Increased associated risk of postoperative periprosthetic femur fracture with cementless versus cemented hemiarthroplasty: a nationwide database study of 24,419 femoral neck fractures

Abstract

Objective:

We evaluated the associated risk of postoperative periprosthetic femur fracture after hemiarthroplasty for the treatment of femoral neck fracture based on stem fixation, cementless compared with cemented.

Design:

A retrospective review of patient data from the PearlDiver database from 2015 to 2020.

Setting:

This database is comprised of 140 million deidentified patients based on national all-claims data including Medicare, Medicaid, government and private insurance.

Patients/Participants:

We identified 24,468 patients with hemiarthroplasty for treatment of femoral neck fracture from 2015 to 2020. Demographic data included; age, sex, diagnosis of obesity, and Charlson Comorbidity Index.

Intervention:

We identified postoperative periprosthetic femur fracture as our primary outcome and stratified femoral fixation as cemented 12,777 (52.2%) or cementless 11,691 (47.8%).

Main Outcome Measurements:

Propensity matched analysis with odds ratios (OR) was performed to evaluate association between fixation and odds of postoperative periprosthetic femur fracture; multivariate analysis was used to adjust for demographic characteristics and comorbidities.

Results:

Cementless fixation was associated with increased odds of post-operative periprosthetic femur fracture (OR 3.32 [95% confidence interval (CI), 2.75–4.00]; P < 0.001). Females with cementless fixation had higher odds for post-operative periprosthetic femur fracture (OR 3.70 [95% CI, 2.94–4.76]; P < 0.001).

Conclusions:

We noted increased associated risk of postoperative periprosthetic femur fracture with cementless femoral fixation for hemiarthroplasty to treat femoral neck fracture, particularly in females. The risk of postoperative periprosthetic femur fracture should be considered as surgeons choose femoral fixation for hemiarthroplasty in the treatment of femoral neck fracture.

Level of Evidence:

Therapeutic Level III.

1. Introduction

Femoral neck fractures are common injuries seen in the United States and as the geriatric population continues to grow, the incidence of these injuries is expected to rise. According to projections, more than 500,000 geriatric femoral neck fractures are expected to occur in the United States annually by 2040.¹ Hemiarthroplasty is an established treatment option for management of displaced femoral neck fractures in the elderly.² Postoperative periprosthetic fracture is a rare, but severe complication following hemiarthroplasty for femoral neck fracture. In an already frail and vulnerable population, postoperative periprosthetic femur fracture after arthroplasty has been shown to add substantial additional morbidity, mortality and cost.^3–6

Broadly, fixation of the femoral component in hemiarthroplasty is either cemented or cementless. Proponents of cemented fixation list immediate fixation in osteoporotic bone, decreased reoperation rates, less pain, and potentially improved functional outcomes as potential benefits over cementless fixation.^7–11 Alternatively, advocates of cementless fixation cite shorter operative times, lower blood loss, and avoidance of bone cement implantation syndrome with no difference in complication rates reported in some studies.^12–15 Currently, the American Academy of Orthopaedic Surgeons Clinical Practice Guideline provides a moderately strong recommendation for preferential use of cemented over cementless femoral fixation in hemiarthroplasty for the treatment of femoral neck fractures.¹⁶

Though reported to be a rare complication, the significant additional morbidity, mortality, and cost associated with postoperative periprosthetic femur fracture after arthroplasty should give surgeons pause when considering implant fixation type. The risk of postoperative periprosthetic femur fracture in cemented and cementless hemiarthroplasty has been investigated previously—with conflicting results. While some investigators have found higher risks of intraoperative and postoperative periprosthetic femur fracture in cementless stems, these studies were limited in size or inadequately powered to evaluate a relatively uncommon complication.^8,12,14 Larger retrospective database studies performed internationally, where cement use is much more common, have identified higher rates of periprosthetic femur fracture with cementless fixation. However, the results may not be applicable to the United States experience where cementless fixation continues to be commonly used.^8,16 Only limited results evaluating the United States experience are available and are limited in generalizability because of the geographic location and source of data.⁹

To our knowledge, there has yet to be a large, nationwide investigation to evaluate the risk of periprosthetic femur fracture after hemiarthroplasty comparing cemented versus cementless fixation. The primary aim for this study was to evaluate the risk of periprosthetic femur fracture in cemented versus cementless femoral stems in patients treated with hemiarthroplasty after femoral neck fracture. Our secondary aim was to evaluate the association of patient demographic factors on risk for periprosthetic femur fracture.

2. Methods

A retrospective review was performed using the Medicare Standard Analytical Files from the PearlDiver database (PearlDiver Technologies, Inc, Fort Wayne, IN). This database is comprised of 140 million deidentified patients based on national all-claims data including Medicare, Medicaid, government and private insurance.

Using this database, we queried using the International Classification of Disease, 10th revision (ICD-10), Procedure Coding System for hemiarthroplasty within 14 days of a diagnosis code for femoral neck fracture in patients older than 65 year from 2015 to 2020. An additional query was performed using the diagnosis code for periprosthetic fracture around internal prosthetic hip joint. Only subjects with matching laterality in all diagnosis and procedure codes were included. Demographic data including age, sex, diagnosis of obesity, and Charlson Comorbidity Index (CCI) were also collected. The study was deemed exempt from Institutional Review Board and Animal Use Committee Review.

2.1. Statistical Analysis

Statistical analysis was performed using R (version 2021; R Core Team, Vienna, Austria). Differences in periprosthetic fracture rate based on continuous variables (age and CCI) and categorical variables (sex and obesity diagnosis) were analyzed using t-tests and χ² tests, respectively. Multivariate analysis with propensity matching was used to investigate the association between periprosthetic fracture after adjusting for demographic characteristics and comorbidities. Kaplan-Meier survival curves were created as a representation to compare fracture risk over time. A P-value of less than 0.05 was considered a statistically significant outcome.

3. Results

We identified 24,468 patients who received hemiarthroplasty for treatment of femoral neck fracture from 2015 to 2020. The overall mean age was 77 years (SD 6.14), 66.9% were female, mean CCI was 3.54 (SD 3.17), and 14% carried a diagnosis of obesity. There were 12,777 in the cemented group and 11,691 in the cementless group. Patients with cemented femoral fixation were older, mean age cement group was 77.05 years versus cementless 75.93 years (P < 0.001). Otherwise, there was no noted group differences in sex, CCI, or obesity (Table 1).

Table 1.

Demographics of cemented versus cementless hemiarthroplasty for femoral neck fractures.

	All, N = 24,468	Cemented fixation, 12,777 (52%)	Cementless fixation, 11,691 (48%)	P
Sex				P < 0.0001
Male	8140 (33%)	4029 (32%)	4111 (35%)
Female	16,328 (67%)	8748 (68%)	7580 (65%)
Age, years				P < 0.0001
Mean (SD)	76.67 (±6.14)	77.05 (±5.66)	75.93 (±6.57)
BMI				P < 0.0001
+Obesity	3374 (14%)	1641 (13%)	1733 (15%)
−Obesity	21,094 (86%)	11,136 (87%)	9958 (85%)
CCI				P = 0.01
Mean (SD)	3.54 (±3.17)	3.58 (±3.23)	3.48 (±3.11)

Out of the 24,468 patients in our cohort, we identified a total of 665 (2.72%) patients with a postoperative periprosthetic femur fracture. Periprosthetic femur fracture occurred in 479 (4.1%) cementless hemiarthroplasties and in 189 (1.46%) cemented. When propensity matching for the demographic factors collected (age, sex, obesity, and CCI), the odds of postoperative periprosthetic femur fracture in cementless is higher compared with cemented with an odds ratio (OR) of 3.32 (95% CI, 2.75–4.00 [P < 0.001]). For female sex, the odds of periprosthetic femur fracture with cementless compared with cemented hemiarthroplasty is even more pronounced, OR 3.70 ([95% CI] = 3.70 [2.94–4.76]).

Figure 1 shows the Kaplan-Meier survival curve for patients without a postoperative periprosthetic femur fracture with regard to time since surgery. For both cemented and cementless femoral fixation, most postoperative periprosthetic femur fracture occurred in the first 12 weeks after hemiarthroplasty. At all time points, cementless femoral fixation had a higher odds of postoperative periprosthetic femur fracture, P < 0.001.

Figure 1.

Kaplan-Meier survival curve free of periprosthetic femur fracture after hemiarthroplasty for treatment of femoral neck fracture with cement versus cementless fixation.

4. Discussion

Despite the American Association of Orthopaedic Surgeons (AAOS) clinical practice guideline recommending cemented fixation in hemiarthroplasty for treatment of femoral neck fracture, the use of cementless fixation in the United States persists, with 47.8% of hemiarthroplasties in this investigation using cementless fixation. In this study of 24,468 patients who received hemiarthroplasty after femoral neck fracture, we found cementless femoral fixation was associated with higher odds of postoperative periprosthetic femur fracture compared with cemented femoral fixation. This risk is further amplified in females and persists across all time points.

The increased odds of periprosthetic femur fracture with cementless fixation is consistent with the limited existing evidence specifically evaluating postoperative periprosthetic fracture in cemented and cementless hemiarthroplasty. Taylor et al⁸ performed a randomized controlled trial comparing cemented versus cementless hemiarthroplasty for displaced femoral neck fractures in the elderly. The authors found a higher rate of periprosthetic femur fracture in the cementless group 15% (12/80) compared with 1.3% (1/80) in the cemented group. However, this New Zealand study was limited by the relatively small size of 160 patients. In addition, this study potentially lacks generalizability to the United States experience as differences in familiarity with cementless or cemented fixation by surgeons in New Zealand is potentially different from the United States experience. Larger retrospective database studies comparing cemented and cementless hemiarthroplasty have noted similar findings. A National Health Systems database from the United Kingdom found a high risk of reoperation at 18 and 48 months in cementless versus cemented stems.⁷ Unfortunately, this study was limited because they compared reoperation in general and not periprosthetic femur fracture specifically. While we can imply reoperations could be due to periprosthetic femur fracture the data are limited to look specifically at fracture risk. Gjersten et al¹⁰ specifically evaluated postoperative periprosthetic fracture in a Norwegian investigation and found a significantly higher rate of periprosthetic fracture in cementless hemiarthroplasty. However, similar to the New Zealand investigation the generalizability of this study to the United States experience is not clear. Surgeons in the United States use cemented fixation technique less frequently compared with our foreign counterparts. For example, about 5% of primary total hip arthroplasty (THA) in the United States are with cemented femoral implants.¹⁷ This is compared with much higher rates of 35% to 50% in the United Kingdom, Sweden, and New Zealand.^18–20

Within the United States, there have been few studies investigating periprosthetic femur fracture in cemented versus cementless hemiarthroplasty. DeAngelis et al reviewed 130 patients who had undergone cemented and uncemented hemiarthroplasty and found no difference in the risk of periprosthetic femur fracture. However, with a smaller cohort, they were potentially underpowered to detect such a rare complication and as a single center experience the results may not be generalizable to other centers. Okike et al⁹ reviewed over 12,000 patients who underwent hemiarthroplasty for femoral neck fracture in the Kaiser Permanente Hip Fracture Registry. They found lower rates of periprosthetic fracture with cemented fixation. Our results support these findings and potential concerns of regarding the generalizability of the Kaiser registry due to it being a single integrated health care system in a specific regional geographic population should be tempered by our similar results in a more generalizable US cohort. Our results, along with the results of Okike et al, suggest that the risks in the United States are similar to those noted in other international studies.

As previously mentioned, the current AAOS Clinical Practice Guideline provides a moderately strong recommendation for preferential use of cemented over cementless femoral fixation for hemiarthroplasty.¹⁶ Despite this recommendation, slightly more than 55% of all hemiarthroplasty procedures for femoral neck fracture in the United States are with cementless fixation.¹⁷ In fact, according to the 2021 American Joint Replacement Registry Annual Report, only about half of patients older than 90 years receive a cemented hemiarthroplasty for femoral neck fracture. Although historically there was an increasing trend towards cementless femoral fixation in hemiarthroplasty for femoral neck fracture, this peaked in 2016 at 61.7% and has remained stable at the current rate over past 3 years.

This study has a number of potential limitations. The retrospective nature of this study has the inherent potential for selection bias. However, the comprehensive and diverse patient population of the PearlDiver database (encompassing all insurance claims data with the exception of Kaiser and TRICARE) helps to reduce selection bias and can provide some confidence in the generalizability of our results to the US population. Despite the size of this study dataset, the PearlDiver database is dependent on proper coding by practitioners to accurately identify and monitor outcomes. Therefore, an element of coding bias is potentially present due to the manual entry of diagnosis and procedural codes. We chose to only include ICD-10 codes in order to decrease any discrepancies with the previously implemented ICD-9 coding system. We attempted to minimize confounding effect of contributing variables by logistical regression, although it is possible that all relevant variables were not identified or controlled for in our analysis. Furthermore, we were unable to perform subgroup analysis of implant type given the limited coding capabilities within the dataset; thus, we were unable to compare periprosthetic femur fracture risk with various classifications of cementless femoral stems. We were unable to obtain preoperative clinical frailty scale or physical activity scale for the elderly scores in the PearlDiver database and future investigations may include these to expand upon our results.

5. Conclusion

Among patients who received hemiarthroplasty for treatment of a femoral neck fracture, we found increased odds of postoperative periprosthetic femur fracture with cementless femoral fixation. These findings should be considered as surgeons consider cemented versus cementless fixation for hemiarthroplasty. The elevated odds of periprosthetic femur fracture with cementless fixation was amplified in females suggesting further scrutiny of femoral fixation strategy in this population.