1 Introduction

Periprosthetic femoral fractures (PFFs) are a serious and increasingly common complication after hip arthroplasty. With the aging of the global population, the number of hip arthroplasty procedures, including both total hip arthroplasty (THA) and bipolar hemiarthroplasty (BHA), is rising annually. Consequently, the incidence of PFF is also increasing, with reported 10-year rates of up to 3.5% after THA and 1.5–2.1% after BHA.^1,2 These fractures significantly impair patients' quality of life, often necessitating revision surgery and prolonged hospitalization, and thus represent a major challenge in orthopedic surgery.

While PFF can occur after both THA and BHA, the patient populations and fracture mechanisms are distinct. THA is typically performed electively for chronic conditions such as osteoarthritis in younger, more active patients. In contrast, BHA is often an emergency procedure for femoral neck fractures in frail, elderly individuals who frequently have multiple comorbidities, such as cardiovascular disease, diabetes, and dementia, in addition to osteoporosis.³ Many BHA recipients already have a limited ambulatory capacity before surgery, making them functionally more vulnerable. As a result, PFF in BHA patients can occur from minor trauma and presents unique challenges that complicate treatment and affect outcomes.

Among PFF types, Vancouver B2 fractures are characterized by a loose stem with adequate bone stock and are generally indicated for revision arthroplasty according to the classification system.⁴ Revision using long-stem prostheses allows for stable fixation and early postoperative weight-bearing, which are considered to be advantages of this approach⁵

However, the significant surgical stress and perioperative risks associated with revision arthroplasty are major concerns for the frail, medically complex patients who undergo BHA. In this context, open reduction and internal fixation (ORIF) has emerged as a less invasive alternative. ORIF may offer benefits such as shorter operative times and reduced physiological burden, particularly in high-risk patients. Nevertheless, achieving stable fixation with ORIF can be difficult due to poor bone quality and the existing intramedullary stem. The efficacy of ORIF remains controversial, with some studies reporting high rates of nonunion and reoperation.^6,7

Although numerous studies have addressed the treatment of PFFs, research focusing specifically on PFFs after BHA is scarce, and evidence is limited in comparison to that for PFFs after THA.

The choice between revision arthroplasty and ORIF for Vancouver B2 fractures after BHA presents a clinical dilemma for orthopedic surgeons. Therefore, this study aimed to retrospectively compare the clinical outcomes of ORIF and revision arthroplasty for these specific fractures. We focused on mortality, the ambulatory function, and reoperation rates to help establish optimal treatment strategies and improve the management of PFF in elderly patients treated with BHA.

2 Materials and methods

The study was approved by the Institutional Ethics Committee of Nagoya University School of Medicine (Ref. 2020-0564, approved on March 24, 2025). In accordance with institutional policy, an opt-out consent process was implemented, and participants had the opportunity to decline participation.

2.1

2.1 Patient selection

Patient data for this multicenter retrospective study were sourced from the Trauma Research Group (TRON) registry. The methodology of this registry, covering 16 tertiary trauma centers in Central Japan, is detailed in our previous work.⁸ In accordance with the previously approved protocol, ethical clearance was obtained from each institution, and an opt-out consent method was utilized.

A search of the registry for the period between 2010 and 2023 yielded 285 PFF cases, of which 238 occurred following BHA for femoral neck fractures. The specific purpose of this study was to compare the outcomes of surgical treatment of fractures with stem instability. Therefore, we limited the inclusion criteria to Vancouver B2 fractures. Patients with other fracture patterns (Vancouver A, B1, B3, C) or those who received conservative treatment were excluded from the analysis. This selection process resulted in a final study cohort of 59 patients (Fig. 1).

Close

Fig. 1 Patient selection flow diagram.

Export to PPT

3 Results

4 Functional outcomes

At the final follow-up examination, there were no significant differences in total Merle d’Aubigné score or its subcategories (pain, range of motion, walking ability) between the ORIF and Revision groups. The mean PMS was also similar (ORIF, 4; Revision, 4; p = 0.461) (Table 4). Given the sample size, the study had limited power to detect small-to-moderate between-group differences in functional scores. Assuming a clinically meaningful between-group difference of 0.5 SD (≈2.0 points), approximately 66 patients per group would be required to achieve 80% power at a two-sided α of 0.05; with the current sample size (n = 23 vs 36), the minimum detectable between-group difference at 80% power was approximately 3.1 points (see Supplement).

4.1

4.1 Survival analysis

At the one-year follow-up, the estimated mortality rate was approximately 30% in both the ORIF and revision arthroplasty groups. Kaplan–Meier analysis demonstrated no statistically significant difference in survival between the two groups (log-rank p = 0.79) (Fig. 2).

Close

Fig. 2 Kaplan–Meier survival curve comparing ORIF and Revision.

Export to PPT

5 Discussion

This study represents the first multicenter comparison of ORIF and revision arthroplasty specifically for Vancouver B2 PFFs in a frail, elderly BHA population. Our principal finding is that while ORIF was a significantly less invasive procedure (evidenced by shorter operative times and lower intraoperative blood loss) and yielded comparable outcomes in terms of one-year mortality, postoperative complications, reoperation rates, stem subsidence, RUST score, and functional status when compared to the more demanding procedure of revision arthroplasty. This suggests that for this vulnerable patient group, the choice of surgical procedure may have less impact on major clinical outcomes than the patient's underlying systemic condition.

ORIF preserves the femoral stem and requires more limited resection of bone and soft tissue, which generally makes it less invasive than revision arthroplasty. Accordingly, shorter operative time and lower blood loss would be expected. In our study, the mean operative time was shorter in the ORIF group, and the mean blood loss was 340 mL, less than half the blood loss observed in the revision group (752 mL). The rate of postoperative complications did not differ between groups. These findings are consistent with the results reported by Gonzalez-Martín et al., who analyzed 57 cases and demonstrated shorter operative time and lower blood loss in the ORIF group, with no differences in complications, reoperation rate, or one-year mortality.¹⁵ In our series, the duration of postoperative non–weight-bearing was 6.0 weeks in the ORIF group and 2.5 weeks in the revision group, and the difference was not statistically significant (P = 0.17), suggesting that postoperative weight-bearing after revision arthroplasty was individualized according to fracture severity and bone quality.

A key finding of this study was the absence of a significant difference in one-year mortality between the ORIF and revision arthroplasty groups. This observation is consistent with other reports, including our previous work,⁸ suggesting that in patient populations with a high comorbidity burden, survival is more closely associated with systemic factors than with the chosen surgical procedure.¹⁶ The overall one-year mortality rate was high at approximately 30% for both treatment groups, demonstrating the clinical severity of this injury. This rate also aligns with our prior findings in the larger cohort from which these patients were drawn.^8,17 Such high mortality is reflective of the baseline frailty of the study population, which was characterized by a mean age of 81 years and a substantial proportion of patients with comorbidities, as over 30% had a CCI of ≥2.^16,18

In terms of functional outcomes, no significant difference in postoperative ambulatory status was found between the ORIF and revision groups. Functional recovery across the cohort was limited, with only approximately 10% of patients regaining independent walking ability. While this outcome is consistent with some reports involving THA patients,¹⁹ it contrasts with a study by Takami and Takegami, which reported superior functional results with revision arthroplasty.²⁰ A critical distinction, however, is that their cohort included THA patients, whereas ours consisted exclusively of BHA recipients. The distinct difference in baseline functional demands and ambulatory potential between these two populations likely explains why neither surgical strategy demonstrated a clear functional advantage in our analysis.

Reoperation rates were comparable between the ORIF and revision groups, falling within the wide ranges previously reported for THA patients (4–33% for ORIF; 7–36% for revision arthroplasty).²¹ A specific characteristic of our cohort, however, was the relatively high proportion of patients treated with ORIF in comparison to other series. This treatment distribution may be attributable to two principal factors. First, the inherent frailty of the BHA population likely influenced surgeons to select a less invasive procedure, a strategy consistent with the recommendations by Pavone et al. for patients with low activity levels or high anesthetic risk.²² Second, the study's setting in tertiary trauma centers, where trauma surgeons often manage these injuries, may have contributed to this trend, as fixation techniques are often favored over revision arthroplasty by this specialty.²³

The findings of the present study support a patient-centered approach to this challenging fracture type. In our cohort, ORIF—typically a less invasive procedure—was associated with clinical outcomes comparable to those of revision arthroplasty, while requiring shorter operative time and resulting in less intraoperative blood loss. These results suggest that, in frail elderly patients with limited physiological reserve (often characterized by sarcopenia), an approach that limits surgical trauma may be appropriate in selected cases.²⁴ Accordingly, for Vancouver B2 fractures after BHA, ORIF may be considered one of the treatment options for carefully selected high-risk patients, provided that adequate fracture stability can be achieved. Ultimately, treatment decisions should be individualized, taking into account fracture characteristics and bone stock, patient comorbidities and functional demands, differences between BHA and THA populations, and surgeon judgment and expertise.

This study has several important limitations. First, the retrospective, non-randomized design introduces a risk of selection bias. Treatment allocation was determined at the surgeon's discretion, raising the possibility of confounding by indication; for example, frailer patients deemed too high-risk for revision arthroplasty may have been preferentially treated with the less invasive ORIF approach. Although baseline measured characteristics did not differ significantly between groups, unmeasured confounding may still have influenced the observed comparisons.

Second, the modest sample size (n = 59) limited statistical power. Because an established minimal clinically important difference (MCID) for the Merle d’Aubigné–Postel score is not available for frail elderly patients with periprosthetic femoral fractures, we used a distribution-based approach to contextualize the potential for type II error. We approximated the standard deviation (SD) from the interquartile range (IQR) assuming an approximately normal distribution (SD ≈ IQR/1.349). Based on the observed IQRs for the total Merle d’Aubigné–Postel score (ORIF: 12 ^9–15; Revision: 12.5 [9–14.25]), the SD was estimated at approximately 4.1 points. Using a clinically meaningful between-group difference of 0.5 SD (≈2.0 points) for sensitivity purposes, approximately 66 patients per group would be required to achieve 80% power at a two-sided α of 0.05. With the current sample size (n = 23 vs 36), the minimum detectable between-group difference at 80% power was approximately 3.1 points, indicating limited power to detect small-to-moderate functional differences. Accordingly, the absence of statistically significant differences should be interpreted cautiously and should not be taken to indicate equivalence between treatment strategies. Third, although all cases were classified as Vancouver B2 fractures, heterogeneity likely remained within this category. Treatment selection may have been influenced by factors not fully captured in the dataset, including the degree of stem loosening, fracture morphology, local bone quality, and surgeon judgment. These factors may have affected both procedure selection and postoperative outcomes. Fourth, the multicenter nature of the study introduces other potential confounders. There were no standardized protocols for surgical techniques, implant selection, or postoperative rehabilitation, which could have introduced variability and influenced outcomes. Functional outcomes were assessed only at the final follow-up, as standardized evaluations at fixed postoperative time points were not available; therefore, time-dependent factors, including aging, may have influenced the results. Similarly, we did not collect data on surgeon experience or case volume, and the retrospective data collection may have led to inconsistencies in the reporting of complications. Fifth, long-term outcomes and complications were not evaluated, preventing an analysis of late implant survivorship. The study's generalizability is also limited by its single-region cohort. Collectively, these limitations indicate that our findings should be confirmed in larger, prospective studies.

6 Conclusion

The results of our study indicate that, for Vancouver B2 PFFs after BHA, stem revision was associated with longer operative time and greater blood loss than ORIF. However, we observed no significant between-group differences in postoperative complications, reoperation rates, ambulatory function, or mortality. These findings suggest that ORIF, as a relatively less invasive procedure, may be considered one treatment option for selected frail elderly patients with limited physiological reserve, rather than as a universal substitute for revision arthroplasty. Treatment decisions should be individualized based on fracture characteristics, patient comorbidities and functional status, and surgeon judgment and expertise. Further prospective studies with larger cohorts and longer follow-up are warranted to validate these findings.

Guardian/patient's consent

The requirement for individual informed consent was waived, and an opt-out process was used.

Ethical statement

This study was approved by the Institutional Review Board of Nagoya University Graduate School of Medicine (IRB No. 2020-564). This study was conducted in accordance with the principles of the Declaration of Helsinki.

Funding

This research received no external funding.