Unipolar or Bipolar Hemiarthroplasty for Femoral Neck Fractures in the Elderly?

Abstract

This prospective randomized trial compared the efficacy of unipolar versus bipolar hemiarthroplasty in elderly patients (≥ 65 years) with displaced femoral neck fractures in terms of quality of life and functional outcomes. One hundred fifteen patients with a mean age of 82.1 years were enrolled in this study and randomized to either unipolar or bipolar hemiarthroplasty. Quality of life and functional outcomes were assessed using the Musculoskeletal Functional Assessment instrument and Short Form-36 health survey. Seventy-eight patients completed 1 year of followup. There were no differences between the groups in estimated blood loss, length of hospital stay, mortality rate, number of dislocations, postoperative complications, or ambulatory status at 1 year. There also were no significant differences between the two groups at either point in postoperative Short Form-36 or Musculoskeletal Functional Assessment instrument scores. Results of this prospective randomized study suggest that the bipolar endoprosthesis provides no advantage in the treatment of displaced femoral neck fractures in elderly patients regarding quality of life and functional outcomes.

Femoral neck fractures are among the most common orthopaedic injuries in patients older than 65 years. According to the American Academy of Orthopaedic Surgeons, there are 350,000 hospital admissions each year resulting from hip fractures, with that number expected to almost double by the year 2050. ³ Elderly patients who sustain a hip fracture have a mortality rate at 1 year ranging from 14% to 36%, ^{1,2,12,22,25,35} with 50% of those who survive losing their ability to walk independently. ³ As the geriatric population increases, treatment of femoral neck fractures will continue to consume a significant portion of the nation’s healthcare resources.

Optimal treatment of femoral neck fractures has been controversial for years. ^15,23,28,30 One of the areas of debate is in the use of unipolar (fixed-head) or bipolar (universal) endoprostheses for treatment of displaced (Garden Types III–IV) ¹⁸ femoral neck fractures in elderly patients. The bipolar prosthesis has a theoretical advantage because it was designed to move at its inner bearing in addition to articulating at the prosthesis-acetabulum interface. ^4,20 This should decrease the amount of acetabular erosion and reduce pain. However, several studies have shown that the inner bearing loses mobility with time and becomes stiff, thereby minimizing the advantage of the bipolar prosthesis. ^10,15,16,31 Furthermore, the increased cost of the bipolar prosthesis compared with the unipolar prosthesis raises the question whether the difference in cost translates to better functional outcomes and quality of life.

The purpose of the current study was to compare the efficacy of unipolar and bipolar hemiarthroplasties in elderly patients (≥ 65 years) with displaced femoral neck fractures in terms of quality of life and functional outcomes. Other studies have evaluated the unipolar versus bipolar debate and either reached differing conclusions or did not have ideal study design. Earlier studies were not randomized and retrospective, introducing selection bias into the groups. ^29,36 More recent studies have improved designs and have addressed functional outcomes, but without looking at overall quality of life issues. ^7,11,14 The current study used well-validated outcome questionnaires to assess the quality of life and physical function of patients at 1-year followup. By examining the results of these questionnaires, a more complete picture of the patients’ well-being was gained.

MATERIALS AND METHODS

From May 1997 to January 2000, patients 65 years or older who presented to the authors’ institution with an acute, displaced femoral neck fracture (Garden Types III–IV) were recruited for this randomized, prospective study. The study was approved by the institutional review board and all patients gave informed consent. Exclusion criteria included patients with dementia, patients who were nonambulatory, patients with pathologic femoral neck fractures, patients with additional acute lower extremity fractures in addition to the femoral neck fracture, and patients residing in nursing homes. One hundred fifteen consecutive patients who met these criteria agreed to be randomized to either a unipolar or a bipolar prosthesis.

Prefracture data were collected by various members of the research team trained in administering questionnaires via patient interviews done in person immediately after admission. Hospital data were obtained with a form filled out in the operating room by the orthopaedic resident or attending physician and a form completed on discharge relating specific details of the patient’s hospital stay. At 3 months and at 1 year after hemiarthroplasty, the patients were contacted either by telephone or in the doctor’s private offices and interviewed for completion of two standard questionnaires.

Patients were assigned to either a Unitrax unipolar or a Centrax bipolar (Howmedica, Rutherford, NJ) prosthesis at the time of surgery by computerized random number generation. Patients were blinded to the type of prosthesis they received. The appropriate-sized cemented Premise stem (Howmedica) was used in each case. All patients were administered heparin preoperatively and taken to the operating room within 24 to 48 hours after admission and medical clearance. Anesthesia type was selected by the anesthesiologist; however, regional anesthesia was used in the majority of patients. A posterolateral approach to the hip was used. Perioperative antibiotic prophylaxis was used and antibiotics were continued for 48 hours postoperatively. A warfarin sliding scale was used for postoperative thromboprophylaxis for 6 weeks. Patients were mobilized to weightbearing as tolerated on postoperative Day 1 with supervision by physical therapists.

On admission, patient characteristics that were examined included age, gender, and ambulatory status before fracture. Patients were classified as community ambulators (with or without assistive devices) or household ambulators (if they walked indoors only). The overall health status of the patients was determined by the number of medical comorbidities at the time of the operation. The comorbidities of each patient were noted and an index of comorbidities for each patient was calculated based on the method of Charlson et al. ^8,9 The Charlson comorbidity index was designed to classify prognostic comorbidity and assigns higher scores to diseases having an increased relative risk of mortality.

Hospital and postoperative data that were analyzed included estimated blood loss, length of hospital stay, and number of blood transfusions required. Operative complications and minor and major postoperative medical complications were recorded. Minor complications included urinary tract infection or retention, hypoxic episode, pneumonia, wound hematoma, fever, and altered mental status. Major complications included pulmonary embolus, deep infection, atrial fibrillation, decubitus ulcer requiring skin graft, cerebrovascular accident, and reoperation. Dislocations up to 1 year postoperatively also were noted.

Functional outcome was assessed at followup by use of the Musculoskeletal Functional Assessment instrument. ¹⁷ The Musculoskeletal Functional Assessment instrument is a well-validated health status instrument that has 100 self-reported health items. A raw score was calculated for each patient using the entire index, and subset scores for mobility and activities of daily living also were determined. For the Musculoskeletal Functional Assessment instrument, a lower score indicates better function. Additional functional outcome data used postoperative ambulatory status (community or household) as defined previously.

Quality of life was assessed using the Medical Outcomes Study 36-item short-form health survey (Short Form-36). ³² The Short Form-36 was designed for use in clinical practice and research. It is a multiitem scale that assesses eight health concepts: (1) limitations in physical activities because of health problems; (2) bodily pain; (3) limitations in usual role activities because of physical health problems; (4) limitations in usual role activities because of emotional problems; (5) general mental health (psychologic distress and well-being); (6) limitations in social activities because of physical or emotional problems; (7) vitality (energy and fatigue); and (8) general health perceptions. It can be self-administered or administered by a trained interviewer in person or by telephone.

Data were analyzed using a chi square test for dichotomized measures or Student’s t test for continuous numeric variables. A p value less than 0.05 was considered significant.

RESULTS

One hundred fifteen patients were enrolled in the study: 60 patients received unipolar prostheses and 55 patients received bipolar prostheses. Ninety-three patients completed 3 months followup: 50 from the unipolar group and 43 from the bipolar group. Twenty-four patients were known to have died before the 1-year followup (12 patients from each group). Of the 91 patients alive at 1 year, 13 patients could not be reached or declined to answer the followup questionnaires, leaving 78 patients who completed the study: 40 in the unipolar group and 38 in the bipolar group. There were no statistically significant differences between the groups in terms of age, gender, medical comorbidities, or ambulatory status before fracture (Table 1).

TABLE 1: Prefracture Characteristics of the Study Groups

Hospital and postoperative data showed no significant differences between the groups with respect to estimated blood loss, number of blood transfusions required, length of stay on the orthopaedic service, or postoperative complications (Table 2). Two patients died while in the hospital and are included among the 24 deaths mentioned previously. One patient who received a unipolar prosthesis had a deep infection of the operative site, requiring removal of the prosthesis after three attempts at incision and drainage of the site. This patient died before 1 year of followup of causes not directly related to the hip fracture. There were two dislocations before the 1-year followup: one in each group. The patient with the bipolar prosthesis with a dislocation had successful closed reduction. A patient with a unipolar prosthesis had three dislocations reduced before an adductor release was done to correct an underlying spastic palsy.

TABLE 2: Postoperative Data

Of the 31 patients with unipolar prostheses who were community ambulators before fracture, 23 (74.2%) remained community ambulators at 1-year followup. For the patients with bipolar prostheses, 21 of 29 patients (72.4%) retained their prefracture status of community ambulators. The difference between these groups was not significant. The percentages at 1 year showed a slight, nonsignificant increase from 3-month followup data, where 71.4% of patients with unipolar prostheses who were community ambulators and 65.8% of the patients with bipolar prostheses who were community ambulators maintained their ambulatory status.

Average Short Form-36 scores were calculated for each group at baseline, 3 months followup, and 1 year followup to assess the patients’ perceptions of quality of life. The different points within each group were compared and no statistically significant differences were found regarding change with time for any of the eight categories (physical function, pain, physical role, emotional role, mental health, social functioning, vitality, and general health). The unipolar and bipolar groups also were compared at each specific point, and no statistically significant differences were detected. The scores for the unipolar and bipolar groups at each point for the eight Short Form-36 categories are shown in Table 3.

TABLE 3: Average Short Form-36 Scores at Each Point

Musculoskeletal Functional Assessment instrument scores were calculated for patients at 1 year of followup. The average raw scores, mobility subset, and self-care subset scores are shown in Table 4. There was no significant difference between the groups at 1-year followup. Three-month followup data (not shown) also failed to show significance between the unipolar and bipolar groups, or significance between points.

TABLE 4: Average Musculoskeletal Functional Assessment Instrument Scores at 1 Year

DISCUSSION

The goal of the current study was to compare functional and quality of life scores in patients randomized to receive either unipolar or bipolar hemiarthroplasty and determine whether bipolar prostheses would result in improved patient outcomes. The bipolar prosthesis was designed so that the primary articulation would be at the inner bearing of the prosthesis and not at the prosthesis-cartilage interface, thereby decreasing the amount of acetabular erosion and pain that the patient encountered. ^4,19,26 One weakness of the current investigation is that the study period of 1 year may not be long enough to determine if bipolar prostheses provide an advantage regarding acetabular erosion. However, radiographic studies have provided conflicting conclusions about whether this inner bearing continues to function with time, with one study reporting in vivo preservation of inner bearing motion at an average 38 months of followup, ⁶ and others reporting a decrease in motion with time, ^10,15,16 including one study that showed the inner bearing became nearly stiff after only 3 months in vivo. ³¹ A study by Dalldorf et al ¹³ comparing the histologic features of acetabular specimens of patients having revision hemiarthroplasty against age-matched subjects suggests that acetabular wear correlated directly with the amount of time the implant remained in the hip rather than the type of implant used.

Despite debate over radiographic evidence of inner bearing motion, some early studies that compared the bipolar prosthesis with the unipolar prosthesis found better clinical results in patients who had bipolar hemiarthroplasty. ^{5,21,26,27,34} In general, however, these studies had one of two types of weakness. The first weakness is that the earlier studies were retrospective or used historic controls. Younger, more active patients may have been selected to receive bipolar prostheses, thereby introducing selection bias into the study. This was the case in the largest clinical series to date by Yamagata et al. ³⁶ In looking at 1001 patients retrospectively, they found that the younger patients with bipolar prostheses had better hip flexion and better Harris hip scores at followup. The second weakness is that earlier studies based their clinical outcomes on physician-based outcomes such as range of motion and radiographic data (erosion rate, loosening) rather than patient-based outcome data such as functional measurements or quality of life questionnaires.

One of the first prospective studies was done by Wathne et al. ³³ Ninety-two patients with bipolar prostheses and 48 patients with unipolar prostheses did not have any differences at minimum 1-year followup in return to preinjury ambulatory status, pain, or activities of daily living. Although this was a prospective study, patients were enrolled in a cohort fashion whereby the bipolar hemiarthroplasties were done from 1987 to 1990 and the unipolar prostheses were inserted between 1991 and 1992.

Two groups have published prospective randomized studies that compare functional outcomes of patients receiving either unipolar or bipolar hemiarthroplasty. Calder et al ⁷ and Davison et al ¹⁴ published studies with a minimum of 2 years followup. The first study used different femoral stem types and compared a cemented unipolar prosthesis (Thompson) with a cemented bipolar prosthesis (Monk) in patients older than 80 years. ⁷ The only statistically significant difference they found was that patients with unipolar prostheses were more likely to return to their preinjury functional state than patients with bipolar prostheses. However, this difference was not present in a second study that compared unipolar hemiarthroplasty, bipolar hemiarthroplasty, and internal fixation with a compression hip screw in patients 65 to 79 years. ¹⁴ No differences were seen between the groups in functional outcomes, return to preinjury status, patient satisfaction, or hip score. Another prospective 6-month preliminary study by Cornell et al ¹¹ randomized 33 patients who received bipolar prostheses and 15 patients who received unipolar prostheses. The cemented femoral stems used in both groups were the same; the only difference in design of the prosthesis was in the head. Patients with bipolar prostheses did better on walk tests and had better range of motion at 6 months. These differences in function apparently were not perceived by the patients, because the patient-oriented hip scores did not differ at 6 months between the unipolar and bipolar groups.

The current randomized, prospective study is the first that takes into account quality of life issues and functional outcomes. It used well-validated outcome instruments to assess how the patients feel about their health in addition to how well they can do certain tasks including activities of daily living. A hip fracture in an elderly patient profoundly affects the patient’s life not only physically, but also emotionally and socially. All of these factors must be evaluated to obtain a true assessment of a patient’s outcome. This was emphasized in the study by Cornell et al, ¹¹ where the patients with bipolar prostheses did better physically but scored similarly to patients with unipolar prostheses in satisfaction scores. Only by looking at physical and emotional aspects can a true picture of a patient’s quality of life be obtained.

At 1 year, 72% to 74% of patients who were community ambulators before fracture retained that status. Other studies have found similar results in that more than 50% of patients can return to ambulatory levels of baseline or near baseline. ^1,7,11 The percentages in the study by Wathne et al, ³³ for example, were between 70% and 75%. The American Academy of Orthopaedic Surgeons states that 50% of patients lose their ability to walk independently after hip fracture. ³ Although the current study did not differentiate between the patient who was a community ambulator with assistive devices from the patient who did not need assistive devices to walk in the community, the ability to remain active and confident enough to ambulate outside one’s home after hip fracture should be considered a positive outcome. When looking at the mortality rate associated with femoral neck fractures, the results of the current study were comparable to those of other studies. ^{1,2,12,22,25,35}

When the Short Form-36 scores were examined between the two groups at each point and within each group at increasing points, no differences were found. The comparisons that were closest to approaching significance were in the Short Form-36 general health category, between the baseline and the 1-year values. The patient scores actually increased from baseline to 1-year followup, but p values remained above 0.10. Most of the other baseline to 1-year comparisons were within 5 points, but general health increased by approximately 10 points in both groups. This may have been because of recall bias in the baseline data, where patients were questioned about prefracture feelings while in pain from their fractures, although this bias probably should have been seen in all of the categories. Another explanation is that in some cases, patients who had been neglecting their overall general health now were given the opportunity to optimize other medical problems (cardiac disease, diabetes) while in the hospital.

With no statistically significant differences seen between any of the data points, it is concluded that bipolar endoprostheses provide no advantage in the treatment of displaced femoral neck fractures in patients older than 65 years who are not institutionalized or otherwise cognitively or physically impaired regarding quality of life or functional outcomes in the study period. However, because of the short-term nature of this prospective study, additional followup of this group of patients is needed to determine whether the rate of symptomatic and radiographic acetabular erosion is affected by the type of prosthesis used.