Nam D

Postoperative dislocation remains a common complication of primary total hip arthroplasties (THAs), affecting less than 1% to more than 10% in reported series.^1,2 In large datasets for modern implants, the incidence of dislocation is 2% to 4%.^3,4 Given that more than 200,000 THAs are performed in the United States each year,⁵ these low percentages represent a large number of patients. The multiplex patient variables that affect THA stability include age, sex, body mass index (BMI), and comorbid conditions.^6-8 Surgical approach, restoration of leg length and femoral offset, femoral head size, and component positioning are also important surgical factors that can increase or decrease the incidence of dislocation.^3,8,9 In particular, appropriate acetabular component orientation is crucial; surgeons can control this factor and thereby limit the occurrence of dislocation.¹⁰ Furthermore, acetabular malpositioning can increase the risk of liner fractures and accelerate bearing-surface wear.^11-14

To minimize the risk of postoperative dislocation, surgeons traditionally have targeted the Lewinnek safe zone, with its mean (SD) inclination of 40° (10°) and mean (SD) anteversion of 15° (10°), for acetabular component orientation.¹⁵ However, the applicability of this target zone to preventing hip instability using modern implant designs, components, and surgical techniques remains unknown. Achieving acetabular orientation based on maximizing range of motion (ROM) before impingement may be optimal, with anteversion from 20° to 30° and inclination from 40° to 45°.^16,17 Furthermore, mean (SD) native acetabular anteversion ranges from 21.3° (6.2°) for men to 24.6° (6.6°) for women.¹⁸ Placing THA acetabular components near the native range for anteversion may best provide impingement-free ROM and thus optimize THA stability,^16,19 but this has not been proved in a clinical study.

Early dislocation is typically classified as occurring within 6 months after surgery,⁹ with almost 80% of dislocations occurring within 3 months after surgery.¹⁰ Surgeon-specific factors, such as acetabular component positioning, are thought to have a predominant effect on dislocations in the early postoperative period.¹⁰ Computer-assisted surgery (CAS), such as imageless navigation, is more accurate than conventional methods for acetabular component placement,^20-23 but the clinical relevance of improving accuracy for acetabular component placement has not been shown with respect to altering patient outcomes.²³

We conducted a study in a large single-surgeon patient cohort to determine the incidence of early postoperative dislocation with target anteversion increased to 25°, approximating mean native acetabular anteversion.^16,19 In addition, we sought to determine the accuracy of imageless navigation in achieving target acetabular component placement.

Materials and Methods

After obtaining institutional review board approval for this retrospective clinical study, we reviewed 671 consecutive cases of primary THA performed by a single surgeon using an imageless CAS system (AchieveCAS; Smith & Nephew, Memphis, Tennessee) between July 2006 and October 2012. THAs were excluded if a metal-on-metal bearing surface was used, if an adequate 6-week postoperative supine anteroposterior (AP) pelvis radiograph was unavailable, or if 6-month clinical follow-up findings were not available (Figure 1). The quality of AP radiographs was deemed poor if they were not centered on the symphysis pubis and if the sacrococcygeal joint was not centered over the symphysis pubis. After exclusion criteria were applied, 553 arthroplasties (479 patients) with a mean (SD) follow-up of 2.4 (1.4) years remained. Perioperative demographic data and component sizes are listed in Table 1.

During surgery, the anterior pelvic plane, defined by the anterior-superior iliac spines and pubic tubercle, was registered with the CAS system with the patient in the supine position. THA was performed with the patient in the lateral decubitus position using a posterolateral technique. For all patients, the surgeon used a hemispherical acetabular component (R3 Acetabular System; Smith & Nephew); bearings that were either metal on highly cross-linked polyethylene (XLPE) or Oxinium (Smith & Nephew) on XLPE; and neutral XLPE acetabular inserts. The goals for acetabular inclination and anteversion were 40° and 25°, respectively, with ±10° each for the target zone. The CAS system was used to adjust target anteversion for sagittal pelvic tilt.²⁴ Uncemented femoral components were used for all patients, and the goal for femoral component anteversion was 15°. Transosseous repair of the posterior capsule and short external rotators was performed after component implantation.²⁵

On each 6-week postoperative radiograph, acetabular orientation was measured with Ein-Bild-Röntgen-Analyse (EBRA; University of Innsbruck, Austria) software, which provides a validated method for measuring acetabular inclination and anteversion on supine AP pelvis radiographs.^10,26 Pelvic boundaries were delineated with grid lines defining pelvic position. Reference points around the projections of the prosthetic femoral head, the hemispherical cup, and the rim of the cup were marked (Figure 2). EBRA calculated radiographic inclination and anteversion of the acetabular component based on the spatial position of the cup center in relation to the plane of the radiograph and the pelvic position.²⁶

Charts were reviewed to identify patients with early postoperative dislocations, as well as dislocation timing, recurrence, and other characteristics. We defined early dislocation as instability occurring within 6 months after surgery. Revision surgery for instability was also identified.

For the statistical analysis, orientation error was defined as the absolute value of the difference between target orientation (40° inclination, 25° anteversion) and radiographic measurements. Repeated-measures multiple regression with the generalized estimating equations approach was used to identify baseline patient characteristics (age, sex, BMI, primary diagnosis, laterality) associated with component positioning outside of our targeted ranges for inclination and anteversion. Fisher exact tests were used to examine the relationship between dislocation and component placement in either the Lewinnek safe zone or our targeted zone. All tests were 2-sided with a significance level of .05. All analyses were performed with SAS for Windows 9.3 (SAS Institute, Cary, North Carolina).

Results

Mean (SD) acetabular inclination was 42.2° (4.9°) (range, 27.6°-65.0°), with a mean (SD) orientation error of 4.2° (3.4°) (Figure 3A). Mean (SD) anteversion was 23.9° (6.5°) (range, 6.2°-48.0°), with a mean (SD) orientation error of 5.2° (4.1°) (Figure 3B). Components were placed outside the Lewinnek safe zone for inclination or anteversion in 46.5% of cases and outside the target zone in 17.7% of cases (Figure 4). Variation in acetabular anteversion alone accounted for 67.3% of target zone outliers (Table 2). Only 0.9% of components were placed outside the target ranges for both inclination and anteversion.

Regression analysis was performed separately for inclination and anteversion to determine the risk factors for placing the acetabular component outside the target orientation ranges. Only higher BMI was associated with malposition with respect to inclination (hazard ratio [HR], 1.059; 95% confidence interval [CI], 1.011-1.111; P = .017). Of obese patients with inclination outside the target range, 90.9% had an inclination angle of more than 50°. Associations between inclination outside the target range and age (P = .769), sex (P = .217), preoperative diagnosis (P > .99), and laterality (P = .106) were statistically insignificant. Only female sex was associated with position of the acetabular component outside the target range for anteversion (HR, 1.871; 95% CI, 1.061-3.299; P = .030). Of female patients with anteversion outside the target range, 70.0% had anteversion of less than 15°. Associations between anteversion outside the target range and age (P = .762), BMI (P = .583), preoperative diagnosis (P > .99), and laterality (P = .235) were statistically insignificant.

Six THAs (1.1%) in 6 patients experienced dislocation within 6 months after surgery (Table 3); mean (SD) time of dislocation was 58.3 (13.8) days after surgery. There was no relationship between dislocation incidence and component placement in the Lewinnek zone (P = .224) or our target zone (P = .287). Of the dislocation cases, 50% involved female patients, and 50% involved right hips. Mean (SD) age of these patients was 53.3 (7.6) years. Mean (SD) BMI was 25.4 (0.9) kg/m². Osteoarthritis was the primary diagnosis for all patients with early dislocation; 32- or 36-mm femoral heads were used in these cases. Two patients had acetabular components placed outside of our target zone. One patient, who had abnormal pelvic obliquity and sagittal tilt from scoliosis (Figures 5A, 5B), had an acetabular component placed outside both the target zone and the Lewinnek safe zone. Mean (SD) acetabular inclination was 39.8° (3.6°), and mean (SD) anteversion was 21.8° (7.3°) (Figure 5C). Two dislocations resulted from trauma, 1 dislocation was related to hyperlaxity, 1 patient had cerebral palsy, and 1 patient had no evident predisposing risk factors. Three patients (0.54%) had multiple episodes of instability requiring revision during the follow-up period.

Discussion

To our knowledge, this study represents the largest cohort of primary THAs performed with an imageless navigation system. Our results showed that increasing targeted acetabular anteversion to 25° using a posterolateral surgical approach and modern implants resulted in a 1.1% incidence of early dislocation and a 0.54% incidence of recurrent instability requiring reoperation. Of the patients with a dislocation, only 1 did not experience trauma and did not have a risk factor for dislocation. Only 1 patient with a dislocation had acetabular components positioned outside both the target zone and the Lewinnek safe zone. The acetabular component was placed within the target zone in 82.3% of cases in which the imageless navigation system was used. In our cohort, BMI was the only risk factor for placement of the acetabular component outside our target range for inclination, and sex was associated with components outside the target range for anteversion.

Early dislocation after THA is often related to improper implant orientation, inadequate restoration of offset and myofascial tension, and decreased femoral head–neck ratio.⁸ Although dislocation rates in the literature vary widely,^1,2 Medicare data suggest that the rate for the first 6 months after surgery can be as high as 4.21%.^3,4 Although use of femoral heads with a diameter of 32 mm or larger may decrease this rate to 2.14%,³ accurate acetabular component orientation helps prevent postoperative dislocation.¹⁰ Using an imageless navigation system to target 25° of anteversion and 40° of inclination resulted in an early-dislocation rate about 49% less than the rate in a Medicare population treated with similar, modern implants.³

Callanan and colleagues¹¹ found that freehand techniques were inaccurate for acetabular positioning in up to 50% of cases, and several studies have demonstrated that imageless navigation systems were more accurate than conventional guides.^20,21,27-29 Higher BMI has been implicated as a risk factor for acetabular malpositioning in several studies of the accuracy of freehand techniques¹¹ and imageless navigation techniques.^23,30 Soft-tissue impediment to the component insertion handle poses a risk of increased inclination and inadequate anteversion, regardless of method used (conventional, CAS). When the acetabular component is placed freehand in obese patients, it is difficult to judge the position of the pelvis on the operating room table. For imageless navigation, a larger amount of adipose tissue over bony landmarks may limit the accuracy of anterior pelvic plane registration.³⁰ Sex typically is not cited as a risk factor for inaccurate acetabular component positioning. We speculate that omitted-variable bias may explain the observed association between female sex and anteversion. For example, changes in postoperative pelvic tilt alter apparent anteversion on plain radiographs,^31-34 but preoperative and postoperative sagittal pelvic tilt was not recorded in this study.

The proper position of the acetabular component has been debated.^15,16,35,36 Although it is generally agreed that inclination of 40° ± 10° balances ROM, stability, and bearing-surface wear,^12,13,15,16 proposed targets for anteversion vary widely, from 0° to 40°.^35,36 Patel and colleagues¹⁶ formulated computer models based on cadaveric specimens to determine that THA impingement was minimized when the acetabular component was placed to match the native anteversion of the acetabulum.In their study model, 20° of anteversion paralleled native acetabular orientation. Tohtz and colleagues¹⁸ reviewed computed tomography scans of 144 female hips and 192 male hips and found that mean (SD) anteversion was 24.6° (6.6°) for women and 21.3° (6.2°) for men. Whether native anatomy is a valid reference for acetabular anteversion is controversial,¹⁹ and definitive recommendations for target anteversion cannot be made, as the effect of acetabular anteversion on the wear of various bearing materials is unknown.¹⁴ Yet, as with inclination, ideal anteversion is likely a compromise between maximizing impingement-free ROM and minimizing wear.

The present study had several limitations. A single-surgeon patient series was reviewed retrospectively, and there was no control group. We determined the incidence only of early dislocation, and 5.3% of THAs that were not metal-on-metal were either lost to follow-up or had inadequate radiographs. However, of the patients excluded for inadequate radiographs, none had an early dislocation. The effects of our surgical techniques on long-term outcomes, bearing wear, and dislocation are unknown. We were not able to comment on the direction of dislocation for any of the 6 patients with early dislocation, as all dislocations were reduced at facilities other than our hospital. Therefore, we cannot determine whether increasing acetabular anteversion resulted in a larger number of anterior versus posterior dislocations.¹⁵

We did not use CAS to place any of the femoral components. Therefore, we could not accurately target combined anteversion, defined as the sum of acetabular and femoral version, which may be an important determinant of THA stability.²⁸ Although restoration of femoral offset and leg length is important in preventing THA dislocation,⁸ the CAS techniques used did not influence these parameters, and they were not measured.

As an imageless navigation system was used, there were no preoperative axial images, which could have been used to assess native acetabular orientation. This limited our assessment with respect to matching each patient’s natural anteversion. Imageless navigation, which references only the anterior pelvic plane, may not be reliable in patients with excessive sagittal pelvic tilt.³⁷ Furthermore, changes in the functional position of the pelvis from supine to sitting to standing were not accounted for, and changes in sagittal tilt between these positions can be significant.³⁸ Changes in sagittal pelvic tilt affect measurement of acetabular anteversion on plain radiographs, with anterior tilt reducing apparent anteversion and posterior tilt increasing it.^32,34 Although postoperative computed tomography is the gold standard for assessing acetabular component orientation, EBRA significantly reduces errors of measurement on plain radiographs.¹⁰ Some variability in measured anteversion may be explained by our surgical technique. In particular, if the cup was uncovered anteriorly, additional anteversion was usually accepted during surgery to minimize anterior impingement and limit the risk of iliopsoas tendonitis.^16,39

Our study results suggested that increasing target acetabular anteversion to 25° may reduce the incidence of early postoperative instability relative to rates reported in the literature. Despite the higher accuracy of component placement with an imageless navigation system, dislocations occurred in patients with acetabular components positioned in our target zone and in the historical safe zone. These dislocations support the notion that there likely is no absolute safe range for acetabular component positioning, as THA stability depends on many factors. Ideal targets for implant orientation for acetabulum and femur may be patient-specific.^16,19 Investigators should prospectively evaluate patient-specific THA component positioning and determine its effect on postoperative dislocation and bearing-surface wear. As specific implant targets are further defined, tools that are more precise and accurate than conventional techniques will be needed to achieve goal component positioning. Our study results confirmed that imageless navigation is an accurate method for achieving acetabular orientation targets.

Postoperative dislocation remains a common complication of primary total hip arthroplasties (THAs), affecting less than 1% to more than 10% in reported series.^1,2 In large datasets for modern implants, the incidence of dislocation is 2% to 4%.^3,4 Given that more than 200,000 THAs are performed in the United States each year,⁵ these low percentages represent a large number of patients. The multiplex patient variables that affect THA stability include age, sex, body mass index (BMI), and comorbid conditions.^6-8 Surgical approach, restoration of leg length and femoral offset, femoral head size, and component positioning are also important surgical factors that can increase or decrease the incidence of dislocation.^3,8,9 In particular, appropriate acetabular component orientation is crucial; surgeons can control this factor and thereby limit the occurrence of dislocation.¹⁰ Furthermore, acetabular malpositioning can increase the risk of liner fractures and accelerate bearing-surface wear.^11-14

To minimize the risk of postoperative dislocation, surgeons traditionally have targeted the Lewinnek safe zone, with its mean (SD) inclination of 40° (10°) and mean (SD) anteversion of 15° (10°), for acetabular component orientation.¹⁵ However, the applicability of this target zone to preventing hip instability using modern implant designs, components, and surgical techniques remains unknown. Achieving acetabular orientation based on maximizing range of motion (ROM) before impingement may be optimal, with anteversion from 20° to 30° and inclination from 40° to 45°.^16,17 Furthermore, mean (SD) native acetabular anteversion ranges from 21.3° (6.2°) for men to 24.6° (6.6°) for women.¹⁸ Placing THA acetabular components near the native range for anteversion may best provide impingement-free ROM and thus optimize THA stability,^16,19 but this has not been proved in a clinical study.

Early dislocation is typically classified as occurring within 6 months after surgery,⁹ with almost 80% of dislocations occurring within 3 months after surgery.¹⁰ Surgeon-specific factors, such as acetabular component positioning, are thought to have a predominant effect on dislocations in the early postoperative period.¹⁰ Computer-assisted surgery (CAS), such as imageless navigation, is more accurate than conventional methods for acetabular component placement,^20-23 but the clinical relevance of improving accuracy for acetabular component placement has not been shown with respect to altering patient outcomes.²³

We conducted a study in a large single-surgeon patient cohort to determine the incidence of early postoperative dislocation with target anteversion increased to 25°, approximating mean native acetabular anteversion.^16,19 In addition, we sought to determine the accuracy of imageless navigation in achieving target acetabular component placement.

Materials and Methods

After obtaining institutional review board approval for this retrospective clinical study, we reviewed 671 consecutive cases of primary THA performed by a single surgeon using an imageless CAS system (AchieveCAS; Smith & Nephew, Memphis, Tennessee) between July 2006 and October 2012. THAs were excluded if a metal-on-metal bearing surface was used, if an adequate 6-week postoperative supine anteroposterior (AP) pelvis radiograph was unavailable, or if 6-month clinical follow-up findings were not available (Figure 1). The quality of AP radiographs was deemed poor if they were not centered on the symphysis pubis and if the sacrococcygeal joint was not centered over the symphysis pubis. After exclusion criteria were applied, 553 arthroplasties (479 patients) with a mean (SD) follow-up of 2.4 (1.4) years remained. Perioperative demographic data and component sizes are listed in Table 1.

During surgery, the anterior pelvic plane, defined by the anterior-superior iliac spines and pubic tubercle, was registered with the CAS system with the patient in the supine position. THA was performed with the patient in the lateral decubitus position using a posterolateral technique. For all patients, the surgeon used a hemispherical acetabular component (R3 Acetabular System; Smith & Nephew); bearings that were either metal on highly cross-linked polyethylene (XLPE) or Oxinium (Smith & Nephew) on XLPE; and neutral XLPE acetabular inserts. The goals for acetabular inclination and anteversion were 40° and 25°, respectively, with ±10° each for the target zone. The CAS system was used to adjust target anteversion for sagittal pelvic tilt.²⁴ Uncemented femoral components were used for all patients, and the goal for femoral component anteversion was 15°. Transosseous repair of the posterior capsule and short external rotators was performed after component implantation.²⁵

On each 6-week postoperative radiograph, acetabular orientation was measured with Ein-Bild-Röntgen-Analyse (EBRA; University of Innsbruck, Austria) software, which provides a validated method for measuring acetabular inclination and anteversion on supine AP pelvis radiographs.^10,26 Pelvic boundaries were delineated with grid lines defining pelvic position. Reference points around the projections of the prosthetic femoral head, the hemispherical cup, and the rim of the cup were marked (Figure 2). EBRA calculated radiographic inclination and anteversion of the acetabular component based on the spatial position of the cup center in relation to the plane of the radiograph and the pelvic position.²⁶

Charts were reviewed to identify patients with early postoperative dislocations, as well as dislocation timing, recurrence, and other characteristics. We defined early dislocation as instability occurring within 6 months after surgery. Revision surgery for instability was also identified.

For the statistical analysis, orientation error was defined as the absolute value of the difference between target orientation (40° inclination, 25° anteversion) and radiographic measurements. Repeated-measures multiple regression with the generalized estimating equations approach was used to identify baseline patient characteristics (age, sex, BMI, primary diagnosis, laterality) associated with component positioning outside of our targeted ranges for inclination and anteversion. Fisher exact tests were used to examine the relationship between dislocation and component placement in either the Lewinnek safe zone or our targeted zone. All tests were 2-sided with a significance level of .05. All analyses were performed with SAS for Windows 9.3 (SAS Institute, Cary, North Carolina).

Results

Mean (SD) acetabular inclination was 42.2° (4.9°) (range, 27.6°-65.0°), with a mean (SD) orientation error of 4.2° (3.4°) (Figure 3A). Mean (SD) anteversion was 23.9° (6.5°) (range, 6.2°-48.0°), with a mean (SD) orientation error of 5.2° (4.1°) (Figure 3B). Components were placed outside the Lewinnek safe zone for inclination or anteversion in 46.5% of cases and outside the target zone in 17.7% of cases (Figure 4). Variation in acetabular anteversion alone accounted for 67.3% of target zone outliers (Table 2). Only 0.9% of components were placed outside the target ranges for both inclination and anteversion.

Regression analysis was performed separately for inclination and anteversion to determine the risk factors for placing the acetabular component outside the target orientation ranges. Only higher BMI was associated with malposition with respect to inclination (hazard ratio [HR], 1.059; 95% confidence interval [CI], 1.011-1.111; P = .017). Of obese patients with inclination outside the target range, 90.9% had an inclination angle of more than 50°. Associations between inclination outside the target range and age (P = .769), sex (P = .217), preoperative diagnosis (P > .99), and laterality (P = .106) were statistically insignificant. Only female sex was associated with position of the acetabular component outside the target range for anteversion (HR, 1.871; 95% CI, 1.061-3.299; P = .030). Of female patients with anteversion outside the target range, 70.0% had anteversion of less than 15°. Associations between anteversion outside the target range and age (P = .762), BMI (P = .583), preoperative diagnosis (P > .99), and laterality (P = .235) were statistically insignificant.

Six THAs (1.1%) in 6 patients experienced dislocation within 6 months after surgery (Table 3); mean (SD) time of dislocation was 58.3 (13.8) days after surgery. There was no relationship between dislocation incidence and component placement in the Lewinnek zone (P = .224) or our target zone (P = .287). Of the dislocation cases, 50% involved female patients, and 50% involved right hips. Mean (SD) age of these patients was 53.3 (7.6) years. Mean (SD) BMI was 25.4 (0.9) kg/m². Osteoarthritis was the primary diagnosis for all patients with early dislocation; 32- or 36-mm femoral heads were used in these cases. Two patients had acetabular components placed outside of our target zone. One patient, who had abnormal pelvic obliquity and sagittal tilt from scoliosis (Figures 5A, 5B), had an acetabular component placed outside both the target zone and the Lewinnek safe zone. Mean (SD) acetabular inclination was 39.8° (3.6°), and mean (SD) anteversion was 21.8° (7.3°) (Figure 5C). Two dislocations resulted from trauma, 1 dislocation was related to hyperlaxity, 1 patient had cerebral palsy, and 1 patient had no evident predisposing risk factors. Three patients (0.54%) had multiple episodes of instability requiring revision during the follow-up period.

Discussion

To our knowledge, this study represents the largest cohort of primary THAs performed with an imageless navigation system. Our results showed that increasing targeted acetabular anteversion to 25° using a posterolateral surgical approach and modern implants resulted in a 1.1% incidence of early dislocation and a 0.54% incidence of recurrent instability requiring reoperation. Of the patients with a dislocation, only 1 did not experience trauma and did not have a risk factor for dislocation. Only 1 patient with a dislocation had acetabular components positioned outside both the target zone and the Lewinnek safe zone. The acetabular component was placed within the target zone in 82.3% of cases in which the imageless navigation system was used. In our cohort, BMI was the only risk factor for placement of the acetabular component outside our target range for inclination, and sex was associated with components outside the target range for anteversion.

Early dislocation after THA is often related to improper implant orientation, inadequate restoration of offset and myofascial tension, and decreased femoral head–neck ratio.⁸ Although dislocation rates in the literature vary widely,^1,2 Medicare data suggest that the rate for the first 6 months after surgery can be as high as 4.21%.^3,4 Although use of femoral heads with a diameter of 32 mm or larger may decrease this rate to 2.14%,³ accurate acetabular component orientation helps prevent postoperative dislocation.¹⁰ Using an imageless navigation system to target 25° of anteversion and 40° of inclination resulted in an early-dislocation rate about 49% less than the rate in a Medicare population treated with similar, modern implants.³

Callanan and colleagues¹¹ found that freehand techniques were inaccurate for acetabular positioning in up to 50% of cases, and several studies have demonstrated that imageless navigation systems were more accurate than conventional guides.^20,21,27-29 Higher BMI has been implicated as a risk factor for acetabular malpositioning in several studies of the accuracy of freehand techniques¹¹ and imageless navigation techniques.^23,30 Soft-tissue impediment to the component insertion handle poses a risk of increased inclination and inadequate anteversion, regardless of method used (conventional, CAS). When the acetabular component is placed freehand in obese patients, it is difficult to judge the position of the pelvis on the operating room table. For imageless navigation, a larger amount of adipose tissue over bony landmarks may limit the accuracy of anterior pelvic plane registration.³⁰ Sex typically is not cited as a risk factor for inaccurate acetabular component positioning. We speculate that omitted-variable bias may explain the observed association between female sex and anteversion. For example, changes in postoperative pelvic tilt alter apparent anteversion on plain radiographs,^31-34 but preoperative and postoperative sagittal pelvic tilt was not recorded in this study.

The proper position of the acetabular component has been debated.^15,16,35,36 Although it is generally agreed that inclination of 40° ± 10° balances ROM, stability, and bearing-surface wear,^12,13,15,16 proposed targets for anteversion vary widely, from 0° to 40°.^35,36 Patel and colleagues¹⁶ formulated computer models based on cadaveric specimens to determine that THA impingement was minimized when the acetabular component was placed to match the native anteversion of the acetabulum.In their study model, 20° of anteversion paralleled native acetabular orientation. Tohtz and colleagues¹⁸ reviewed computed tomography scans of 144 female hips and 192 male hips and found that mean (SD) anteversion was 24.6° (6.6°) for women and 21.3° (6.2°) for men. Whether native anatomy is a valid reference for acetabular anteversion is controversial,¹⁹ and definitive recommendations for target anteversion cannot be made, as the effect of acetabular anteversion on the wear of various bearing materials is unknown.¹⁴ Yet, as with inclination, ideal anteversion is likely a compromise between maximizing impingement-free ROM and minimizing wear.

The present study had several limitations. A single-surgeon patient series was reviewed retrospectively, and there was no control group. We determined the incidence only of early dislocation, and 5.3% of THAs that were not metal-on-metal were either lost to follow-up or had inadequate radiographs. However, of the patients excluded for inadequate radiographs, none had an early dislocation. The effects of our surgical techniques on long-term outcomes, bearing wear, and dislocation are unknown. We were not able to comment on the direction of dislocation for any of the 6 patients with early dislocation, as all dislocations were reduced at facilities other than our hospital. Therefore, we cannot determine whether increasing acetabular anteversion resulted in a larger number of anterior versus posterior dislocations.¹⁵

We did not use CAS to place any of the femoral components. Therefore, we could not accurately target combined anteversion, defined as the sum of acetabular and femoral version, which may be an important determinant of THA stability.²⁸ Although restoration of femoral offset and leg length is important in preventing THA dislocation,⁸ the CAS techniques used did not influence these parameters, and they were not measured.

As an imageless navigation system was used, there were no preoperative axial images, which could have been used to assess native acetabular orientation. This limited our assessment with respect to matching each patient’s natural anteversion. Imageless navigation, which references only the anterior pelvic plane, may not be reliable in patients with excessive sagittal pelvic tilt.³⁷ Furthermore, changes in the functional position of the pelvis from supine to sitting to standing were not accounted for, and changes in sagittal tilt between these positions can be significant.³⁸ Changes in sagittal pelvic tilt affect measurement of acetabular anteversion on plain radiographs, with anterior tilt reducing apparent anteversion and posterior tilt increasing it.^32,34 Although postoperative computed tomography is the gold standard for assessing acetabular component orientation, EBRA significantly reduces errors of measurement on plain radiographs.¹⁰ Some variability in measured anteversion may be explained by our surgical technique. In particular, if the cup was uncovered anteriorly, additional anteversion was usually accepted during surgery to minimize anterior impingement and limit the risk of iliopsoas tendonitis.^16,39

Our study results suggested that increasing target acetabular anteversion to 25° may reduce the incidence of early postoperative instability relative to rates reported in the literature. Despite the higher accuracy of component placement with an imageless navigation system, dislocations occurred in patients with acetabular components positioned in our target zone and in the historical safe zone. These dislocations support the notion that there likely is no absolute safe range for acetabular component positioning, as THA stability depends on many factors. Ideal targets for implant orientation for acetabulum and femur may be patient-specific.^16,19 Investigators should prospectively evaluate patient-specific THA component positioning and determine its effect on postoperative dislocation and bearing-surface wear. As specific implant targets are further defined, tools that are more precise and accurate than conventional techniques will be needed to achieve goal component positioning. Our study results confirmed that imageless navigation is an accurate method for achieving acetabular orientation targets.

Postoperative dislocation remains a common complication of primary total hip arthroplasties (THAs), affecting less than 1% to more than 10% in reported series.^1,2 In large datasets for modern implants, the incidence of dislocation is 2% to 4%.^3,4 Given that more than 200,000 THAs are performed in the United States each year,⁵ these low percentages represent a large number of patients. The multiplex patient variables that affect THA stability include age, sex, body mass index (BMI), and comorbid conditions.^6-8 Surgical approach, restoration of leg length and femoral offset, femoral head size, and component positioning are also important surgical factors that can increase or decrease the incidence of dislocation.^3,8,9 In particular, appropriate acetabular component orientation is crucial; surgeons can control this factor and thereby limit the occurrence of dislocation.¹⁰ Furthermore, acetabular malpositioning can increase the risk of liner fractures and accelerate bearing-surface wear.^11-14

To minimize the risk of postoperative dislocation, surgeons traditionally have targeted the Lewinnek safe zone, with its mean (SD) inclination of 40° (10°) and mean (SD) anteversion of 15° (10°), for acetabular component orientation.¹⁵ However, the applicability of this target zone to preventing hip instability using modern implant designs, components, and surgical techniques remains unknown. Achieving acetabular orientation based on maximizing range of motion (ROM) before impingement may be optimal, with anteversion from 20° to 30° and inclination from 40° to 45°.^16,17 Furthermore, mean (SD) native acetabular anteversion ranges from 21.3° (6.2°) for men to 24.6° (6.6°) for women.¹⁸ Placing THA acetabular components near the native range for anteversion may best provide impingement-free ROM and thus optimize THA stability,^16,19 but this has not been proved in a clinical study.

Early dislocation is typically classified as occurring within 6 months after surgery,⁹ with almost 80% of dislocations occurring within 3 months after surgery.¹⁰ Surgeon-specific factors, such as acetabular component positioning, are thought to have a predominant effect on dislocations in the early postoperative period.¹⁰ Computer-assisted surgery (CAS), such as imageless navigation, is more accurate than conventional methods for acetabular component placement,^20-23 but the clinical relevance of improving accuracy for acetabular component placement has not been shown with respect to altering patient outcomes.²³

We conducted a study in a large single-surgeon patient cohort to determine the incidence of early postoperative dislocation with target anteversion increased to 25°, approximating mean native acetabular anteversion.^16,19 In addition, we sought to determine the accuracy of imageless navigation in achieving target acetabular component placement.

Materials and Methods

After obtaining institutional review board approval for this retrospective clinical study, we reviewed 671 consecutive cases of primary THA performed by a single surgeon using an imageless CAS system (AchieveCAS; Smith & Nephew, Memphis, Tennessee) between July 2006 and October 2012. THAs were excluded if a metal-on-metal bearing surface was used, if an adequate 6-week postoperative supine anteroposterior (AP) pelvis radiograph was unavailable, or if 6-month clinical follow-up findings were not available (Figure 1). The quality of AP radiographs was deemed poor if they were not centered on the symphysis pubis and if the sacrococcygeal joint was not centered over the symphysis pubis. After exclusion criteria were applied, 553 arthroplasties (479 patients) with a mean (SD) follow-up of 2.4 (1.4) years remained. Perioperative demographic data and component sizes are listed in Table 1.

During surgery, the anterior pelvic plane, defined by the anterior-superior iliac spines and pubic tubercle, was registered with the CAS system with the patient in the supine position. THA was performed with the patient in the lateral decubitus position using a posterolateral technique. For all patients, the surgeon used a hemispherical acetabular component (R3 Acetabular System; Smith & Nephew); bearings that were either metal on highly cross-linked polyethylene (XLPE) or Oxinium (Smith & Nephew) on XLPE; and neutral XLPE acetabular inserts. The goals for acetabular inclination and anteversion were 40° and 25°, respectively, with ±10° each for the target zone. The CAS system was used to adjust target anteversion for sagittal pelvic tilt.²⁴ Uncemented femoral components were used for all patients, and the goal for femoral component anteversion was 15°. Transosseous repair of the posterior capsule and short external rotators was performed after component implantation.²⁵

On each 6-week postoperative radiograph, acetabular orientation was measured with Ein-Bild-Röntgen-Analyse (EBRA; University of Innsbruck, Austria) software, which provides a validated method for measuring acetabular inclination and anteversion on supine AP pelvis radiographs.^10,26 Pelvic boundaries were delineated with grid lines defining pelvic position. Reference points around the projections of the prosthetic femoral head, the hemispherical cup, and the rim of the cup were marked (Figure 2). EBRA calculated radiographic inclination and anteversion of the acetabular component based on the spatial position of the cup center in relation to the plane of the radiograph and the pelvic position.²⁶

Charts were reviewed to identify patients with early postoperative dislocations, as well as dislocation timing, recurrence, and other characteristics. We defined early dislocation as instability occurring within 6 months after surgery. Revision surgery for instability was also identified.

For the statistical analysis, orientation error was defined as the absolute value of the difference between target orientation (40° inclination, 25° anteversion) and radiographic measurements. Repeated-measures multiple regression with the generalized estimating equations approach was used to identify baseline patient characteristics (age, sex, BMI, primary diagnosis, laterality) associated with component positioning outside of our targeted ranges for inclination and anteversion. Fisher exact tests were used to examine the relationship between dislocation and component placement in either the Lewinnek safe zone or our targeted zone. All tests were 2-sided with a significance level of .05. All analyses were performed with SAS for Windows 9.3 (SAS Institute, Cary, North Carolina).

Results

Mean (SD) acetabular inclination was 42.2° (4.9°) (range, 27.6°-65.0°), with a mean (SD) orientation error of 4.2° (3.4°) (Figure 3A). Mean (SD) anteversion was 23.9° (6.5°) (range, 6.2°-48.0°), with a mean (SD) orientation error of 5.2° (4.1°) (Figure 3B). Components were placed outside the Lewinnek safe zone for inclination or anteversion in 46.5% of cases and outside the target zone in 17.7% of cases (Figure 4). Variation in acetabular anteversion alone accounted for 67.3% of target zone outliers (Table 2). Only 0.9% of components were placed outside the target ranges for both inclination and anteversion.

Regression analysis was performed separately for inclination and anteversion to determine the risk factors for placing the acetabular component outside the target orientation ranges. Only higher BMI was associated with malposition with respect to inclination (hazard ratio [HR], 1.059; 95% confidence interval [CI], 1.011-1.111; P = .017). Of obese patients with inclination outside the target range, 90.9% had an inclination angle of more than 50°. Associations between inclination outside the target range and age (P = .769), sex (P = .217), preoperative diagnosis (P > .99), and laterality (P = .106) were statistically insignificant. Only female sex was associated with position of the acetabular component outside the target range for anteversion (HR, 1.871; 95% CI, 1.061-3.299; P = .030). Of female patients with anteversion outside the target range, 70.0% had anteversion of less than 15°. Associations between anteversion outside the target range and age (P = .762), BMI (P = .583), preoperative diagnosis (P > .99), and laterality (P = .235) were statistically insignificant.

Six THAs (1.1%) in 6 patients experienced dislocation within 6 months after surgery (Table 3); mean (SD) time of dislocation was 58.3 (13.8) days after surgery. There was no relationship between dislocation incidence and component placement in the Lewinnek zone (P = .224) or our target zone (P = .287). Of the dislocation cases, 50% involved female patients, and 50% involved right hips. Mean (SD) age of these patients was 53.3 (7.6) years. Mean (SD) BMI was 25.4 (0.9) kg/m². Osteoarthritis was the primary diagnosis for all patients with early dislocation; 32- or 36-mm femoral heads were used in these cases. Two patients had acetabular components placed outside of our target zone. One patient, who had abnormal pelvic obliquity and sagittal tilt from scoliosis (Figures 5A, 5B), had an acetabular component placed outside both the target zone and the Lewinnek safe zone. Mean (SD) acetabular inclination was 39.8° (3.6°), and mean (SD) anteversion was 21.8° (7.3°) (Figure 5C). Two dislocations resulted from trauma, 1 dislocation was related to hyperlaxity, 1 patient had cerebral palsy, and 1 patient had no evident predisposing risk factors. Three patients (0.54%) had multiple episodes of instability requiring revision during the follow-up period.

Discussion

To our knowledge, this study represents the largest cohort of primary THAs performed with an imageless navigation system. Our results showed that increasing targeted acetabular anteversion to 25° using a posterolateral surgical approach and modern implants resulted in a 1.1% incidence of early dislocation and a 0.54% incidence of recurrent instability requiring reoperation. Of the patients with a dislocation, only 1 did not experience trauma and did not have a risk factor for dislocation. Only 1 patient with a dislocation had acetabular components positioned outside both the target zone and the Lewinnek safe zone. The acetabular component was placed within the target zone in 82.3% of cases in which the imageless navigation system was used. In our cohort, BMI was the only risk factor for placement of the acetabular component outside our target range for inclination, and sex was associated with components outside the target range for anteversion.

Early dislocation after THA is often related to improper implant orientation, inadequate restoration of offset and myofascial tension, and decreased femoral head–neck ratio.⁸ Although dislocation rates in the literature vary widely,^1,2 Medicare data suggest that the rate for the first 6 months after surgery can be as high as 4.21%.^3,4 Although use of femoral heads with a diameter of 32 mm or larger may decrease this rate to 2.14%,³ accurate acetabular component orientation helps prevent postoperative dislocation.¹⁰ Using an imageless navigation system to target 25° of anteversion and 40° of inclination resulted in an early-dislocation rate about 49% less than the rate in a Medicare population treated with similar, modern implants.³

Callanan and colleagues¹¹ found that freehand techniques were inaccurate for acetabular positioning in up to 50% of cases, and several studies have demonstrated that imageless navigation systems were more accurate than conventional guides.^20,21,27-29 Higher BMI has been implicated as a risk factor for acetabular malpositioning in several studies of the accuracy of freehand techniques¹¹ and imageless navigation techniques.^23,30 Soft-tissue impediment to the component insertion handle poses a risk of increased inclination and inadequate anteversion, regardless of method used (conventional, CAS). When the acetabular component is placed freehand in obese patients, it is difficult to judge the position of the pelvis on the operating room table. For imageless navigation, a larger amount of adipose tissue over bony landmarks may limit the accuracy of anterior pelvic plane registration.³⁰ Sex typically is not cited as a risk factor for inaccurate acetabular component positioning. We speculate that omitted-variable bias may explain the observed association between female sex and anteversion. For example, changes in postoperative pelvic tilt alter apparent anteversion on plain radiographs,^31-34 but preoperative and postoperative sagittal pelvic tilt was not recorded in this study.

The proper position of the acetabular component has been debated.^15,16,35,36 Although it is generally agreed that inclination of 40° ± 10° balances ROM, stability, and bearing-surface wear,^12,13,15,16 proposed targets for anteversion vary widely, from 0° to 40°.^35,36 Patel and colleagues¹⁶ formulated computer models based on cadaveric specimens to determine that THA impingement was minimized when the acetabular component was placed to match the native anteversion of the acetabulum.In their study model, 20° of anteversion paralleled native acetabular orientation. Tohtz and colleagues¹⁸ reviewed computed tomography scans of 144 female hips and 192 male hips and found that mean (SD) anteversion was 24.6° (6.6°) for women and 21.3° (6.2°) for men. Whether native anatomy is a valid reference for acetabular anteversion is controversial,¹⁹ and definitive recommendations for target anteversion cannot be made, as the effect of acetabular anteversion on the wear of various bearing materials is unknown.¹⁴ Yet, as with inclination, ideal anteversion is likely a compromise between maximizing impingement-free ROM and minimizing wear.

The present study had several limitations. A single-surgeon patient series was reviewed retrospectively, and there was no control group. We determined the incidence only of early dislocation, and 5.3% of THAs that were not metal-on-metal were either lost to follow-up or had inadequate radiographs. However, of the patients excluded for inadequate radiographs, none had an early dislocation. The effects of our surgical techniques on long-term outcomes, bearing wear, and dislocation are unknown. We were not able to comment on the direction of dislocation for any of the 6 patients with early dislocation, as all dislocations were reduced at facilities other than our hospital. Therefore, we cannot determine whether increasing acetabular anteversion resulted in a larger number of anterior versus posterior dislocations.¹⁵

We did not use CAS to place any of the femoral components. Therefore, we could not accurately target combined anteversion, defined as the sum of acetabular and femoral version, which may be an important determinant of THA stability.²⁸ Although restoration of femoral offset and leg length is important in preventing THA dislocation,⁸ the CAS techniques used did not influence these parameters, and they were not measured.

As an imageless navigation system was used, there were no preoperative axial images, which could have been used to assess native acetabular orientation. This limited our assessment with respect to matching each patient’s natural anteversion. Imageless navigation, which references only the anterior pelvic plane, may not be reliable in patients with excessive sagittal pelvic tilt.³⁷ Furthermore, changes in the functional position of the pelvis from supine to sitting to standing were not accounted for, and changes in sagittal tilt between these positions can be significant.³⁸ Changes in sagittal pelvic tilt affect measurement of acetabular anteversion on plain radiographs, with anterior tilt reducing apparent anteversion and posterior tilt increasing it.^32,34 Although postoperative computed tomography is the gold standard for assessing acetabular component orientation, EBRA significantly reduces errors of measurement on plain radiographs.¹⁰ Some variability in measured anteversion may be explained by our surgical technique. In particular, if the cup was uncovered anteriorly, additional anteversion was usually accepted during surgery to minimize anterior impingement and limit the risk of iliopsoas tendonitis.^16,39

Our study results suggested that increasing target acetabular anteversion to 25° may reduce the incidence of early postoperative instability relative to rates reported in the literature. Despite the higher accuracy of component placement with an imageless navigation system, dislocations occurred in patients with acetabular components positioned in our target zone and in the historical safe zone. These dislocations support the notion that there likely is no absolute safe range for acetabular component positioning, as THA stability depends on many factors. Ideal targets for implant orientation for acetabulum and femur may be patient-specific.^16,19 Investigators should prospectively evaluate patient-specific THA component positioning and determine its effect on postoperative dislocation and bearing-surface wear. As specific implant targets are further defined, tools that are more precise and accurate than conventional techniques will be needed to achieve goal component positioning. Our study results confirmed that imageless navigation is an accurate method for achieving acetabular orientation targets.

Metal-on-metal hip resurfacing arthroplasty (HRA) remains an alternative to total hip arthroplasty (THA) in appropriately selected, younger, active adults with degenerative hip disease.^1-4 While concerns remain regarding the potential for adverse local tissue reactions from wear of the metal-on-metal bearing surface,^5-8 10-year data from the Australian Orthopaedic Association National Joint Replacement Registry Annual Report⁹ showed a revision rate of only 6.3% when the Birmingham Hip Resurfacing (BHR) System was used (Smith & Nephew Inc, Memphis, Tennessee).In addition, in an independent review of 230 consecutive BHRs at a mean follow-up of 10.4 years, Coulter and colleagues¹⁰ showed encouraging clinical results, with a mean Oxford Hip Score of 45.0 and a mean University of California at Los Angeles (UCLA) activity score of 7.4.

Similar to the prior increase in popularity of HRA, hip arthroscopy has also become much more commonplace, and its indications continue to evolve.¹¹ Hip arthroscopy has been used in the native hip joint to manage femoroacetabular impingement, labral tears, and iliopsoas tendinopathy, among other conditions.¹² In addition, the use of hip arthroscopy has not been limited to the native hip but also has increased as a diagnostic and therapeutic procedure after hip arthroplasties. Bajwa and Villar¹² found hip arthroscopy to be diagnostic in 23 of 24 patients who underwent the procedure after a hip arthroplasty, concluding that arthroscopy is a useful adjunct in the diagnosis of symptomatic arthroplasties.

Therefore, hip arthroscopy has been shown to be an effective modality to treat pathology in both the native hip and after hip arthroplasties. However, the effect of a prior hip arthroscopy on the outcome of a subsequent metal-on-metal HRA has not been determined. Piedade and colleagues¹³ showed a prior knee arthroscopy to increase the risk of postoperative complications and subsequent revision after total knee arthroplasty. Complications included reflex sympathetic dystrophy, undiagnosed pain, infection, stiffness, and component loosening. A prior osteochondroplasty at the femoral head-neck junction could increase the risk of femoral neck fracture after a subsequent HRA. Thus, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy and to compare these results with a cohort of patients who received an HRA with no prior hip surgeries. Our hypothesis is that a prior hip arthroscopy will lead to inferior outcomes in patients undergoing HRA. 

Materials and Methods

This study is a retrospective, case-control study using a 1:2 matching analysis. Dr. Su performed all HRAs, which were enrolled in an institutional review board–approved arthroplasty registry. All HRAs were performed using the BHR System. 

The surgical technique for hip resurfacing arthroplasty has been described.¹ All procedures were performed via a posterior approach with the patient in the lateral decubitus position. All patients received a hybrid metal-on-metal hip resurfacing, with an uncemented acetabular component and cemented femoral component. Intraoperative anesthesia for all patients was performed via a combined spinal-epidural anesthetic, and an epidural patient-controlled analgesic was used for the first day postoperatively, followed by a transition to oral analgesics. The sizes of the acetabular and femoral components were recorded for each hip resurfacing. Postoperatively, intermittent pneumatic compression devices were placed upon arrival in the recovery room, and active ankle flexion and extension exercises were initiated immediately after the patient’s neurologic function returned.¹⁴ Aspirin was used for chemical deep venous thrombosis prophylaxis in all patients postoperatively for a period of 6 weeks. Full weight-bearing, with the use of crutches for assistance with balance, was permitted immediately. Crutches were used for a period of 3 weeks prior to being discontinued. 

From a database of 1357 HRAs (all BHR implants) performed between June 2006 and June 2012, 51 patients were identified who received an HRA after a prior hip arthroscopy. Eight patients were excluded because they did not possess adequate clinical documentation or were lost to follow-up. In the remaining 43 patients, there were 32 men and 11 women (21 right hips, 22 left hips), which formed the arthroscopy cohort. Two patients had a history of multiple hip arthroscopies (1 patient with 2 prior procedures, 1 patient with 3 prior procedures). The mean (SD) time from the most recent hip arthroscopy to the HRA was 2.5 (2.5) years. Table 1 presents a summary of the hip arthroscopy procedures (including only the most recent hip arthroscopy procedure in those with multiple arthroscopies).

Patient demographic variables (age, body mass index [BMI]) were recorded preoperatively, along with the Harris Hip Score (HHS),¹⁵ UCLA activity score,¹⁶ Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score,¹⁷ and preoperative hip range of motion (flexion, extension, abduction, adduction, internal rotation, and external rotation). The same clinical indices were assessed postoperatively along with the Short Form-12 (SF-12) Health Survey Score,¹⁸ at the 6-week, 3-month, 6-month, 1-year, and most recent follow-up visits.

Radiographic assessment consisted of a low anteroposterior (AP) pelvic radiograph (with the radiographic beam centered on the pubic symphysis) and a cross-table lateral radiograph obtained at the most recent follow-up visit. Both the acetabular component abduction relative to the inter-teardrop line, and the angle between the femoral stem and the anatomic axis of the femoral shaft (stem-shaft angle) were measured on AP radiographs.^19,20 Acetabular component anteversion was measured on the cross-table lateral radiographs as the angle between the projected long axis of the acetabular opening and a line drawn perpendicular to the long axis plane of the body (Figures A, B).²¹

The same registry database was used to identify patients who received an HRA without a prior history of arthroscopy or hip surgery. A 1:2 matching analysis for those patients with a prior hip arthroscopy to those without a prior hip arthroscopy was performed to formulate a control group (control cohort) of 86 patients. Each patient in the arthroscopy cohort was matched with 2 patients in the control cohort based on the following parameters: age (± 6 years), sex (same), BMI (± 4 kg/m²), femoral head size (± 4 mm), and preoperative HHS and WOMAC scores (± 7 points). In the event an arthroscopy patient matched to 2 or more control patients, the patients who minimized the least squared error among the matching variables were selected. 

Statistical Analysis

All data were collected and analyzed using Microsoft Excel software (Microsoft Corporation, Redmond, Washington). Statistical comparisons between the 2 cohorts regarding demographic variables, clinical outcomes, and radiographic alignment were performed using an unpaired, Student 2-tailed t test, with statistical significance set at P ≤ .05. 

Results

A comparison of the results of the 1:2 matching analysis between the arthroscopy and control cohorts is presented in Table 2. There was no significant difference in the preoperative age, BMI, femoral head size, HHS, or WOMAC score between the 2 cohorts. However, the control cohort did show a more severe, preoperative flexion contracture (as expressed by a decreased amount of extension) and a decreased amount of preoperative abduction (Table 3). The preoperative UCLA activity score was also decreased in the control cohort, but this was not statistically significant.

The mean (SD) follow-up was 2.0 (1.0) years in the arthroscopy cohort and 2.1 (1.1) years in the control cohort. There was no significant difference in radiographic alignment between the 2 cohorts. The stem-shaft angle was 139.3° (SD, 5.4°) in the arthroscopy cohort (vs 138.3° [SD, 5.5°] in the control cohort; P = .3), the acetabular abduction was 43.9° (SD, 5.8°) in the arthroscopy cohort (vs 42.9° [SD, 6.1°] in the control cohort; P = .4), and the acetabular anteversion was 21.1° (SD, 7.5°) in the arthroscopy cohort (vs 20.8° [SD, 7.1°] in the control cohort; P = .8). 

At 6-week follow-up, the arthroscopy cohort showed a significantly decreased WOMAC score compared with the control cohort (72.9 [SD, 15.5] vs 80.5 [SD, 11.8], respectively; P = .05). In addition, there was a trend towards a decreased SF-12 mental component score in the arthroscopy cohort (52.2 [SD, 9.3] vs 56.5 [SD, 7.8] in the control cohort; P = .06). However, none of the remaining clinical indices showed a significant difference between the 2 cohorts, and there was no difference in range of motion between the 2 cohorts at the 6-week follow-up visit (Table 4).

In addition, at 3-month follow-up, no statistically significant differences were seen between the 2 cohorts for any of the clinical indices or range of motion values. Both groups continued to improve rapidly, with HHS of 96.9 (SD, 3.5) in the arthroscopy cohort and 95.5 (SD, 6.6) in the control cohort, and WOMAC scores of 88.7 (SD, 10.2) and 89.5 (SD, 9.8), respectively (Table 5). Similarly, at the 6-month and 1-year follow-up intervals, the 2 cohorts showed continued improvement in their clinical measures, with no statistically significant differences between the 2 cohorts (Tables 6, 7). 

At the most recent follow-up visit, more than 1 year after surgery, the HHS was 99.5 (SD, 1.3) in the arthroscopy cohort and 99.2 (SD, 9.7) in the control cohort (P = .9), and the WOMAC score was 93.5 (SD, 11.3) and 92.4 (SD, 12.2), respectively (P = .8). No significant perioperative complications were seen in the arthroscopy cohort. In the arthroscopy cohort, 1 patient was diagnosed with a deep venous thrombosis 2 weeks after the procedure and was placed on low-molecular-weight heparin and coumadin for treatment. A second patient in the arthroscopy cohort had continued serosanguinous drainage for 4 days postoperatively, which resolved with continued compressive dressings. To date, no patients in the arthroscopy or control cohorts have required a second operation or revision of their components.

Discussion

Given the increasing prevalence of hip arthroscopies to treat multiple disorders of the native joint, it is important to assess the potential consequences of these procedures on future arthroplasties. Piedade and colleagues,¹³ in a retrospective review of 1474 primary total knee arthroplasties, showed a prior bony procedure (high tibial osteotomy, tibial plateau fracture, patellar realignment) to be predictive of decreased range of motion postoperatively. In addition, a prior knee arthroscopy was associated with a higher rate of postoperative complications, with 30% of the complications requiring a reoperation, and 8.3% of the complications requiring a revision total knee arthroplasty. Kaplan-Meier survival curves showed a survival rate of only 86.8% in those patients with a prior knee arthroscopy (vs 98.1% in those without a prior knee surgery).²² Therefore, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy. After the initial 6-week follow-up visit, no significant difference was seen in the functional outcomes between those patients with or without a history of prior hip arthroscopy who received an HRA.

After analysis of patient outcomes using multiple clinical measurement tools, at 6-week, 3-month, 6-month, 1-year, and most recent follow-up intervals, the only significant difference between the 2 cohorts was the WOMAC score at 6-week follow-up. Interestingly, there was no significant difference seen in the other clinical assessments, including the SF-12 score, HHS, range of motion, or UCLA activity score (although this did trend towards significance). This can be explained by the difference in both the mode of administration and various metrics assessed by these instruments. In comparison to the HHS evaluation, the patient completes the WOMAC (rather than the clinician) and also provides a more detailed assessment of symptoms, pain, stiffness, and activities of daily living.¹⁷ Therefore, this study suggests that patients with a prior hip arthroscopy may require more time to return to their activities of daily living after an HRA. However, whether the statistically significant difference between the 2 scores translates into a clinically significant difference can be questioned.

The clinical outcomes of this series of patients were excellent at the short-term follow-up, and both groups achieved clinical results comparable to prior reported results of HRA.^1,10,23,24 However, despite these results, there are several limitations to this study. First, longer-term follow-up is required to determine if any significant differences (such as aseptic loosening, infection, and prosthesis survival) are associated with a prior hip arthroscopy. In addition, this study included a relatively small cohort of patients who had a prior hip arthroscopy. However, a relatively large, single-surgeon database of 1357 HRAs was reviewed, with only 51 cases being reported (3.7%). With the increasing popularity of hip arthroscopy, the number of patients presenting for HRA will likely continue to increase. However, despite these limitations, this study shows that a prior hip arthroscopy does not appear to affect the short-term, clinical outcomes of a metal-on-metal HRA.

Metal-on-metal hip resurfacing arthroplasty (HRA) remains an alternative to total hip arthroplasty (THA) in appropriately selected, younger, active adults with degenerative hip disease.^1-4 While concerns remain regarding the potential for adverse local tissue reactions from wear of the metal-on-metal bearing surface,^5-8 10-year data from the Australian Orthopaedic Association National Joint Replacement Registry Annual Report⁹ showed a revision rate of only 6.3% when the Birmingham Hip Resurfacing (BHR) System was used (Smith & Nephew Inc, Memphis, Tennessee).In addition, in an independent review of 230 consecutive BHRs at a mean follow-up of 10.4 years, Coulter and colleagues¹⁰ showed encouraging clinical results, with a mean Oxford Hip Score of 45.0 and a mean University of California at Los Angeles (UCLA) activity score of 7.4.

Similar to the prior increase in popularity of HRA, hip arthroscopy has also become much more commonplace, and its indications continue to evolve.¹¹ Hip arthroscopy has been used in the native hip joint to manage femoroacetabular impingement, labral tears, and iliopsoas tendinopathy, among other conditions.¹² In addition, the use of hip arthroscopy has not been limited to the native hip but also has increased as a diagnostic and therapeutic procedure after hip arthroplasties. Bajwa and Villar¹² found hip arthroscopy to be diagnostic in 23 of 24 patients who underwent the procedure after a hip arthroplasty, concluding that arthroscopy is a useful adjunct in the diagnosis of symptomatic arthroplasties.

Therefore, hip arthroscopy has been shown to be an effective modality to treat pathology in both the native hip and after hip arthroplasties. However, the effect of a prior hip arthroscopy on the outcome of a subsequent metal-on-metal HRA has not been determined. Piedade and colleagues¹³ showed a prior knee arthroscopy to increase the risk of postoperative complications and subsequent revision after total knee arthroplasty. Complications included reflex sympathetic dystrophy, undiagnosed pain, infection, stiffness, and component loosening. A prior osteochondroplasty at the femoral head-neck junction could increase the risk of femoral neck fracture after a subsequent HRA. Thus, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy and to compare these results with a cohort of patients who received an HRA with no prior hip surgeries. Our hypothesis is that a prior hip arthroscopy will lead to inferior outcomes in patients undergoing HRA. 

Materials and Methods

This study is a retrospective, case-control study using a 1:2 matching analysis. Dr. Su performed all HRAs, which were enrolled in an institutional review board–approved arthroplasty registry. All HRAs were performed using the BHR System. 

The surgical technique for hip resurfacing arthroplasty has been described.¹ All procedures were performed via a posterior approach with the patient in the lateral decubitus position. All patients received a hybrid metal-on-metal hip resurfacing, with an uncemented acetabular component and cemented femoral component. Intraoperative anesthesia for all patients was performed via a combined spinal-epidural anesthetic, and an epidural patient-controlled analgesic was used for the first day postoperatively, followed by a transition to oral analgesics. The sizes of the acetabular and femoral components were recorded for each hip resurfacing. Postoperatively, intermittent pneumatic compression devices were placed upon arrival in the recovery room, and active ankle flexion and extension exercises were initiated immediately after the patient’s neurologic function returned.¹⁴ Aspirin was used for chemical deep venous thrombosis prophylaxis in all patients postoperatively for a period of 6 weeks. Full weight-bearing, with the use of crutches for assistance with balance, was permitted immediately. Crutches were used for a period of 3 weeks prior to being discontinued. 

From a database of 1357 HRAs (all BHR implants) performed between June 2006 and June 2012, 51 patients were identified who received an HRA after a prior hip arthroscopy. Eight patients were excluded because they did not possess adequate clinical documentation or were lost to follow-up. In the remaining 43 patients, there were 32 men and 11 women (21 right hips, 22 left hips), which formed the arthroscopy cohort. Two patients had a history of multiple hip arthroscopies (1 patient with 2 prior procedures, 1 patient with 3 prior procedures). The mean (SD) time from the most recent hip arthroscopy to the HRA was 2.5 (2.5) years. Table 1 presents a summary of the hip arthroscopy procedures (including only the most recent hip arthroscopy procedure in those with multiple arthroscopies).

Patient demographic variables (age, body mass index [BMI]) were recorded preoperatively, along with the Harris Hip Score (HHS),¹⁵ UCLA activity score,¹⁶ Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score,¹⁷ and preoperative hip range of motion (flexion, extension, abduction, adduction, internal rotation, and external rotation). The same clinical indices were assessed postoperatively along with the Short Form-12 (SF-12) Health Survey Score,¹⁸ at the 6-week, 3-month, 6-month, 1-year, and most recent follow-up visits.

Radiographic assessment consisted of a low anteroposterior (AP) pelvic radiograph (with the radiographic beam centered on the pubic symphysis) and a cross-table lateral radiograph obtained at the most recent follow-up visit. Both the acetabular component abduction relative to the inter-teardrop line, and the angle between the femoral stem and the anatomic axis of the femoral shaft (stem-shaft angle) were measured on AP radiographs.^19,20 Acetabular component anteversion was measured on the cross-table lateral radiographs as the angle between the projected long axis of the acetabular opening and a line drawn perpendicular to the long axis plane of the body (Figures A, B).²¹

The same registry database was used to identify patients who received an HRA without a prior history of arthroscopy or hip surgery. A 1:2 matching analysis for those patients with a prior hip arthroscopy to those without a prior hip arthroscopy was performed to formulate a control group (control cohort) of 86 patients. Each patient in the arthroscopy cohort was matched with 2 patients in the control cohort based on the following parameters: age (± 6 years), sex (same), BMI (± 4 kg/m²), femoral head size (± 4 mm), and preoperative HHS and WOMAC scores (± 7 points). In the event an arthroscopy patient matched to 2 or more control patients, the patients who minimized the least squared error among the matching variables were selected. 

Statistical Analysis

All data were collected and analyzed using Microsoft Excel software (Microsoft Corporation, Redmond, Washington). Statistical comparisons between the 2 cohorts regarding demographic variables, clinical outcomes, and radiographic alignment were performed using an unpaired, Student 2-tailed t test, with statistical significance set at P ≤ .05. 

Results

A comparison of the results of the 1:2 matching analysis between the arthroscopy and control cohorts is presented in Table 2. There was no significant difference in the preoperative age, BMI, femoral head size, HHS, or WOMAC score between the 2 cohorts. However, the control cohort did show a more severe, preoperative flexion contracture (as expressed by a decreased amount of extension) and a decreased amount of preoperative abduction (Table 3). The preoperative UCLA activity score was also decreased in the control cohort, but this was not statistically significant.

The mean (SD) follow-up was 2.0 (1.0) years in the arthroscopy cohort and 2.1 (1.1) years in the control cohort. There was no significant difference in radiographic alignment between the 2 cohorts. The stem-shaft angle was 139.3° (SD, 5.4°) in the arthroscopy cohort (vs 138.3° [SD, 5.5°] in the control cohort; P = .3), the acetabular abduction was 43.9° (SD, 5.8°) in the arthroscopy cohort (vs 42.9° [SD, 6.1°] in the control cohort; P = .4), and the acetabular anteversion was 21.1° (SD, 7.5°) in the arthroscopy cohort (vs 20.8° [SD, 7.1°] in the control cohort; P = .8). 

At 6-week follow-up, the arthroscopy cohort showed a significantly decreased WOMAC score compared with the control cohort (72.9 [SD, 15.5] vs 80.5 [SD, 11.8], respectively; P = .05). In addition, there was a trend towards a decreased SF-12 mental component score in the arthroscopy cohort (52.2 [SD, 9.3] vs 56.5 [SD, 7.8] in the control cohort; P = .06). However, none of the remaining clinical indices showed a significant difference between the 2 cohorts, and there was no difference in range of motion between the 2 cohorts at the 6-week follow-up visit (Table 4).

In addition, at 3-month follow-up, no statistically significant differences were seen between the 2 cohorts for any of the clinical indices or range of motion values. Both groups continued to improve rapidly, with HHS of 96.9 (SD, 3.5) in the arthroscopy cohort and 95.5 (SD, 6.6) in the control cohort, and WOMAC scores of 88.7 (SD, 10.2) and 89.5 (SD, 9.8), respectively (Table 5). Similarly, at the 6-month and 1-year follow-up intervals, the 2 cohorts showed continued improvement in their clinical measures, with no statistically significant differences between the 2 cohorts (Tables 6, 7). 

At the most recent follow-up visit, more than 1 year after surgery, the HHS was 99.5 (SD, 1.3) in the arthroscopy cohort and 99.2 (SD, 9.7) in the control cohort (P = .9), and the WOMAC score was 93.5 (SD, 11.3) and 92.4 (SD, 12.2), respectively (P = .8). No significant perioperative complications were seen in the arthroscopy cohort. In the arthroscopy cohort, 1 patient was diagnosed with a deep venous thrombosis 2 weeks after the procedure and was placed on low-molecular-weight heparin and coumadin for treatment. A second patient in the arthroscopy cohort had continued serosanguinous drainage for 4 days postoperatively, which resolved with continued compressive dressings. To date, no patients in the arthroscopy or control cohorts have required a second operation or revision of their components.

Discussion

Given the increasing prevalence of hip arthroscopies to treat multiple disorders of the native joint, it is important to assess the potential consequences of these procedures on future arthroplasties. Piedade and colleagues,¹³ in a retrospective review of 1474 primary total knee arthroplasties, showed a prior bony procedure (high tibial osteotomy, tibial plateau fracture, patellar realignment) to be predictive of decreased range of motion postoperatively. In addition, a prior knee arthroscopy was associated with a higher rate of postoperative complications, with 30% of the complications requiring a reoperation, and 8.3% of the complications requiring a revision total knee arthroplasty. Kaplan-Meier survival curves showed a survival rate of only 86.8% in those patients with a prior knee arthroscopy (vs 98.1% in those without a prior knee surgery).²² Therefore, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy. After the initial 6-week follow-up visit, no significant difference was seen in the functional outcomes between those patients with or without a history of prior hip arthroscopy who received an HRA.

After analysis of patient outcomes using multiple clinical measurement tools, at 6-week, 3-month, 6-month, 1-year, and most recent follow-up intervals, the only significant difference between the 2 cohorts was the WOMAC score at 6-week follow-up. Interestingly, there was no significant difference seen in the other clinical assessments, including the SF-12 score, HHS, range of motion, or UCLA activity score (although this did trend towards significance). This can be explained by the difference in both the mode of administration and various metrics assessed by these instruments. In comparison to the HHS evaluation, the patient completes the WOMAC (rather than the clinician) and also provides a more detailed assessment of symptoms, pain, stiffness, and activities of daily living.¹⁷ Therefore, this study suggests that patients with a prior hip arthroscopy may require more time to return to their activities of daily living after an HRA. However, whether the statistically significant difference between the 2 scores translates into a clinically significant difference can be questioned.

The clinical outcomes of this series of patients were excellent at the short-term follow-up, and both groups achieved clinical results comparable to prior reported results of HRA.^1,10,23,24 However, despite these results, there are several limitations to this study. First, longer-term follow-up is required to determine if any significant differences (such as aseptic loosening, infection, and prosthesis survival) are associated with a prior hip arthroscopy. In addition, this study included a relatively small cohort of patients who had a prior hip arthroscopy. However, a relatively large, single-surgeon database of 1357 HRAs was reviewed, with only 51 cases being reported (3.7%). With the increasing popularity of hip arthroscopy, the number of patients presenting for HRA will likely continue to increase. However, despite these limitations, this study shows that a prior hip arthroscopy does not appear to affect the short-term, clinical outcomes of a metal-on-metal HRA.

Metal-on-metal hip resurfacing arthroplasty (HRA) remains an alternative to total hip arthroplasty (THA) in appropriately selected, younger, active adults with degenerative hip disease.^1-4 While concerns remain regarding the potential for adverse local tissue reactions from wear of the metal-on-metal bearing surface,^5-8 10-year data from the Australian Orthopaedic Association National Joint Replacement Registry Annual Report⁹ showed a revision rate of only 6.3% when the Birmingham Hip Resurfacing (BHR) System was used (Smith & Nephew Inc, Memphis, Tennessee).In addition, in an independent review of 230 consecutive BHRs at a mean follow-up of 10.4 years, Coulter and colleagues¹⁰ showed encouraging clinical results, with a mean Oxford Hip Score of 45.0 and a mean University of California at Los Angeles (UCLA) activity score of 7.4.

Similar to the prior increase in popularity of HRA, hip arthroscopy has also become much more commonplace, and its indications continue to evolve.¹¹ Hip arthroscopy has been used in the native hip joint to manage femoroacetabular impingement, labral tears, and iliopsoas tendinopathy, among other conditions.¹² In addition, the use of hip arthroscopy has not been limited to the native hip but also has increased as a diagnostic and therapeutic procedure after hip arthroplasties. Bajwa and Villar¹² found hip arthroscopy to be diagnostic in 23 of 24 patients who underwent the procedure after a hip arthroplasty, concluding that arthroscopy is a useful adjunct in the diagnosis of symptomatic arthroplasties.

Therefore, hip arthroscopy has been shown to be an effective modality to treat pathology in both the native hip and after hip arthroplasties. However, the effect of a prior hip arthroscopy on the outcome of a subsequent metal-on-metal HRA has not been determined. Piedade and colleagues¹³ showed a prior knee arthroscopy to increase the risk of postoperative complications and subsequent revision after total knee arthroplasty. Complications included reflex sympathetic dystrophy, undiagnosed pain, infection, stiffness, and component loosening. A prior osteochondroplasty at the femoral head-neck junction could increase the risk of femoral neck fracture after a subsequent HRA. Thus, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy and to compare these results with a cohort of patients who received an HRA with no prior hip surgeries. Our hypothesis is that a prior hip arthroscopy will lead to inferior outcomes in patients undergoing HRA. 

Materials and Methods

This study is a retrospective, case-control study using a 1:2 matching analysis. Dr. Su performed all HRAs, which were enrolled in an institutional review board–approved arthroplasty registry. All HRAs were performed using the BHR System. 

The surgical technique for hip resurfacing arthroplasty has been described.¹ All procedures were performed via a posterior approach with the patient in the lateral decubitus position. All patients received a hybrid metal-on-metal hip resurfacing, with an uncemented acetabular component and cemented femoral component. Intraoperative anesthesia for all patients was performed via a combined spinal-epidural anesthetic, and an epidural patient-controlled analgesic was used for the first day postoperatively, followed by a transition to oral analgesics. The sizes of the acetabular and femoral components were recorded for each hip resurfacing. Postoperatively, intermittent pneumatic compression devices were placed upon arrival in the recovery room, and active ankle flexion and extension exercises were initiated immediately after the patient’s neurologic function returned.¹⁴ Aspirin was used for chemical deep venous thrombosis prophylaxis in all patients postoperatively for a period of 6 weeks. Full weight-bearing, with the use of crutches for assistance with balance, was permitted immediately. Crutches were used for a period of 3 weeks prior to being discontinued. 

From a database of 1357 HRAs (all BHR implants) performed between June 2006 and June 2012, 51 patients were identified who received an HRA after a prior hip arthroscopy. Eight patients were excluded because they did not possess adequate clinical documentation or were lost to follow-up. In the remaining 43 patients, there were 32 men and 11 women (21 right hips, 22 left hips), which formed the arthroscopy cohort. Two patients had a history of multiple hip arthroscopies (1 patient with 2 prior procedures, 1 patient with 3 prior procedures). The mean (SD) time from the most recent hip arthroscopy to the HRA was 2.5 (2.5) years. Table 1 presents a summary of the hip arthroscopy procedures (including only the most recent hip arthroscopy procedure in those with multiple arthroscopies).

Patient demographic variables (age, body mass index [BMI]) were recorded preoperatively, along with the Harris Hip Score (HHS),¹⁵ UCLA activity score,¹⁶ Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score,¹⁷ and preoperative hip range of motion (flexion, extension, abduction, adduction, internal rotation, and external rotation). The same clinical indices were assessed postoperatively along with the Short Form-12 (SF-12) Health Survey Score,¹⁸ at the 6-week, 3-month, 6-month, 1-year, and most recent follow-up visits.

Radiographic assessment consisted of a low anteroposterior (AP) pelvic radiograph (with the radiographic beam centered on the pubic symphysis) and a cross-table lateral radiograph obtained at the most recent follow-up visit. Both the acetabular component abduction relative to the inter-teardrop line, and the angle between the femoral stem and the anatomic axis of the femoral shaft (stem-shaft angle) were measured on AP radiographs.^19,20 Acetabular component anteversion was measured on the cross-table lateral radiographs as the angle between the projected long axis of the acetabular opening and a line drawn perpendicular to the long axis plane of the body (Figures A, B).²¹

The same registry database was used to identify patients who received an HRA without a prior history of arthroscopy or hip surgery. A 1:2 matching analysis for those patients with a prior hip arthroscopy to those without a prior hip arthroscopy was performed to formulate a control group (control cohort) of 86 patients. Each patient in the arthroscopy cohort was matched with 2 patients in the control cohort based on the following parameters: age (± 6 years), sex (same), BMI (± 4 kg/m²), femoral head size (± 4 mm), and preoperative HHS and WOMAC scores (± 7 points). In the event an arthroscopy patient matched to 2 or more control patients, the patients who minimized the least squared error among the matching variables were selected. 

Statistical Analysis

All data were collected and analyzed using Microsoft Excel software (Microsoft Corporation, Redmond, Washington). Statistical comparisons between the 2 cohorts regarding demographic variables, clinical outcomes, and radiographic alignment were performed using an unpaired, Student 2-tailed t test, with statistical significance set at P ≤ .05. 

Results

A comparison of the results of the 1:2 matching analysis between the arthroscopy and control cohorts is presented in Table 2. There was no significant difference in the preoperative age, BMI, femoral head size, HHS, or WOMAC score between the 2 cohorts. However, the control cohort did show a more severe, preoperative flexion contracture (as expressed by a decreased amount of extension) and a decreased amount of preoperative abduction (Table 3). The preoperative UCLA activity score was also decreased in the control cohort, but this was not statistically significant.

The mean (SD) follow-up was 2.0 (1.0) years in the arthroscopy cohort and 2.1 (1.1) years in the control cohort. There was no significant difference in radiographic alignment between the 2 cohorts. The stem-shaft angle was 139.3° (SD, 5.4°) in the arthroscopy cohort (vs 138.3° [SD, 5.5°] in the control cohort; P = .3), the acetabular abduction was 43.9° (SD, 5.8°) in the arthroscopy cohort (vs 42.9° [SD, 6.1°] in the control cohort; P = .4), and the acetabular anteversion was 21.1° (SD, 7.5°) in the arthroscopy cohort (vs 20.8° [SD, 7.1°] in the control cohort; P = .8). 

At 6-week follow-up, the arthroscopy cohort showed a significantly decreased WOMAC score compared with the control cohort (72.9 [SD, 15.5] vs 80.5 [SD, 11.8], respectively; P = .05). In addition, there was a trend towards a decreased SF-12 mental component score in the arthroscopy cohort (52.2 [SD, 9.3] vs 56.5 [SD, 7.8] in the control cohort; P = .06). However, none of the remaining clinical indices showed a significant difference between the 2 cohorts, and there was no difference in range of motion between the 2 cohorts at the 6-week follow-up visit (Table 4).

In addition, at 3-month follow-up, no statistically significant differences were seen between the 2 cohorts for any of the clinical indices or range of motion values. Both groups continued to improve rapidly, with HHS of 96.9 (SD, 3.5) in the arthroscopy cohort and 95.5 (SD, 6.6) in the control cohort, and WOMAC scores of 88.7 (SD, 10.2) and 89.5 (SD, 9.8), respectively (Table 5). Similarly, at the 6-month and 1-year follow-up intervals, the 2 cohorts showed continued improvement in their clinical measures, with no statistically significant differences between the 2 cohorts (Tables 6, 7). 

At the most recent follow-up visit, more than 1 year after surgery, the HHS was 99.5 (SD, 1.3) in the arthroscopy cohort and 99.2 (SD, 9.7) in the control cohort (P = .9), and the WOMAC score was 93.5 (SD, 11.3) and 92.4 (SD, 12.2), respectively (P = .8). No significant perioperative complications were seen in the arthroscopy cohort. In the arthroscopy cohort, 1 patient was diagnosed with a deep venous thrombosis 2 weeks after the procedure and was placed on low-molecular-weight heparin and coumadin for treatment. A second patient in the arthroscopy cohort had continued serosanguinous drainage for 4 days postoperatively, which resolved with continued compressive dressings. To date, no patients in the arthroscopy or control cohorts have required a second operation or revision of their components.

Discussion

Given the increasing prevalence of hip arthroscopies to treat multiple disorders of the native joint, it is important to assess the potential consequences of these procedures on future arthroplasties. Piedade and colleagues,¹³ in a retrospective review of 1474 primary total knee arthroplasties, showed a prior bony procedure (high tibial osteotomy, tibial plateau fracture, patellar realignment) to be predictive of decreased range of motion postoperatively. In addition, a prior knee arthroscopy was associated with a higher rate of postoperative complications, with 30% of the complications requiring a reoperation, and 8.3% of the complications requiring a revision total knee arthroplasty. Kaplan-Meier survival curves showed a survival rate of only 86.8% in those patients with a prior knee arthroscopy (vs 98.1% in those without a prior knee surgery).²² Therefore, the purpose of this study was to evaluate the clinical outcomes of a series of patients who received an HRA after a prior hip arthroscopy. After the initial 6-week follow-up visit, no significant difference was seen in the functional outcomes between those patients with or without a history of prior hip arthroscopy who received an HRA.

After analysis of patient outcomes using multiple clinical measurement tools, at 6-week, 3-month, 6-month, 1-year, and most recent follow-up intervals, the only significant difference between the 2 cohorts was the WOMAC score at 6-week follow-up. Interestingly, there was no significant difference seen in the other clinical assessments, including the SF-12 score, HHS, range of motion, or UCLA activity score (although this did trend towards significance). This can be explained by the difference in both the mode of administration and various metrics assessed by these instruments. In comparison to the HHS evaluation, the patient completes the WOMAC (rather than the clinician) and also provides a more detailed assessment of symptoms, pain, stiffness, and activities of daily living.¹⁷ Therefore, this study suggests that patients with a prior hip arthroscopy may require more time to return to their activities of daily living after an HRA. However, whether the statistically significant difference between the 2 scores translates into a clinically significant difference can be questioned.

The clinical outcomes of this series of patients were excellent at the short-term follow-up, and both groups achieved clinical results comparable to prior reported results of HRA.^1,10,23,24 However, despite these results, there are several limitations to this study. First, longer-term follow-up is required to determine if any significant differences (such as aseptic loosening, infection, and prosthesis survival) are associated with a prior hip arthroscopy. In addition, this study included a relatively small cohort of patients who had a prior hip arthroscopy. However, a relatively large, single-surgeon database of 1357 HRAs was reviewed, with only 51 cases being reported (3.7%). With the increasing popularity of hip arthroscopy, the number of patients presenting for HRA will likely continue to increase. However, despite these limitations, this study shows that a prior hip arthroscopy does not appear to affect the short-term, clinical outcomes of a metal-on-metal HRA.