, representing 32.5% (24 patients) of all patients [Table 1]. The median duration of injury before presentation was 4 days with a range of 1–156 days among which 49.3% (36) patients had had previous treatment at the traditional bone setters’ place, followed by 32.9% (24 patients) had no previous treatment, and the remaining 17.8% (13 patients) had received some form of treatment at private health institution before presentation at the tertiary hospital where the operations were performed. Further analysis showed that 15% (11 patients) of the patients who had previous treatment at the traditional hospital had various grades of SSI. This was statistically significant [Table 2]. The mechanism of injury (MOI) was a domestic fall in 71.2% (52 patients) and a road traffic injury in 20.5% (21 patients). Further analysis found that 61.5% (32 patients) of those with domestic falls were female, compared with 38.5% (20 patients) who were male. There was a statistically significant difference between the MOI and sex (p = 0.028) [Table 2]. The premorbid disease was present in 53.4% (39) of the patients, of which hypertension accounted for 42.5% (31 patients), followed by diabetic mellitus in 9.6% (7) patients, and 1.4% (1 patient) had chronic obstructive pulmonary disease. There was no premorbid disease in 46.6% (34 patients). The mean pre-operative hematocrit (HCT) was 35.4% ± 4.2 SD, with a range of 25–44%. Nine (12.3%) patients with an HCT of <30% had pre-operative transfusion of 1–3 pints of blood to optimize them for anesthesia. The remaining 64 patients (87.7%) had pre-operative HCT >30%. The fracture pattern was Garden’s type 3 in 29 patients (39.3%) and type 4 in 44 patients (60.3%). There was no statistically significant association between pre-operative HCT and radiological grading of the fracture neck of the femur. Pre-operative American Society of Anesthesiologists grades were grade 2 in 35 patients (48.0%), grade 3 in 28 patients (38.4%), and grade 4 in 10 patients (13.7%). The duration of admission before surgery ranged from 1 to 37 days, with a median of 7.0 days. However, further analysis showed that 7 patients (18.7%) with ≤1 week of admission before the operation had SSI, and 7 patients (19.4%) with more than 1 week of admission had SSI. This was not statistically significant. Spinal anesthesia was the most commonly used in 70 patients (95.9%), epidural in 2 (2.7%), and general anesthesia in 1 (1.4%). The most commonly used approach to the hip was the direct lateral approach of Hardinge in 53 patients (72.5%), the anterolateral approach in 13 patients (17.8%), and the posterior approach in 7 patients (9.5%). An uncemented BHA was used in 37 patients (50.7%) and a cemented BHA in 36 patients (49.3%). The median operating time was 95 min with a range of 38–215 min. The median intraoperative blood loss was 250 cc with a range of 75–1000 cc. The mean LHS was 19.3 ± 11.7 SD days, with a range of 4–54 days. The longest was a 68-year-old female patient who had a grade 3 SSI but no premorbid medical conditions. She had repeated joint washout and intravenous culture-specific antibiotics. Eventually, the infection resolved, and she was discharged.

The mean HHS for pain was 34 ± 10.3 SD. Although 63% (46 patients) had HHS of 40 and 44, which was good to excellent, using the status of the patients at the clinic attendance as a functional outcome showed that 18 patients (24.7%) were walking without aid, and 29 patients (39.7%) were using walking sticks as an assistive device. This represented 64.4% of the total number of patients, while the remaining patients used a walking frame in 3 (4.1%) or bilateral axillary crutches in 23 (31.5%) as assistive devices.

There was a statistically significant association between age group and the patient’s status at the last clinic attendance [p = 0.001; Table 3]. There was a statistically significant association between operating time and surgical approach (p = 0.004) [Table 3]. However, there was no statistically significant association between the operation time and the estimated blood loss. The mean LHS was 19.3 ± 11.7 SD days, with a range of 4–54 days. There was a statistically significant association between LHS and SSI [p = 0.001; Table 2], but no significant association with surgical approach. The mortality rate was 4.1% (3 patients). Two of the deaths were not related to the operation because the patient had a complication of a pre-morbid medical condition. The deaths occurred at 2 and 3 months postoperatively. The diabetic patient was brought unconscious to the hospital and was managed by the endocrinologists. The second patient had probably had dysthymia before arrival in the hospital and died during cardiopulmonary resuscitation. He was also treated by a cardiologist. The third patient had a grade 2 pressure sore before admission, following a visit and treatment by the traditional bonesetters (TBS). Subsequently, he developed deep SSI.

DISCUSSION

The aim of treatment for the fracture of the neck of the femur is an early return to premorbid functional state, although this is often dictated by the treatment options available in the environment. There is a consensus that internal fixation should be performed in younger patients with high functional demands; the choice of treatment for the elderly is between THR and BHA. Surgeons preferred BHA as the treatment of choice for the elderly with low functional demand. THR is the preferred treatment for the elderly with a displaced fracture of the neck of the femur who have functional demands, active status, and well-controlled comorbidities.^[17,18] But for surgeons practicing in developing countries where affordability, availability of expertise, and surgical instrumentation may be lacking or not easily accessible, then BHA remains the only option available. This is because it is affordable, readily available, and surgical expertise is accessible. It also has the advantage of less complex surgery, shorter operating time, less blood loss, and an outcome similar to that of THR in the short term.^[17,19] In this study, the mean age of the patients was 75 years, with an almost equal male-to-female ratio, similar to those reported by Khorami et al.^[20] Although their study included both unipolar and bipolar hemiarthroplasty as well as younger patients compared to the present study. Our study included only those aged 60 and above. A second possible explanation for our findings is the small sample size. In a small sample, the pattern of injuries might coincidentally reflect the pattern seen in different sociocultural strata of the country, even if that’s not true for the whole population. The eighth decade of life was most predominantly affected; this is akin to reports by Abd El-Naby et al. and Fahad S. et al.^[21,22]

Although complications arising from the treatment of long bone fractures with TBS have been widely reported,^[23] their involvement in the treatment of femoral neck fractures remains poorly reported in the literature.^[24] This often leads to delays in presentation to the hospital for early treatment, which guarantees a very early return to function. Approximately half of the patients in this study received their initial treatment from TBS before presenting to the tertiary hospitals, where they underwent surgery. This common practice continues to plague orthopedic care in low-resource countries.^[25-27] Our study’s overall SSI of 19.1% is considerably higher than the 1.3–2.14% range for superficial-to-deep infections in studies with undefined criteria and is higher than the 2.94% rate in studies with defined criteria.^[28] This might be due to delayed presentation to the hospital caused by a visit to the TBS, because 25% of those who had their initial treatment with the TBS had SSI compared to 13.5% of those who did not.^[28] In addition, our small sample could have magnified this occurrence. Previously published data on outcomes from TBS treatment reported varying degrees of infection in closed fractures but a higher rate in those with open fractures.^[23,27,29] The association between SSI and BHA might be a new finding contributing to the body of knowledge on TBS-related complications. A recent systematic review recommended intersectoral collaboration to stem the tide of their menace.^[25,30,31] In addition, these results underscore the role of pre-hospital care –seeking behavior as a modifiable risk factor for SSI in our setting.^[32] The MOI was almost equally shared between domestic falls and road traffic injuries. This was statistically significant with a moderate association. This compared favorably with a multicenter study from southeastern Nigeria that focused on the mechanisms of hip fractures and associated risk factors.^[33] This may be due to shared sociodemographic status, although our study involved the north-west and south-west geopolitical zones of the country.

Preoperative anemia is commonly reported among patients with displaced femoral neck fractures. Garden 3 and 4 grades of the fracture of the femoral neck have been associated with increased risk of morbidity and mortality.^[24,34] In our study, only eight patients (11%) had preoperative transfusion to optimize them for anesthesia. There was no statistically significant association between preoperative anemia and Garden’s grade of displacement of the femoral neck fractures. This is likely because Garden’s grade of displaced fracture describes fracture displacement based on biomechanical factors, whereas HCT is influenced by the baseline patient comorbidities, hydration status, and the timing of laboratory assessment.^[35]

Spinal anesthesia was the most predominantly used anesthesia in this study. This was consistent with studies by Sharma et al., which favored its use in more than 60% of patients who underwent HA.^[34]

A recent scoping analysis of approaches to hip fracture, comparing Hardinge’s direct lateral approach with the posterior approach, was inconclusive in its recommendation. The study suggested that a randomized clinical trial should be conducted to provide appropriate recommendations.^[36] Although some surgeons favor the use of the posterior approach because of its ease of use, shorter operating time, and reduced blood loss, its major flaw lies in the higher rate of post-operative dislocation. A few others favor Hardinge’s direct lateral approach for its low dislocation rates, though it is associated with hip pain that may hinder post-operative ambulation.^[37] In our study, the direct lateral approach was most commonly used, and its selection was based on the surgeon’s experience and preference. This may be the explanation why over one-third of our patients had mild pain, which warranted occasional use of analgesics but did not negatively impact the activity of daily living.

The operative time is usually a function of the surgeon’s experience, the surgical approach, and sometimes the physical build of the patient. The mean operating time of 72.8 min was reported in a study that included only octogenarian patients, a single surgeon, and a posterior-lateral approach.^[38] This was less than the median operating time of 95 min in our study. This may be because our study population was heterogeneous and multicenter, with different approaches to the hip. A meta-analysis of hip approaches and operative time found no statistically significant association, although the conclusion was based on two approaches: Posterior lateral and lateral approaches. This differs from our study, where three approaches were analyzed.^[39]

The mean LHS across most studies ranges from 4 to 12 days, usually <1 week, as reported by Wang et al. and Marya et al.^[18,19] Our findings, 19.5 days, were longer than compared to this study but it was shorter than those reported in another study where it was reported that those <85 years had a shorter LHS of 25 days compared to those above 85 years who spent a mean of 41.5 days. Furthermore, the study also reported that factors associated with prolonged LHS following BHA were pre-operative morbid conditions, such as diabetes and high-level walking activity.^[40] In addition, another study also found that those who had a direct anterior approach had a significantly shorter LOHS compared to those who had an anterolateral approach.^[41] There was a statistically significant association between the LHS and SSI. This is expected because those with SSI will require extra days for their wound care, coupled with delayed mobilization and discharge planning.^[40] The functional outcome, using the HHS for pain and patient status at the last clinic visit, was good to excellent in 63% of our patients, which compares favorably with reports by Yurdakul et al. and Kerakkanavar et al.^[42,43]

CONCLUSION

BHA had low SSI, low mortality, and good to excellent functional outcome, making it a viable option in the surgical treatment of displaced femoral neck fractures in a resource-constrained environment. However, this should be interpreted within the context of its retrospective design, selection bias, and small sample size, which are the limitations of the study.