As the population ages, hip fractures and associated morbidity and mortality are becoming more common1. Studies have shown that, within one year after a hip fracture, only 50% to 65% of patients regain their pre-fracture walking status and as many as 20% are unable to walk2,3. Mortality rates one year after a hip fracture have ranged from 14% to 36% among the elderly1,2.
Boston reported that a history of hip fracture is a notable risk factor for a second hip fracture4. Therefore, as the total number of patients sustaining hip fractures increases and perioperative management improves, more patients will be living with a surgically treated hip fracture and will subsequently produce a possible increase in the number of second hip fractures5,6. Between 1.7% and 11.8% of patients with an initial hip fracture will sustain a subsequent contralateral hip fracture4-11. These subsequent fractures are associated with an increase in mortality and morbidity compared with an initial hip fracture4,6,10. According to Boston, patients sustaining sequential hip fractures experienced a 17% increase in three-month mortality4. Similarly, Shabat et al. reported that 27% fewer patients with sequential bilateral hip fractures were able to return to their initial mobility status following rehabilitation, compared with patients with only one hip fracture9. Although the rate of subsequent contralateral hip fractures has remained fairly constant recently7, a second hip fracture may become more common as the population ages and the prevalence of osteoporosis increases9. Therefore, identifying risk factors for a subsequent hip fracture provides valuable information to implement interventions to decrease mortality and morbidity following hip fracture.
Associations between several different risk factors and bilateral hip fractures have been reported5,6,9, including comorbidities, institutionalization, biochemical changes associated with osteomalacia, and surgical technique5,6,9. While up to 11.8% of patients undergoing surgical treatment for an initial hip fracture sustain a subsequent contralateral hip fracture7, it is not known whether this rate is similar in patients managed with different surgical procedures. Although it has been suggested as a possible associated factor, we are aware of no study that has evaluated the initial surgical procedure that was used to treat the fracture and the related likelihood of a subsequent contralateral hip fracture. More specifically, a review of the PubMed database revealed no articles specifically addressing whether closed reduction and percutaneous pinning is associated with a difference in the relative risk of a contralateral hip fracture compared with initial arthroplasty treatment. Of all possible factors associated with the subsequent rate of additional fracture, the chosen surgical procedure is the most controllable variable available to the operating surgeon.
Generally, femoral neck fractures are treated with either closed reduction and percutaneous pinning or arthroplasty12. Closed reduction and percutaneous pinning is less invasive and less time-consuming, is used more frequently for nondisplaced and impacted fractures, and usually is associated with decreased blood loss and a shorter hospital stay12,13. Closed reduction and percutaneous pinning for the treatment of displaced fractures has been associated with a higher rate of revision procedures, resulting in an overall inferior cost-effectiveness14-18. In comparison, arthroplasty (hemiarthroplasty and total hip arthroplasty) has been associated with better initial postoperative skeletal stability and lower rates of failure, resulting in a decreased likelihood of revision procedures12,19,20. Arthroplasty also results in increased surgical morbidity and a trend toward increased early mortality18,21-23. Although the outcomes of these two treatment methods have been frequently reported, we are not aware of any study that has investigated the relationship of either technique to the risk of a contralateral hip fracture.
The primary purpose of the present study was to compare the rates of subsequent contralateral hip fractures in patients initially managed with closed reduction and percutaneous pinning as compared with the rates in patients initially managed with arthroplasty.
The present study is a retrospective, comparative chart review of a cohort of patients within a single health-care system.
Patients
Prior to initiation, the present study was approved by our institutional review board. All patients who underwent treatment of a femoral neck fracture with closed reduction and percutaneous pinning or arthroplasty between 1999 and 2008 within a single health-care system, which included a Level-I trauma center, were identified. Patients who had malignant disease, those who were younger than fifty years of age, and those who had sustained the fracture as the result of high-energy trauma were excluded from the present study. High-energy trauma was defined as an injury resulting from a mechanism other than a fall from a standing or sitting height. Patients also were excluded if they had had an initial procedure involving the proximal femoral region other than arthroplasty or closed reduction and percutaneous pinning for the treatment of a femoral neck fracture. For the present study, 1238 patient records were reviewed.
Intervention
All patients were managed by an orthopaedic senior staff surgeon within the same health-care system. For each patient, treatment was based on the fracture pattern identified on radiographs and the experience of the treating surgeon. The treatment algorithm utilized was based on the classification of stability of femoral neck fractures as described by Garden24 in 1964. The displacement was evaluated on the lateral radiograph, and the fractures were classified as either stable or unstable. Garden Type-I and II fractures were considered stable and were treated with subsequent closed reduction and percutaneous pinning. Partially displaced Garden Type-III fractures that were reduced easily also were considered stable and were treated accordingly. Displaced Garden Type-III and IV femoral neck fractures were regarded as unstable and were treated with arthroplasty.
The closed reduction and percutaneous pinning group included patients who were managed with closed reduction and percutaneous fixation of a femoral neck fracture with two to four percutaneous 6.5-mm cannulated screws that were placed under fluoroscopic guidance. Patients who initially underwent closed reduction and percutaneous pinning with subsequent failure and revision to an arthroplasty, a Girdlestone procedure, or other treatments were included in the closed reduction and percutaneous pinning group. The arthroplasty group consisted of patients who were managed with a primary unipolar or bipolar hemiarthroplasty or total hip arthroplasty for the initial treatment of a femoral neck fracture.
While in the hospital, patients were allowed weight-bearing and unrestricted hip motion as tolerated postoperatively and were evaluated and managed by a physical therapist. All patients either were discharged to nursing homes, skilled nursing facilities, or rehabilitation centers with continuation of physical therapy or were discharged to home with home health physical therapy. Occupational therapy was provided on an individual basis according to the observed need while the patient was in the hospital.
Outcome Measures
Data were retrieved for all patients for whom the CPT (Current Procedural Terminology) code was 27235 or 27236. Electronic medical records and digital radiographs were used to determine all variable outcomes. For the primary outcome of the contralateral fracture rate, the procedures performed for the treatment of the initial femoral neck fracture and the presence of a subsequent contralateral proximal femoral fracture were recorded. A proximal femoral fracture was defined as a femoral neck, intertrochanteric, or subtrochanteric fracture. Secondary variables included the time that had elapsed between the initial and subsequent fractures, age, sex, bisphosphonate use, history of diabetes, and smoking history. Secondary variables were documented as follows. The diagnosis of diabetes included patients with any history of type-1, type-2, and/or diet-controlled diabetes. Bisphosphonate use was noted as “yes” or “no” with no distinction of the type of drug utilized. Smoking history was classified according to five different grades: Grade 1 represented patients who had never smoked or had smoked less than ten pack-years, Grade 2 represented former smokers of ten to fifty pack-years, Grade 3 represented former smokers of greater than fifty pack-years, Grade 4 represented current smokers of less than one pack per day, and Grade 5 represented current smokers of greater than one pack per day. A meaningful smoking history was defined as Grade 2 or higher. Age and sex were recorded directly. No other comorbidities were included in the data set.
Statistical Analysis
All demographic characteristics of the patients were summarized according to the surgical procedure group (closed reduction and percutaneous pinning compared with arthroplasty) and the fracture group (fracture compared with no fracture) with use of descriptive statistics; the mean and the standard deviation were used for continuous variables, and the frequency (percentage) was used for categorical variables. The two procedure groups were compared with regard to each characteristic with use of a two-sample t test for continuous variables and a chi-square test or Fisher exact test for categorical variables. Univariate logistic regression analyses were performed to assess the probability of having a subsequent fracture in association with each of the patient characteristics. A reduced multivariable logistic regression model that included all significant variables was presented, and the Hosmer-Lemeshow goodness-of-fit test was performed. The propensity score, defined as the predicted probability of treatment with closed reduction and percutaneous pinning compared with arthroplasty, was estimated with use of the multivariable logistic regression model that adjusted for age, sex, smoking status, diabetes, and bisphosphonate use. A multivariate logistic regression model adjusting the propensity score as a covariate was utilized to compare the two surgical procedures in terms of the probability of a subsequent fracture. A p value of <0.05 indicated significance. SAS version 9.2 (SAS Institute, Cary, North Carolina) was used for data management and statistical analysis.
Source of Funding
There were no external sources of funding for this retrospective review.
Of the 1238 patients who were reviewed, a total of 1177 met the inclusion criteria. The study population included 906 female patients and 271 male patients with an overall average age of 79.96 years (range, fifty to 102 years). Fifty-two patients were excluded because they were less than fifty years old or had sustained the injury as the result of a high-energy mechanism, and nine patients were excluded because of an associated primary or metastatic tumor.
Four hundred and ninety-five patients (42%) underwent initial closed reduction and percutaneous pinning, and 682 (58%) underwent arthroplasty (Table I). The closed reduction and percutaneous pinning group included 364 female patients (73.54%) and 131 male patients (26.46%) with an average age (and standard deviation) of 77.98 ± 10.51 years. The arthroplasty group included 542 female patients (79.47%) and 140 male patients (20.53%) with an average age of 81.40 ± 9.10 years.
The contralateral fracture rate was 10.10% (n = 50) in the closed reduction and percutaneous pinning group and 5.57% (n = 38) in the arthroplasty group; this difference was significant (p = 0.0035). The odds ratio (OR) in the closed reduction and percutaneous pinning group was about two times that in the arthroplasty group both on univariate analysis (OR = 1.904, 95% confidence interval [CI] = 1.228 to 2.953; p = 0.0040) (Table II) and in the reduced multivariate model (OR = 2.106, 95% CI = 1.351 to 3.282; p = 0.0010) (Table III). The propensity score was estimated as the predicted probability of treatment with closed reduction and percutaneous pinning as compared with arthroplasty. A multivariable logistic regression model adjusting for propensity score as a covariate, in addition to age, sex, smoking history, presence of diabetes, and bisphosphonate use, was then utilized to compare the two surgical procedures with regard to the probability of having a subsequent fracture. Propensity score analysis also gave a similar result when the closed reduction and percutaneous pinning group was compared with the arthroplasty group (OR = 2.021, 95% CI = 1.298 to 3.147; p = 0.0018) (Table III).
Significant differences were demonstrated between the two procedure groups in terms of age (p < 0.0001), sex (p = 0.0169), and smoking history (p = 0.0161). The arthroplasty group was older, included more female patients, and had more patients in the Grade-1 smoking category as compared with the closed reduction and percutaneous pinning group. There was no difference between the groups in terms of the percentage of patients who had diabetes or who used bisphosphonates (Table I).
Among patients who sustained a contralateral fracture, there was no difference between the procedure groups in terms of age, smoking grade, diabetes, or bisphosphonate use. There was a significantly larger percentage of male patients with contralateral fractures in the closed reduction and percutaneous pinning group as compared with the arthroplasty group (26% compared with 7.89%; p = 0.0292). The average duration of time from the initial fracture to the contralateral fracture was similar in both the closed reduction and percutaneous pinning and the arthroplasty groups (24.91 compared with 21.17 months) (Table IV). The interval between fractures in both groups ranged from one week to eighty-five months.
In the total study population, the rate of contralateral fracture was 7.5%. No significant difference was found between sex or smoking grade and the presence of a contralateral fracture. However, surgical procedure, age, presence of diabetes, and bisphosphonate use were found to be significantly different in patients who sustained a contralateral fracture as compared with those who did not (Table II). Seventy-six (86%) of the eighty-eight patients with a subsequent contralateral hip fracture had a femoral neck fracture.
To our knowledge, no studies have evaluated specific methods of hip fracture treatment and the associated risk of a subsequent contralateral hip fracture. The purpose of the present study was to determine if the rate of a subsequent contralateral hip fracture after femoral neck fracture was different between patients undergoing initial treatment with closed reduction and percutaneous pinning and those undergoing arthroplasty. The unique design of the present study allowed the focus to be on an isolated, and frequent, type of hip fracture in patients managed with two predominant and well-accepted methods of treatment. Although the high revision rate associated with closed reduction and percutaneous pinning for the treatment of displaced hip fractures has been well documented12,13,18, the rate of subsequent contralateral fractures following closed reduction and percutaneous pinning as compared with arthroplasty has not been evaluated previously, to our knowledge.
The present study indicated that the risk of a contralateral hip fracture is significantly higher after the treatment of an initial femoral neck fracture with closed reduction and percutaneous pinning as compared with arthroplasty. It is noteworthy that when the study population was limited to patients who sustained only femoral neck fractures as the initial injury, the overall rate of contralateral fractures was 7.5%. This rate was consistent with the findings of previously published studies in which the rate of contralateral hip fracture was 1.7% to 11.8%4-11. The rate of similarities between the initial and subsequent hip fractures also was consistent with previous reports. Several studies have demonstrated that patients with a second hip fracture will tend to have a similar type of fracture as the initial injury4-6,9,10. In the present study, both treatment groups had a strong tendency for the second fracture to occur once again within the femoral neck region. In the present cohort, 86% of the subsequent contralateral proximal femoral fractures were femoral neck fractures. This predisposition was postulated by Ferris et al. to be due to longer femoral necks in patients with subcapital fractures as compared with patients with intertrochanteric fractures25. Shabat et al. theorized that this tendency was possibly due to each individual's unique gait pattern or bone architecture or a combination of both9. In the present study, both treatment groups experienced a similar average length of time until the second fracture (24.91 months in the closed reduction and percutaneous pinning group, compared with 21.17 months in the arthroplasty group). Previously reported intervals until subsequent fracture have ranged from 23.9 months to seven years5,6,9-11.
Overall, patients who incurred a subsequent hip fracture were older than those who did not, with eighty-six of eighty-eight patients having an age of more than sixty-five years, producing an OR of 4.588 (95% CI = 1.113 to 18.911). This finding differs from previously published data that showed no significant difference in age between patients with a first hip fracture and those with a subsequent fracture5,9. Despite the association of advanced age with increased risk, the closed reduction and percutaneous pinning group consisted of a younger population while still incurring a higher rate of subsequent fracture, suggesting that factors other than age were responsible for the increased risk in this group. Neurological conditions and dementia, which can be related to an increased risk of falling, were not included as variables. There was no significant difference between the sexes with regard to the rate of subsequent fracture. This finding is consistent with previously published data5,9.
Diabetes is associated with decreased bone mineral density26. In the present study, diabetes was also a significant covariate when comparing patients with and without a subsequent fracture. Diabetic patients were found to have a significantly increased risk of sustaining a contralateral hip fracture (OR = 1.725, 95% CI = 1.090 to 2.730), which was a plausible finding as diabetes has been linked with decreased bone mineral density in the hip27. The fact that both procedure groups included a similar percentage of diabetic patients implies that a variable other than diabetes was responsible for the difference in the contralateral fracture rate.
The data revealed that patients using bisphosphonates had an elevated risk of a contralateral fracture compared with patients not using bisphosphonates (OR = 1.724, 95% CI = 1.008 to 2.947; p = 0.0466). The true meaning of this finding is unknown. However, because there was no significant difference between the procedure groups in terms of the number of patients taking bisphosphonates, bisphosphonate use does not appear to be influential in the increased fracture rate found in the closed reduction and percutaneous pinning group.
Cumulatively, these data showed that patients who were older than sixty-five years of age, those who were diabetic, those who were managed with closed reduction and percutaneous pinning, and those who were taking bisphosphonates had an increased risk of sustaining a contralateral hip fracture after an initial femoral neck fracture. The presence of diabetes, age, and many other patient-related variables are not controllable. However, the surgical intervention performed is one of the few variables over which the surgeon has direct control. As the present study is the only one, to our knowledge, in which the rate of contralateral fracture in patients who have been managed with closed reduction and percutaneous pinning is compared with that in patients who have been managed with arthroplasty, additional investigations are needed to confirm or deny these results.
Several studies have compared the cost of closed reduction and percutaneous pinning with that of arthroplasty for the treatment of femoral neck fractures. Although closed reduction and percutaneous pinning is less costly initially because of decreased implant cost, it is less cost-effective for the treatment of displaced femoral neck fractures after accounting for the cost of revision procedures14-17,21. Those studies included costs associated with the initial hospitalization and revision surgical procedures but failed to include costs related to any subsequent hip fractures. The total cost of treatment can be as much as doubled when secondary surgical procedures are required28. The high rate of contralateral fracture after closed reduction and percutaneous pinning in these patients suggests that these outcomes underestimate the true total associated cost of closed reduction and percutaneous pinning. Further studies accounting for costs associated with subsequent hip fractures are warranted to better understand the true total cost of each of these procedures.
The limitations of the present study include its retrospective nature, which limited the data to those that were retrievable from historical documentation. Also, the population studied consisted of patients with both displaced and nondisplaced fractures. There is widespread consensus that nondisplaced or impacted fractures need to be treated with percutaneous pinning without a reduction attempt, and the inclusion of patients with these types of fractures is a major confounder. Surgeon bias concerning radiographic interpretation is inevitably present in clinical practice and therefore affects the selection of the resultant surgical procedure. However, in the present study, this bias was dampened by the large number of physicians responsible for treating fractures, the exclusion criteria applied, and the high volume of patients. Another limitation was the lack of identification of potentially meaningful associated patient-related factors, including the cause of the original fall that led to the fracture. Walking status, mental status, institutionalization, the presence of osteoporosis, and associated comorbidities affect the outcome of different treatment methods5,6. We recommend that future studies investigate the role of these factors in sustaining contralateral proximal femoral fractures.
Furthermore, these data also question what role, if any, that surgical intervention has in the increased rate of contralateral fractures following closed reduction and percutaneous pinning. The present study allows only for suggestions of the responsible factor or factors. It seems logical that the shortening at the fracture site that often occurs with closed reduction and percutaneous pinning has the potential to permanently alter the gait of the individual, which could lead to an increased risk of subsequent falls. Zlowodzki et al. showed that approximately 30% of femoral neck fractures that were treated with closed reduction and percutaneous pinning underwent shortening of >3 mm29. This shortening alters the moment arm of the hip abductors and results in an increase in the joint reactive force as well as an increase in the abductor force required for walking. Zlowodzki et al. additionally revealed that patients with shortened fractures also had decreased Physical Functioning and Role Physical subscores on the Short Form-36 (SF-36)29. There remains the possibility of additional factors, in isolation or in combination, that potentially contribute to the increased rate of contralateral hip fractures. The similarities in time elapsed from initial fixation to contralateral fractures between the two treatment groups seem to eliminate the suggestion of differences in rehabilitation as a likely contributor to the disparity. In addition, differences between the demographic characteristics of the two populations that lead to an increase were not identified in the present study. Critiques of the results of each of the surgical procedures were beyond the scope of the present study. As the number of hip fractures increases and the treatment of hip fractures improves, more patients will be living with surgically treated hip fractures leading to a subsequent increase in the possibility of contralateral hip fractures. The prevention of a contralateral hip fracture is fundamental in appropriately managing elderly patients after an initial fracture, and identification of associated risk factors is paramount to prevention. The present study demonstrated a higher risk of sustaining a contralateral hip fracture in patients who underwent closed reduction and percutaneous pinning for initial fracture treatment as compared with those who underwent arthroplasty. Unlike other variables, the method of surgical intervention can be controlled by the surgeon and may be instrumental in preventing future contralateral hip fractures and calculating overall costs associated with a particular surgical procedure.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.