The Cost and Consequences of Failed Osteosynthesis of Intertrochanteric Femur Fractures: A Matched Cohort Study

Abstract

Background

We aimed to evaluate the cost and consequences of failed osteosynthesis of intertrochanteric femur fracture (ITFF) patients and compare with primary ITFF patients.

Methods

We retrospectively evaluated 689 patients who underwent surgery due to ITFF via cephalomedullary nail. 31 patients (5.8%) had revision surgery because of osteosynthesis failure of ITFF. Each revision case included in the study was matched with four primary ITFF cases as control group based on age, gender, year of operation, type of fracture and American Society of Anesthesiologists (ASA) grade. Total cost for the admission that patients underwent surgery, mortality rate at first year, infection rate, length of stay at hospital, length of stay at intensive care unit, and erythrocyte transfusion amounts were recorded from hospital registry records. Tip apex distances (TAD) were noted.

Results

The mean total cost of the revision cases and primary cases was 10,027 ± 6387 and 5261 ± 1773 Turkish Liras, respectively (p < 0.001). TAD was ≥ 20 mm in 32.3% (10/31) of patients in revision group while 2.4% (3/124) of the patients in control group (p < 0.001). The mean length of stay at hospital, length of stay at intensive care unit, erythrocyte transfusion amounts, infection rate and mortality rate at first year were significantly higher in revision cases compared to matched primary control cases (p < 0.05).

Conclusion

Revision surgeries due to failed osteosynthesis of ITFFs with cephalomedullary nail have at least two times higher mean total cost than primary cases. The awareness of the cost, morbidity and mortality of the revision surgeries may reduce the modifiable risk factors of osteosynthesis failure including maintenance of TAD below 20 mm, obtaining optimal lag screw position and reduction quality.

Level of Evidence

Level 3, retrospective cohort study.

Introduction

Osteoporotic hip fracture in the elderly population is a common injury with a reported annual incidence of 1.6 million in worldwide [1, 2]. Approximately, 45% of these fractures are intertrochanteric femur fractures (ITFF) [1]. The optimal treatment method for the ITFF remains controversial [3]. Most of the ITFF could be treated with osteosynthesis utilizing dynamic hip screws or cephalomedullary nails. In the recent literature, osteosynthesis with cephalomedullary nail has been intensively used in the treatment of ITFF [4]. Failure of osteosynthesis of ITFF following treatment with cephalomedullary nail is an important concern for orthopedic surgeons since it could lead to severe morbidities for the patients [5].

Failure rate of osteosynthesis of ITFF has been reported 4.7–8% in the literature [5, 6]. Although the cost and complications of primary osteosynthesis of ITFF have been studied well [7], there are only a few studies concerning the failed cases [8, 9]. In this matched control study, we aimed to evaluate the cost and consequences of failed osteosynthesis of ITFF patients and compare with primary ITFF patients.

Materials and Methods

We retrospectively evaluated patients who underwent surgery between January 2010 and June 2019 in our clinic due to an ITFF after obtaining local ethics council approval (no: 43278876-929). A written informed consent was obtained from each patient. Patients with pathological fractures, periprosthetic fractures, initial osteosynthesis performed with an implant other than cephalomedullary nail, who had follow-up < 12 months and lost during follow-ups were excluded. Patients who were treated with cephalomedullary nail were identified. Patients who had ITFF and underwent revision surgery because of osteosynthesis failure following treatment with cephalomedullary nails were included in the study (Fig. 1).

Fig. 1.

Flowchart showing the patient population included

Total cost of the patients for the admission that patients underwent surgery (revision or primary surgery) was obtained from the local finance department of our hospital and given with Turkish Liras (TL). Patients' age, gender, American Society of Anesthesiologists (ASA) grade, mortality rate at first year, infection rate, length of stay at hospital, length of stay at intensive care unit, and erythrocyte transfusion amounts were recorded from hospital registry records. ITFFs of the patients were classified based on AO classification system [10]. Mode of failure of osteosynthesis, type of implants used during revision, complications were noted. Tip apex distance (TAD) was measured on early postoperative radiographs as previously described [11]. TAD ≥ 20 mm was considered as a risk factor for failure of osteosynthesis [12] (Fig. 2).

Fig. 2.

a Preoperative and b–c early postoperative radiographs of an 80-year-old female patient with AO type 2 intertrochanteric femur fracture treated with a cephalomedullary nail. Tip apex distance was above 20 mm. d The patient had failed osteosynthesis of intertrochanteric femur fracture at the first month of initial fixation due to a mechanical failure of the implant. e The patient underwent revision surgery with hemiarthroplasty and primary femoral stem

For the revision cases, the types of arthroplasty components were decided based on bone quality, bone loss of the patient, and surgeons' preference. Patients were followed at least 1 year. Absence of callus or bony bridge on at least three cortices at anteroposterior and lateral radiographs at 6 months follow-up was considered to be non-union [13].

We decided to compose a matched control group to make a valid comparison in homogenous groups. The primary outcome considered the total cost of the admission that patients had their surgeries. The priori power analysis revealed that control group requires four times more patients than the revision group (n = 31) to satisfy the 0.05 significance level and 80% statistical power. Thus, each revision case included in the study was randomly matched with four control cases (primary ITFF cases treated with cephalomedullary nail) based on age, gender, year of operation, fracture type, and ASA grade. Revision cases and matched primary control cases were compared using the aforementioned parameters.

Descriptive statistics were expressed as mean ± standard deviation for continuous numerical variables, categorical variables were expressed as number of patients and percentage. Distribution of variables was measured with the Kolmogorov–Smirnov test. Statistical analysis was performed with Mann–Whitney U test to compare mean values. Categorical variables were compared with Pearson Chi-square test. Analyses of the data were performed using the IBM SPSS Statistics 23.0 (IBM Corporation, Armonk, NY, USA) program. The results were considered statistically significant when the p value was < 0.05.

Results

A total of 689 patients with ITFF were evaluated. 527 ITFF patients met the inclusion criteria and were included in the study. There were 31 (5.8%) ITFF patients who had revision surgery due to failed osteosynthesis. Demographics of the patients regarding age, gender, year of operation, fracture classification, ASA grade were similar between revision (n = 31) and matched primary control cases (n = 124) (p > 0.05) (Table 1). 48% (n = 15) of the revisions were performed during the first postoperative month and 74% (n = 23) during the first 3 months.

Table 1.

Patient demographics of revision cases and matched primary control cases

	Revision cases (n = 31)	Matched primary control cases (n = 124)	p value
Age	83.4 ± 7.6	82.9 ± 7.4	0.858
Female gender (%)	71%	66.1%	0.608
ASA grade	2.9 ± 0.3	2.8 ± 0.6	0.275
Fracture type			0.756
A1	6 (19.4%)	30 (24.2%)
A2	18 (58.1%)	63 (50.8%)
A3	7 (22.6%)	31 (25.0%)

The mode of failure of osteosynthesis was mechanical failure of the implant (11 cut-outs, 16 cut-throughs) in 87% (27/31) of the patients and non-union in 13% (4/31) of the patients. All the mechanical failures were treated with arthroplasty. Two non-union patients were also treated with arthroplasty and the other two non-union patients were treated with autografting and refixation with cephalomedullary nails. Hemiarthroplasty was performed in 77.4% (n = 24) of the patients, total hip arthroplasty was performed in 16.1% (n = 5) of the patients. Primary femoral stems were used in ten patients, long femoral stems were used in 13 patients, calcar replacement femoral stems were used in 6 patients. 4 patients had an infection following revision with hemiarthroplasty. Three of the infections occurred in revisions that were performed within the first month of initial fixation. During the revision cases, two intraoperative femoral fissures were observed and treated with cerclage wires. 1 patient had dislocation of the hemiarthroplasty during follow-up and treated with revision total hip arthroplasty.

The mean total cost of the revision cases and primary cases was 10,027 ± 6387 TL and 5261 ± 1773 TL, respectively (p < 0.001). In the revision group, 32.3% (10/31) patients had TAD ≥ 20 mm while 2.4% (3/124) of the patients in control group (p < 0.001). The mean length of stay at hospital, length of stay at intensive care unit, erythrocyte transfusion amounts, infection rate, and mortality rate at first year were significantly higher in revision cases compared to matched primary control cases (p < 0.05) (Table 2).

Table 2.

Statistical analysis of mean total cost, mean length of stay at hospital, mean length of stay at intensive care unit, erythrocyte transfusion amounts, infection rate, mortality rate at first year, and tip apex distance ≥ 20 mm among revision and matched primary control cases

	Revision cases (n = 31)	Matched primary control cases (n = 124)	p value
Mean total cost (TL)	10,027 ± 6387 TL	5261 ± 1773 TL	< 0.001*
Mean length of stay at hospital (days)	17.6 ± 21.5	7.0 ± 3.7	< 0.001*
Mean length of stay at intensive care unit (days)	2.4 ± 1.3	1.5 ± 2.0	< 0.001*
Mean transfusion amount (units)	4.3 ± 3.4	1.4 ± 1.3	< 0.001*
Infection rate	4/31 (12.9%)	2/124 (1.6%)	0.004*
Mortality rate at first year	12/31 (38.7%)	22/124 (17.7%)	0.012*
Tip apex distance ≥ 20 mm	10/31 (32.3%)	3/124 (2.4%)	< 0.001*

TL Turkish Lira

*indicates statistical significance

Discussion

ITFFs have been an important health problem in elderly owing to increased life expectancy [1]. As the number of operated ITFFs increases, the failed osteosynthesis of ITFFs has been observed more frequently [14]. There are a few studies examining the impact of osteosynthesis failure on the health system as well as patients [8, 9]. We investigated the cost and impact of failed osteosynthesis of ITFFs in a matched control study and demonstrated that revision cases are significantly associated with higher cost, morbidity, and mortality rates.

Mechanical failure of the intramedullary nail is one of the most feared complications following internal fixation of ITFFs. Failure rates of osteosynthesis due to mechanical failure have been reported 4.7–8% in the literature [5, 6]. Risk factors of mechanical failure have been intensively investigated. TAD, position of lag screw or blade, and postoperative reduction quality were reported as the most important factors to prevent mechanical problems [5, 11]. These risk factors could be considered as modifiable risk factors for failure. In the current study, we observed 5.8% osteosynthesis failure following ITFF treatment with cephalomedullary nail and 87% of our failed osteosynthesis occurred because of mechanical failures. TAD of ≥ 20 mm was observed 32.3% (10/31) in revision group and 2.4% (3/124) in control group. 48% (n = 15) of the revisions were performed during the first postoperative month and 74% (n = 23) during the first 3 months. Considering the percentage of early mechanical failures and higher percentage of patients with TAD of ≥ 20 mm in the current study, we believe a remarkable percentage of mechanical failures could be prevented hence revision surgeries could be avoided by reducing the risk factors.

Arthroplasty has been commonly used as a salvage procedure for failed osteosynthesis of ITFF. The choice of hemiarthroplasty versus total hip arthroplasty is commonly made based on patients' comorbidity and functional status in addition to the arthritis grade in the hip joint at revision [15]. Various femoral stem types have been used based on bone quality and femoral canal geometry. Considering the poor bone quality of the patients, long stems or calcar-replacement stems are often required to achieve adequate bone fixation. Cho et al. reported that long stems or calcar-replacement femoral stems were used in 89% of the patients who were operated after failed internal fixation of ITFF [16]. Haidukewych et al. used long stems or calcar-replacement femoral stems in 51 of their 60 salvage arthroplasties for failed internal fixation of ITFF [17]. In our approach during revision cases, we do not remove the implants as the initial step of the surgery if the lag screw is within the femoral head and there are no broken implant parts. We slightly adjust the failed lag screw to its original position because subluxation of the hip joint is easier when fractured segments move in single piece. We consider that metaphysial fixation of femoral stems may not have adequate strength, we mostly prefer longer stems which pass the distal locking screw of the nail especially in osteoporotic patients. In the current study, 19 of 31 patients underwent arthroplasty utilizing long stems or calcar replacement femoral stems.

Arthroplasty could be an option for the primary treatment of ITFFs. It may reduce implant-related complications, reoperation rates, and immobilization period; however, it could increase intraoperative blood loss compared to cephalomedullary nailing [18]. In a recent meta-analysis, Ju et al. recommended hip arthroplasty in selected elderly patients with unstable ITFFs who have several risk factors of mechanical complications. Nevertheless, the authors stated that further high-quality randomized controlled trials are needed to determine the optimal treatment options for ITFFs [19]. We believe arthroplasty could be used in selected patients who are prone to complications related to prolonged immobilization and mechanical failure. More evidence is required for the utility of arthroplasty as a primary treatment option in ITFFs.

Revision of a failed osteosynthesis of ITFF is a challenging procedure. Some problematic situations could be encountered during the surgery including femoral and acetabular bone defects, osteopenic bone, removal of the nail or its broken parts, presence of a scar tissue due to previous surgery, and bone loss from previous fractures [20]. Mortazavi et al. evaluated both femoral neck fracture and ITFF conversions to arthroplasty and they stated that intramedullary nail conversions could be considered in the complex case group in which also there are cases required an extended trochanteric osteotomy [21]. Pui et al. reported 3 intraoperative femur fractures in 31 patients who underwent total hip arthroplasty after failed ITFF [22]. Lee et al. had 5 intraoperative femur fractures in 33 patients who underwent conversion to total hip arthroplasty following failed ITFF fixation [23]. In the current study, 2 of our 31 patients had intraoperative femur fractures and were treated with cerclage wiring.

Revision surgeries secondary to failed osteosynthesis of ITFFs commonly lead to high rates of morbidity and mortality. Tetsunaga et al. reported 6.3% infection rate following the conversions of failed ITFFs to total hip arthroplasty [24]. In a meta-analysis conducted by Luthringer et al. including failed internal fixation of ITFFs, 2 of 55 hemiarthroplasty and 3 of 71 total hip arthroplasty patients had infection after surgery [15]. In the current study, we had 12.9% (4/31) infection rate. As revision cases have high risk of infection, intraoperative and postoperative precautions should be taken to minimize the risk of infection. Thakar et al. stated that the strongest predictor of postoperative complications was the need for postoperative erythrocyte transfusion [8]. In the current study, primary cases had a mean of 1.4 ± 1.3 units transfusion and revision cases had a mean of 4.3 ± 3.4 units erythrocyte transfusion. Transfusion has been shown to cause a reduction in the patient's immune response [25]. This significant difference in transfusion amounts could have been contributed to the higher infection rate in revision cases. To reduce the erythrocyte transfusion rate following revision cases, careful coagulation control and utility of tranexamic acid during surgery would be beneficial.

Dislocation is not an infrequent problem following the revision of failed osteosynthesis of ITFF. Smith et al. reported 8.1% dislocation rate after the revision of failed osteosynthesis of ITFF with total hip arthroplasty [26]. Luthringer et al. compared hemiarthroplasty with total arthroplasty as a salvage procedure after failed osteosynthesis of ITFF regarding dislocation rate and they reported 10.4% dislocation rate after total hip arthroplasty and 5% after hemiarthroplasty [15]. We observed only one dislocation during follow-up which was occurred in a hemiarthroplasty patient.

Mortality is a major issue even for primary ITFF cases. In a systematic review containing 36 studies, the average 1-year mortality rate after hip fractures has been reported as 22% [27]. There are controversial studies in the literature reporting various mortality rates of revision cases compared to primary cases. Soreide and Lillestol reported that mortality rate was higher in the first 6 months in revision cases [28]; however, Palmer et al. noted no overall difference in mortality between primary and revision cases [29]. In the current study, 1-year mortality was significantly higher in revision cases (38.7%) compared to primary cases (17.7%). As the mortality rate of revision cases even higher, patients may benefit from close follow-ups by means of a multidisciplinary approach.

It is not hard to predict that revision cases would have higher costs than primary cases. However, less is known about how many times revision cases cost more compared to primary cases. Few studies investigated the economic burden of secondary surgeries due to failed osteosynthesis of ITFFs [8, 9]. Tiihonen et al. reported that the mean cost of primary cases was €7481 and revision cases was €9791 in Finland. Their mean length of stay at hospital was 9 days for primary cases and 16.2 days for revision cases [9]. Thakar et al. stated that mean cost of primary cases was £8610 (£918.54–£45 601.30) and the mean cost of secondary surgeries was £18 709 (£2606.30–£60.827.10) in the United Kingdom. Thakar et al. concluded that revision cases after failed osteosynthesis of ITFFs doubled the costs [8]. In the current study, the mean cost of primary cases was 5261 ± 1773 Turkish Liras and the mean cost of revision cases was 10,027 ± 6387 Turkish Liras. The mean length of stay at hospital and at intensive care unit was 7.0 ± 3.7 days and 1.5 ± 2.0 days for primary cases, 17.6 ± 21.5 days and 2.4 ± 1.3 days for revision cases, respectively. Our results demonstrated that revision cases double the mean total cost, mean length of stay at hospital and intensive care unit.

The strength of this study is the 4:1 matching of age, gender, ASA, type of fracture and year of the operation that allowed a valid comparison between the two groups. There are also some limitations of our study. The current study was a retrospective study thus had all of the drawbacks associated with that method. Further studies examining the functional status of primary and revision cases would be beneficial.

In conclusion, revision surgeries due to failed osteosynthesis of ITFFs with cephalomedullary nail at least double the mean total cost, the mean length of stay at hospital, the mean length of stay at intensive care unit, erythrocyte transfusion amounts, infection rate and mortality rate at the first year. The awareness of the cost, morbidity, and mortality of the revision surgeries may reduce the modifiable risk factors of osteosynthesis failure including maintenance of TAD below 20 mm, obtaining optimal lag screw position and reduction quality. In addition, arthroplasty might be an option for the primary treatment of patients with unstable ITFFs or patients with poor bone quality to eliminate the catastrophic consequences of osteosynthesis failure.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflicts of interest.

Ethical standard statement

This study was ethically approved by the Keçiören Health Practice and Research Center Council (date: 06/07/2020, documents number: 43278876-929).

Informed consent

A written informed consent was obtained from each patient.