Outcomes of internal fixation versus hemiarthroplasty for elderly patients with an undisplaced femoral neck fracture: a systematic review and meta-analysis

In current practice, internal fixation has been the treatment of choice for undisplaced femoral neck fractures. However, most studies have reported a high reoperation rate after internal fixation (ranged from 8 to 34.6%) [1‐10]. Therefore, several alternative options have been discussed. One of the most commonly performed surgeries is hemiarthroplasty [11‐14]. In current literature, the perioperative outcomes between hemiarthroplasty and internal fixation remain inconclusive with regard to reoperation rate, mortality rate, and functional outcome. Sikand et al. validated that hemiarthroplasty surgery was an independent risk factor for increased 1-month and 1-year mortality [12] while two other studies did not find a difference [11, 13]. In terms of reoperation rate, two studies noted similar results between the two treatment modalities. In contrast, Dolatowski et al. found a lower reoperation rate in the hemiarthroplasty group. On the other hand, improved functional outcome was noted in patients that underwent hemiarthroplasty [11], while results from another study did not reveal a difference [13].

Due to these inconclusive results, we conducted this meta-analysis to evaluate several outcome parameters for elderly patients that underwent either hemiarthroplasty or internal fixation for undisplaced femoral neck fractures. We hypothesize that patients receiving hemiarthroplasty is associated with a lower risk of reoperation and will have improved functional status compared with patients that underwent internal fixation.

In this meta-analysis, we evaluated the outcome of elderly patients with undisplaced femoral neck fractures that underwent internal fixation or hemiarthroplasty. This study provides a synthesis of evidence from two randomized controlled trials and one prospective observational study to determine the optimal procedure in dealing with elderly patients presenting with undisplaced femoral neck fracture.

Our primary outcome comprised of reoperation rate and 1-month and 1-year mortality rate. Our analysis revealed a significantly higher reoperation rate in the internal fixation group (OR 4.489, 95% CI 2.030 to 9.927). One-month and 1-year mortality rate was equivalent between internal fixation and hemiarthroplasty. Secondary outcome included Harris Hip Score at postoperative 1 year and 2 years, length of hospital stay, and operation duration. The internal fixation group was associated with a shorter length of hospital stay (SMD − 0.800, 95% CI − 1.011 to − 0.589) and a shorter operation duration (SMD − 2.497, − 2.801 to − 2.193). Harris Hip Score at postoperative 1 year and 2 years was not different.

In current literature, the treatment of choice for undisplaced femoral neck fractures is with internal fixation. However, there is a high reoperation rate (8 to 34.6% [1‐10]), which has led to several alternative options being proposed. Several complications such as loss of fixation, nonunion, and avascular necrosis are the most common reasons for reoperation. A subsequent conversion to hip arthroplasty might be required to restore function and relieve pain. The reported rate of conversion to hip arthroplasty ranged from 8 to 16% [1, 5, 9]. In a prospective case series of 383 patients that received internal fixation for garden type I or II femoral neck fractures, 10% of patients had a salvage arthroplasty. It was estimated that up to one fourth of long-term survivors needed a conversion to arthroplasty [5]. Therefore, several studies have been conducted to validate hip arthroplasty as a possible alternative treatment for undisplaced femoral neck fracture [11‐14].

According to current literature, we defined “elderly” patients as a chronological age of 65 years or older [15]. Based on this definition, there was one prospective observational study and two randomized controlled trials that met our inclusion criteria. Sikand et al. enrolled 139 elderly patients with undisplaced femoral neck fracture surgically treated with internal fixation (N = 110) or hemiarthroplasty (N = 29). The authors noted a shorter operation duration and length of hospital stay in the internal fixation group. The reoperation rate was not significantly different between internal fixation (N = 8 of 110, 7.2%) and hemiarthroplasty (N = 1 of 29, 3%). However, patients that had undergone hemiarthroplasty surgery had a higher 1-month (21% vs. 2%) and 1-year (38% vs. 16%) mortality rate. In addition, pre-injury residential status was also a significant factor that influenced early mortality. Therefore, the authors concluded that they did not recommend hemiarthroplasty as the initial treatment option for undisplaced femoral neck fractures [12]. However, the functional outcome for these patients was not assessed in this study. Two randomized controlled trials were conducted to compare the outcome between internal fixation and hemiarthroplasty, including reoperation rate, mortality rate, and functional scores. Lu et al. [13] conducted a randomized controlled trial including 78 patients over 80 years of age that were treated with internal fixation (N = 41) or hemiarthroplasty (N = 37) for undisplaced femoral neck fractures. In patients that received internal fixation, there was a shorter length of surgical incision, operation duration, and length of hospital stay. There were also less blood loss and lower transfusion rates. There was a trend toward a higher reoperation rate in the internal fixation group (9/41, 22%) versus hemiarthroplasty group (2/37, 5.4%, p = 0.051). In contrast to the results from Sikand et al. [12], patient cumulative survival rates were similar between the two treatment groups. The Harris Hip Score at postoperative 12, 24, 36, 48, and 60 months was assessed, but there was no significant difference between the two groups [16]. In a multicenter randomized controlled trial conducted by Dolatowski et al. [11], 219 patients ≥ 70 years old with a nondisplaced femoral neck fracture received either internal fixation (N = 111) or hemiarthroplasty (N = 108). The Harris Hip Score (HHS) and other functional outcome domains including the timed “Up & Go” (TUG) test, pain intensity numerical rating scale (PI-NRS), EuroQol-5 Dimension-3 level scale (EQ-5D), and mini-mental state examination (MMSE) were assessed at prefracture and postoperative 3 months, 12 months, and 24 months [16‐21]. The patients in the hemiarthroplasty group demonstrated a significantly improved mobility (TUG test) in postoperative 12 and 24 months. The HHS, PI-NRS, and MMSE were similar at all time points between the two groups. For patients that received internal fixation, a shorter operation duration and length of hospital stay and less intraoperative blood loss were noted. There was a lower major operation rate (hemiarthroplasty vs. internal fixation, 5% vs. 20%, p = 0.002) and combined major and minor reoperation rate (hemiarthroplasty vs. internal fixation, 7.4% vs. 24.3%, p < 0.05) in the hemiarthroplasty group. Mortality rate was not different at postoperative 3, 12, and 24 months. The authors found hemiarthroplasty to be superior to internal fixation with regard to a lower major reoperation rate and improved function outcome as assessed by the TUG test. We conducted this meta-analysis because of the inconclusive results with regard to outcome domains including mortality rate, reoperation rate, and functional outcomes. Our analysis revealed that hemiarthroplasty for undisplaced femoral neck fracture in the elderly patients might be a viable treatment option compared with internal fixation in terms of a lower reoperation rate and an equivalent mortality rate.

The reasons for reoperation after an internal fixation or hemiarthroplasty surgery for an undisplaced femoral neck fracture are quite distinct from each other. In patients who had undergone internal fixation, the most common causes include fixation failure, nonunion, and osteonecrosis. A conversion to hip arthroplasty is usually necessary to restore mobility and relieve pain [1, 3, 5]. Several other reasons such as periimplant fracture and hardware irritation that required a revision surgery (fracture fixation and removal of implant, respectively) are other reasons that may result in additional surgeries [1, 3‐5, 9, 11‐13, 22]. For patients that received hemiarthroplasty, periprosthetic joint infection, dislocation, and prosthesis loosening were the most common causes of reoperation [11, 13]. In patients with periprosthetic infection, debridement and/or exchange arthroplasty may be required. For patients with recurrent hip dislocations or loosening of prosthesis, a revision surgery is often required [11‐13].

In addition to the medical benefits associated with hemiarthroplasty surgery (lower risk of reoperation etc.), the quality of life and healthcare-related costs are also important outcome domains that should be assessed. Dolatowski et al. conducted the only study that compared quality of life between the two groups using the EQ-5D index. The authors noted a higher EQ-5D index in the HA group 2 weeks before fracture and remained proportionate throughout the study [11]. Further studies are required to clarify whether this postoperative difference resulted from the type of surgery, preoperative status, or patient characteristics. There were several randomized controlled trials comparing total costs between hip arthroplasty and internal fixation for displaced femoral neck fracture in the first 1 or 2 years after the surgery [23‐25]. One study found similar total costs between internal fixation and hip arthroplasty when secondary surgeries were included [25], while results from other studies revealed that internal fixation was associated with higher total costs [23, 24]. Frihagen et al. reported a lower average cost for initial in-hospital stay (€9044 vs. €11,887, p < 0.01) but a subtle higher average total cost (€47,186 vs. €38,165, p = 0.09) in the internal fixation group [23]. The lower initial average cost in the internal fixation group was outweighed by the subsequent costs resulted from a higher reoperation rate. However, there is lack of similar study in patients with an undisplaced femoral neck fracture. Further study concerning average total cost, cost per quality-adjusted life year (QALY), disability-adjusted life year (DALY), and life years (LY) gained in patients with an undisplaced femoral neck fracture would be necessary to strengthen the conclusion of an ideal treatment choice.

This study is currently the first meta-analysis to compare the outcome between internal fixation and hemiarthroplasty for elderly patients with an undisplaced femoral neck fracture. However, there are several limitations that should be recognized. First, we searched only for English articles but not articles in other languages or unpublished data. This could be potential source of publication bias. Second, heterogeneity of clinical setting between studies including age, sex, medical comorbidities, internal fixation methods, surgical approaches, and types of implants for hemiarthroplasty should be recognized. Third, we were not able to analyze several outcome domains with clinical importance such as estimated blood loss, drop in hemoglobin, transfusion rate, health-related quality of life, and healthcare direct or indirect costs because of the limited literature.

The present meta-analysis revealed that hemiarthroplasty led to a lower reoperation rate compared with that of internal fixation. Mortality rate and functional outcome were not different. The findings suggest that hemiarthroplasty might be a better choice than internal fixation in treating elderly patients with an undisplaced femoral neck fracture.