Surgical Hip Dislocation in the Era of Hip Arthroscopy Demonstrates High Survivorship and Improvements in Patient-reported Outcomes for Complex Femoroacetabular Impingement

Abstract

Background

Over the past decade, hip arthroscopy has become more commonly used in the treatment of patients with femoroacetabular impingement (FAI) as well as those with many other intra-articular hip pathologies. As such, the indications for open surgical hip dislocation have narrowed to include complex intra-articular and extra-articular bony morphologies and revision hip preservation. Although surgical hip dislocation has been established as an effective approach, previous research using contemporary indications has been limited primarily to smaller cohorts with short-term follow-up. Furthermore, factors associated with reoperation, conversion arthroplasty, and not achieving clinical improvement remain ambiguous.

Questions/purposes

At a minimum of 2 years of follow-up in patients treated for FAI with surgical hip dislocation, (1) how much did patient-reported outcome measures (PROMs) improve, and what percentage of patients achieved the minimum clinically important difference (MCID) and patient-acceptable symptom state (PASS)? (2) What was the survivorship free from reoperation of the hip for any reason? (3) What was the survivorship free from conversion to arthroplasty (THA or hip resurfacing)?

Methods

Between February 2011 and April 2021, 211 patients were treated at one academic institution with surgical dislocation and osteoplasty of the femoral head-neck junction or greater trochanter for FAI. Of these, patients with concomitant diagnoses including Legg-Calvé-Perthes disease, slipped capital femoral epiphysis, osteochondritis dissecans, a benign tumor, or another neurologic or metabolic disorder were excluded. This yielded 74% (156 of 211) of patients as potentially eligible. A further 12% (18 of 156) of patients without postoperative PROMs were excluded, and another 13% (21 of 156) were lost before the minimum study follow-up of 2 years, leaving 75% (117) for analysis at a median of 4.9 years (range 2 to 12 years) in this retrospective study. A total of 88% of the patients were women (103 of 117), and 40% (47 patients) underwent the procedure as a revision. All patients underwent surgical hip dislocation. Intraoperatively, 66% (77 of 117) of patients demonstrated cam impingement, and the remainder had either pincer impingement or combined cam and pincer; 82% (96 of 117) had some component of extra-articular impingement. To answer our first research question, we obtained PROM scores for the modified Harris hip score (mHHS) and international hip outcome tool 12 (iHOT-12) before surgery and at the most-recent follow-up from our longitudinally maintained institutional database, and we determined the percentage of patients who achieved the MCID or PASS. To answer our second research question, we calculated Kaplan-Meier survivorship free from any reoperation on the hip at 5.8 years (mean follow-up) after the index procedure. To answer our third research question, we calculated the Kaplan-Meier survivorship free from conversion to arthroplasty at 5.8 years. Using univariate logistic regression analysis, factors associated with meeting these defined endpoints were identified.

Results

At a mean of 6 ± 3 years, the mHHS of patients treated with surgical hip dislocation for FAI who had not had previous surgery on the hip improved from 53 ± 15 to 85 ± 13 (mean difference 31 [95% confidence interval 27 to 36]; p < 0.01) and the mean iHOT-12 score improved from 25 ± 14 to 74 ± 24 (mean difference 49 [95% CI 42 to 56]; p < 0.01). At a mean of 5 ± 3 years, for patients with prior hip surgery, mHHS improved from 55 ± 13 to 79 ± 15 (mean difference 24 [95% CI 18 to 30]; p < 0.01) and iHOT-12 scores improved from 25 ± 16 to 66 ± 26 (mean difference 41 [95% CI 31 to 50]; p < 0.01). The proportion of patients who achieved the MCID for the mHHS was 87% (55 of 63) and 83% (33 of 40), while the proportion who achieved the PASS for the mHHS was 66% (46 of 70) and 51% (24 of 47) for primary and revision surgery, respectively. The proportion of patients who achieved the MCID for the iHOT-12 was 87% (53 of 61) and 83% (33 of 40), while the proportion who achieved the PASS for iHOT-12 was 68% (47 of 69) and 57% (26 of 46) for primary and revision surgery, respectively. Higher lateral center-edge angles (odds ratio 1.2 per 1° [95% CI 1.1 to 1.3]; p < 0.01), a lower acetabular version at 2:00 (OR 0.92 per 1° [95% CI 0.85 to 0.99]; p = 0.05), the absence of a labral tear (OR 0.21 [95% CI 0.06 to 0.79]; p = 0.02), a higher preoperative mHHS (OR 1.1 per 1 point [95% CI 1.02 to 1.2]; p < 0.01), and a higher preoperative iHOT-12 score (OR 1.1 per 1 point [95% CI 1.01 to 1.1]; p = 0.01) were associated with not achieving the MCID. In patients without a history of ipsilateral hip surgery, survivorship free from reoperation at 5.8 years after surgical hip dislocation was 90% (95% CI 82% to 98%). In the setting of prior ipsilateral hip surgery, survivorship free from reoperation at 5.8 years after surgical hip dislocation was 75% (95% CI 60% to 95%). No patients for whom surgical hip dislocation was their primary hip procedure underwent conversion to arthroplasty. Among the revision patient cohort, survivorship free from conversion to arthroplasty at 5.8 years after surgical hip dislocation was 97% (95% CI 91% to 100%).

Conclusion

Surgical hip dislocation effectively improves hip pain and function with high survivorship (free of reoperation or conversion to arthroplasty) for complex intra-articular and extra-articular FAI that is deemed less suitable for arthroscopic treatment, but when the procedure is used in patients who have had prior hip surgery, they should be told that revision surgery is a realistic possibility. If insufficient acetabular coverage is anticipated, acetabular reorientation osteotomy should be considered. Future studies should evaluate long-term survivorship and PROMs in larger patient cohorts, as well as identify other factors associated with reoperation and conversion to arthroplasty.

Level of Evidence

Level III, therapeutic study.

Introduction

Femoroacetabular impingement (FAI) is associated with hip pain and the early development of osteoarthritis [3, 10, 13, 17, 39, 40]. Over the past decade, technical advances have facilitated less-invasive arthroscopic correction of proximal femoral and acetabular bony morphologies associated with FAI, including cam, pincer, and subspine impingement [32]. This approach often improves pain and function, may slow the degenerative process, and is associated with relatively few serious complications [14-16, 19, 25]. When compared directly, hip arthroscopy and open surgical hip dislocation for FAI demonstrate similar results [22, 28]. Although the efficacy of hip arthroscopy is well established, there is a subset of patients for whom arthroscopic intervention may be less applicable as an approach. Allowing for full, dynamic evaluation of the femoral head and acetabulum, surgical hip dislocation is preferred for more extensive correction of bony morphologies including global acetabular overcoverage, posterior impingement, bony dysmorphology of the proximal femur or acetabulum with concurrent femoral version abnormalities, or for patients with a history of hip arthroscopy undergoing further correction. Additionally, extra-articular impingement (a less common cause of FAI, where regions of the proximal femur outside the hip capsule impinge with the pelvis) including ischiofemoral impingement, greater trochanter-pelvis impingement, or subspine impingement can be diagnosed and addressed with surgical hip dislocation.

Because these conditions are uncommon, and because arthroscopic approaches are seeing wider use, research on survivorship and patient-reported outcome measures (PROMs) after surgical hip dislocation to treat FAI in the era of hip arthroscopy remains limited. Previous studies have shown short-term improvements in pain and function [1, 8, 11, 27, 29]; however, few have evaluated PROMs and survivorship at a follow-up interval beyond 3 years [2, 4, 20, 21, 22, 35, 36]. Because only two of these studies [21, 22] examined more than 100 patients, further investigation is required to determine postoperative satisfaction and survivorship that can be generalized to the contemporary patient population. Furthermore, these studies did not assess the role of surgical hip dislocation specifically in the revision setting.

We therefore asked: At a minimum of 2 years of follow-up in patients treated for FAI with surgical hip dislocation, (1) how much did PROM scores improve, and what percentage of patients achieved the minimum clinically important difference (MCID) and patient-acceptable symptom state (PASS)? (2) What was the survivorship free from reoperation of the hip for any reason? (3) What was the survivorship free from conversion to arthroplasty (THA or hip resurfacing)?

Patients and Methods

Study Design and Setting

This was a retrospective study drawn from a longitudinally maintained, single-center hip preservation registry. It was performed at an urban, tertiary academic medical center with the approval of the local steering committee. Patients were treated by one surgeon as part of a referral hip preservation practice.

Patients

Between February 2011 and April 2021, we treated 211 patients with surgical dislocation and osteoplasty of the femoral head-neck junction or regions of the greater trochanter for intra-articular or extra-articular FAI. Of these, patients with concomitant diagnoses including Legg-Calvé-Perthes disease, slipped capital femoral epiphysis, osteochondritis dissecans, a benign tumor, or another neurologic or metabolic disorder were excluded. This yielded 74% (156 of 211) of patients as potentially eligible. A further 12% (18 of 156) of patients without postoperative PROMs were excluded and another 13% (21 of 156) were lost before the minimum study follow-up of 2 years, leaving 75% (117 patients) for analysis in this retrospective study at a median of 4.9 years (range 2 to 12 years) (Fig. 1). For patients who underwent bilateral surgery, only the first hip was included. Data from 24% (28 of 117) of these patients have been published elsewhere [29], but these patients were included in the current study because they had longer durations of follow-up.

Fig. 1.

This STROBE study flow diagram shows patients who were lost to follow-up and excluded from this study.

Indications for surgical hip dislocation (as opposed to arthroscopic treatments for FAI) during this period included a subtle cam lesion with signs of impingement on examination such as poor hip flexion (less than 90°) and rotation (internal rotation less than 30° and external rotation less than 45° with hip flexion), an examination finding consistent with extra-articular impingement (greater trochanter-pelvis impingement, subspine or anterior inferior iliac spine impingement, or ischiofemoral impingement), FAI with concurrent femoral version abnormalities, posterior or global impingement, and persistent impingement-related hip pain after a prior hip preservation procedure (most often hip arthroscopy). Although treatment indications remained consistent over the study period, during the 6 years before the study was initiated, patients with more complex conditions and histories were presented at a multidisciplinary hip preservation conference consisting of hip arthroscopists, physiatrists, joint replacement specialists, and open hip preservation surgeons to collectively decide on appropriate surgical interventions. With the conference to aid in determining the surgical treatment, revision surgery represented a larger proportion of surgical hip dislocations for FAI (50% of patients from 2015 to 2021 versus 34% in 2011 to 2014).

Descriptive Data

The total patient cohort (117 patients, mean ± standard deviation follow-up 5.8 ± 3.3 years) included 70 patients without a history of an ipsilateral hip procedure and 47 patients for whom the surgical dislocation was a revision procedure (Table 1). When stratified by the absence or presence of a prior hip surgery, 73% (51 of 70) and 85% (40 of 47) of patients presented with intra-articular and extra-articular impingement for the primary and revision subgroups, respectively. In the revision cohort, 96% (45 of 47) of patients had idiopathic intra-articular or extra-articular impingement that was either inadequately corrected or not addressed by prior surgery. Of these, 9% (4 of 45) of patients had additional iatrogenic FAI related to scar tissue or material from prior labral or capsular repair, while 2% (1 of 45) had additional post-traumatic deformities from prior acetabular fractures. One patient had isolated iatrogenic FAI from prior ablation of an osteoid osteoma, and one patient had chondral changes and a valgus deformity at the head and neck after a femoral neck fracture. Eleven patients undergoing revision (23% [11 of 47]) had borderline dysplasia as defined by a lateral center-edge angle (LCEA) of 18° to 25° [38]. The most common prior ipsilateral surgery was hip arthroscopy (85% [40 of 47]), with 28% (13 of 47) of patients having undergone more than one arthroscopy. Other prior surgeries included periacetabular osteotomy (6% [3 of 47]), hip pinning (6% [3 of 47]), tumor or mass excision (4% [2 of 47]), and closed hip reduction with tenotomy (2% [1 of 47]).

Table 1.

Characteristics of included patients

Characteristic	Primary surgery (n = 70)	Revision surgery (n = 47)
Demographics
Median age in years	22 (range 14 to 46)	26 (range 14 to 47)
Median follow-up in years	5 (range 2 to 12)	5 (range 2 to 12)
BMI in kg/m2	22 ± 3	23 ± 4
Women	89% (62 of 70)	87% (41 of 47)
Laterality (right)	51% (36 of 70)	55% (26 of 47)
Impingement type
Intra-articular impingement	94% (66 of 70)	98% (46 of 47)
Isolated cam	60% (42 of 70)	74% (35 of 47)
Isolated pincer	1% (1 of 70)	0% (0 of 47)
Mixed	33% (23 of 70)	23% (11 of 47)
Extra-articular Impingement	79% (55 of 70)	87% (41 of 47)
Greater trochanteric pelvic impingement	71% (50 of 70)	87% (41 of 47)
Subspine/AIIS impingement	13% (9 of 70)	6% (3 of 47)
Ischiofemoral impingement	1% (1 of 70)	2% (1 of 47)
Radiographic parameters
Plain radiograph	n = 70	n = 46
Lateral center edge angle in °	32 ± 7	30 ± 6
Anterior center edge in °	37 ± 9	36 ± 10
Tönnis grade
0	71% (50 of 70)	46% (21 of 46)
1	27% (19 of 70)	52% (24 of 46)
2	1% (1 of 70)	2% (1 of 46)
3	0% (0 of 70)	0% (0 of 46)
CT	n = 55	n = 34
Alpha angle in °	56 ± 11	57 ± 12
Normal (< 60°)	64% (35 of 55)	55% (18 of 33)
Elevated (> 60°)	36% (20 of 55)	45% (15 of 33)
Femoral version in °	13 ± 14	19 ± 11
Normal (5° to 20°)	38% (21 of 55)	41% (14 of 34)
Anteverted (> 20°)	27% (15 of 55)	50% (17 of 34)
Retroverted (< 5°)	35% (19 of 55)	9% (3 of 34)
Acetabular version 1:00 in °	4 ± 12	7 ± 10
Normal (0° to 10°)	35% (17 of 48)	41% (13 of 32)
Anteverted (> 10°)	27% (13 of 48)	37% (12 of 32)
Retroverted (< 0°)	38% (18 of 48)	22% (7 of 32)
Acetabular version 2:00 in °	11 ± 10	14 ± 10
Normal (10° to 15°)	19% (10 of 53)	25% (8 of 32)
Anteverted (> 15°)	36% (19 of 53)	47% (15 of 32)
Retroverted (< 10°)	45% (24 of 53)	28% (9 of 32)
Acetabular version 3:00 in °	16 ± 7	20 ± 8
Normal (15° to 20°)	36% (19 of 53)	44% (14 of 32)
Anteverted (> 20°)	26% (14 of 53)	37% (12 of 32)
Retroverted (< 15°)	38% (20 of 53)	19% (6 of 32)
Femoral neck-shaft angle in °	135 ± 5	133 ± 7
Normal (126° to 139°)	74% (14 of 19)	68% (13 of 19)
Coxa valga (> 139°)	21% (4 of 19)	16% (3 of 19)
Coxa vara (< 126°)	5% (1 of 19)	16% (3 of 19)
MRI	n = 67	n = 45
Labral tear	85% (57 of 67)	44% (20 of 45)

Data presented as mean ± SD or % (n) unless noted otherwise.

Surgical Procedures

All patients underwent surgical hip dislocation as described by Ganz et al. [9]. Concomitant procedures included femoral osteochondroplasty for cam-type impingement (95% [111 of 117]), osteoplasty of the anterior facet of the greater trochanter (53% [62 of 117]) and osteoplasty of the posterior facet of the greater trochanter (45% [53 of 117]) for greater trochanter-pelvis impingement, labral repair (41% [48 of 117]), acetabular rim trimming (25% [29 of 117]), scar debridement (15% [17 of 117]), labral debridement (15% [17 of 117]), relative neck lengthening for coxa breva and varus neck-shaft angle with a high-riding greater trochanter (11% [13 of 117]), subspine or anterior inferior iliac spine decompression for subspine impingement (10% [12 of 117]), labral reconstruction for a deficient labrum (7% [8 of 117]), lesser trochanteric resection for ischiofemoral impingement (2% [2 of 117]), acetabular microfracture (2% [2 of 117]), removal of heterotopic ossification (2% [2 of 117]), femoral osteotomy (1% [1 of 117]), and periacetabular osteotomy for concurrent hip dysplasia (1% [1 of 117]).

Data and Data Sources

Demographic, radiographic, and surgical data are regularly recorded by the senior author and surgeon (ES). These data are subsequently entered into the electronic medical record. Hip-specific PROMs are administered annually per registry protocol beginning at 1 year of follow-up through the electronic medical record or by a member of the hip preservation team (ZT). Data are automatically queried from the electronic medical record and transferred into the registry, which were extracted by a member of the hip preservation team (ZT). Factors included age, gender, BMI, presence of a psychiatric disorder, primary diagnosis, concurrent procedures, and any prior or subsequent hip or pelvis surgical procedures. Preoperative radiographic measurements were collected, including the LCEA, anterior center-edge angle (ACEA), and Tönnis grade. Alpha angle; femoral version; acetabular version at 1:00, 2:00, and 3:00; and femoral neck-shaft angle were recorded from routinely obtained preoperative 3D CT images. Labral tears, if present on MRI, were noted.

Endpoints

Hip-specific PROMs included the modified Harris hip score (mHHS) and International Hip Outcome Tool 12 (iHOT-12). For pediatric patients, only applicable iHOT-12 questions were scored. For each PROM, higher scores on a scale of 0 to 100 indicated pain and functional improvement. The MCID was defined as an outcome change ≥ 8 for the mHHS and ≥ 13 for iHOT-12. This mHHS threshold has been used to evaluate surgical hip dislocation for extra-articular impingement and was calculated using anchor-based methods for revision hip arthroscopy [26, 29]. To our knowledge, the MCID for the iHOT-12 has not been defined for surgical hip dislocation; thus, thresholds defined for hip arthroscopy for FAI in a sample of predominantly women were applied [18, 24]. The PASS was evaluated against anchor-based, time-dependent scores defined for the arthroscopic treatment of FAI: 83.3 and 83.6, and 72.2 and 74.3 at the 2-year and 5-year timepoints for the mHHS and iHOT-12, respectively [23]. Because PASS values have not been defined for surgical hip dislocation, to our knowledge, anchor-based values for the midterm treatment of FAI in a sample of predominantly women were used [23].

Primary and Secondary Study Outcomes

Our primary study goal was to define improvements in patients undergoing surgical hip dislocation for FAI. To answer the first research question, we collected outcome scores for the mHHS and iHOT-12 preoperatively and at the most-recent follow-up from our longitudinally maintained institutional database. The mean improvement and percentage of patients who achieved the MCID or PASS were determined.

Our secondary goals were to report survivorship free from reoperation and survivorship free from conversion to arthroplasty. To answer the second research question, we calculated Kaplan-Meier survivorship free from any reoperation (excluding routine hardware removal) at 5.8 years (mean follow-up) after the index procedure. To answer the third research question, we calculated Kaplan-Meier survivorship free from conversion to arthroplasty (THA or hip resurfacing) at 5.8 years. Finally, to determine variables associated with negative endpoints (not meeting the MCID or PASS, reoperation, and conversion to arthroplasty), univariate logistic regression analyses were used for recorded data and endpoints.

Ethical Approval

This study received approval from the hip preservation steering committee to use data from the Hospital for Special Surgery Institutional Review Board–approved, longitudinally maintained hip preservation registry (#2014-044-CR9).

Statistical Analysis

Normality was assessed using Shapiro-Wilk tests. The mean improvement in PROM scores was calculated using the independent-sample t-test or Wilcoxon Mann-Whitney test, where appropriate. Descriptive analyses (frequency and percentage) are used to report the percentage of patients achieving the MCID and PASS at the most-recent follow-up interval. Categorical variables were analyzed using the chi-square or Fisher exact test, where appropriate. Survivorship was determined using a Kaplan-Meier survival analysis. Variables associated with negative endpoints were analyzed through univariate logistic regression. Significance was defined as p values less than 0.05.

Results

PROMs

At a mean of 6 ± 3 years postoperatively in patients without prior hip surgery, the mHHS of those treated with surgical hip dislocation for FAI improved from 53 ± 15 to 85 ± 13 (mean difference 31 [95% confidence interval 27 to 36]; p < 0.01) and the mean iHOT-12 score improved from 25 ± 14 to 74 ± 24 (mean difference 49 [95% CI 42 to 56]; p < 0.01). At a mean of 5 ± 3 years postoperatively, in patients with prior hip surgery, the mHHS improved from 55 ± 13 to 79 ± 15 (mean difference 24 [95% CI 18 to 30]; p < 0.01) and iHOT-12 score improved from 25 ± 16 to 66 ± 26 (mean difference 41 [95% CI 31 to 50]; p < 0.01).

The proportion of patients who achieved the MCID for the mHHS was 87% (55 of 63) and 83% (33 of 40) for primary and revision surgery, respectively, while the proportion who achieved the PASS for mHHS was 66% (46 of 70) and 51% (24 of 47), respectively. The proportion of patients who achieved the MCID for the iHOT-12 was 87% (53 of 61) and 83% (33 of 40) for primary and revision surgery, respectively, while the proportion who achieved the PASS for the iHOT-12 was 68% (47 of 69) and 57% (26 of 46), respectively.

Factors associated with not achieving the MCID were a higher LCEA (odds ratio 1.2 per 1° [95% CI 1.1 to 1.3]; p < 0.01), lower acetabular version at 2:00 (OR 0.92 per 1° [95% CI 0.85 to 0.99]; p = 0.05), the absence of a labral tear (OR 0.21 [95% CI 0.06 to 0.79]; p = 0.02), a higher preoperative mHHS (OR 1.1 per 1 point [95% CI 1.02 to 1.2]; p < 0.01), and a higher preoperative iHOT-12 score (OR 1.1 per 1 point [95% CI 1.01 to 1.1]; p = 0.01).

Factors associated with not achieving the PASS were a lower preoperative iHOT-12 score (OR 0.96 per 1 point [95% CI 0.93 to 0.99]; p = 0.02).

Survivorship Free From Reoperation

Survivorship free from reoperation in the hip at 5.8 years after surgical hip dislocation for FAI in patients who had no prior ipsilateral surgery was 90% (95% CI 82% to 98%) (Fig. 2). In the setting of prior ipsilateral hip surgery, survivorship free from reoperation at 5.8 years after surgical hip dislocation was 75% (95% CI 60% to 95%) (Fig. 3). Among the 117 included patients, 14% (16 of 117) underwent reoperations. Revision hip arthroscopy was the most common procedure, performed in six patients, four of whom had persistent bony impingement, one had trochanteric bursitis, and one had capsular adhesions. Periacetabular osteotomy was the next-most-common reoperation, performed in four patients, of whom two had iatrogenic instability from prior arthroscopic rim resection, one had idiopathic acetabular retroversion, and one had trauma-induced instability. Additional procedures included open reduction and internal fixation for two patients with trochanteric nonunion, femoral osteotomy for one patient with unaddressed femoral retroversion, combined hip arthroscopy and periacetabular osteotomy for one patient with iatrogenic instability and capsular adhesions, and open excision of heterotopic ossification with repair of a gluteus medius tear for one patient. Reoperation occurred 0.2 to 10 years after surgical hip dislocation. No variables we studied were associated with subsequent reoperation.

Fig. 2.

The Kaplan-Meier survival analysis of patients without prior hip surgery is shown with reoperation as the endpoint. Survivorship is shown with 95% confidence intervals in the shaded area.

Fig. 3.

The Kaplan-Meier survival analysis of patients with prior hip surgery is shown with reoperation as the endpoint. Survivorship is shown with 95% confidence intervals in the shaded area.

Survivorship Free From Conversion to THA

No patient in the primary subgroup underwent conversion arthroplasty. Among the revision patient cohort, survivorship free from conversion to arthroplasty at 5.8 years was 97% (95% CI 91% to 100%) (Fig. 4). Among all patients, 3% (3 of 117) underwent conversion to arthroplasty (two THAa and one hip resurfacing). The only factor associated with conversion to arthroplasty was Tönnis Grade 2 changes (OR 70 [95% CI 2 to 2103]; p = 0.01). Conversion arthroplasty occurred 3 to 11 years after surgical hip dislocation.

Fig. 4.

The Kaplan-Meier survival analysis of patients with prior hip surgery is shown with conversion to arthroplasty as the endpoint. Survivorship is shown with 95% confidence intervals in the shaded area.

Discussion

FAI is among the leading causes of joint degeneration in the pre-arthritic hip [3, 10, 13, 17, 39, 40]. Although technical advances in hip arthroscopy have resulted in its wider use for the treatment of FAI, there is a subset of patients for whom arthroscopic intervention may be inadequate. With the advantage of circumferential dynamic evaluation of the femoral head and acetabulum, surgical hip dislocation in the era of hip arthroscopy is preferred for more-severe intra-articular and extra-articular impingement as well as in revision hip preservation [5, 33, 41]. Given the paucity of evidence examining these contemporary indications, it is important to define PROM score improvement and survivorship after surgical hip dislocation as the approach to these complex diagnoses. We found that surgical hip dislocation as a primary procedure provides improvements in pain and function, as well as excellent survivorship free from reoperation and conversion to arthroplasty. As a revision procedure, surgical hip dislocation provides comparable improvements in pain and function, with lower survivorship. In the contemporary management of FAI, these results suggest that surgical hip dislocation is a relevant approach in hip preservation for complex intra-articular and extra-articular impingement that would be incompletely addressed with hip arthroscopy, but patients undergoing surgical hip dislocation as a revision procedure should have their expectations managed thoughtfully because of the increased risk of reoperation.

Limitations

Selection bias was a factor here, in that it is likely that patients with more challenging clinical conditions were treated with surgical hip dislocation during this time; more straightforward morphologies tended to be managed arthroscopically. In a study by Nepple et al. [22], an unmatched surgical dislocation cohort was younger, more frequently involved in competitive sports, had higher rates of pincer and combined FAI, and had a higher percentage of Tönnis Grade 2 changes than the hip arthroscopy cohort. Younger, physically active patients with complex bony morphologies may have symptoms with increased severity or psychosocial factors affecting PROM scores and willingness to undergo revision procedures or conversion to arthroplasty. As such, MCID and PASS thresholds for hip arthroscopy may not accurately reflect the improvement expectations of patients who undergo surgical hip dislocation. However, because these values have not yet been defined specifically for surgical hip dislocation, cutoffs were identified from studies of comparable patients with hip arthroscopy. Furthermore, to mitigate the confounding effect of psychosocial factors, the presence of a psychologic disorder was recorded and analyzed. A younger age may have contributed to a preference toward revision hip preservation after surgical hip dislocation as opposed to conversion arthroplasty.

Second, transfer bias was a concern here, because 25% (39 of 156) of potentially eligible patients were excluded from the analysis either because of missing postoperative PROMs or lack of 2-year minimum follow-up. Of these, 46% (18 of 39) were patients undergoing revision surgery after a prior hip arthroscopy. In general, the results reported here may overestimate treatment benefit and survivorship, because missing patients are usually not doing as well as those who are accounted for. The lower ranges of the 95% CIs on our survivorship curves, and not just the point estimates, should therefore be considered realistic. Indications of excluded patients did not differ from those who were included, however, providing some confidence that the 117 patients represent those undergoing surgical hip dislocation for FAI during this period.

PROMs

Analysis of PROMs demonstrated improvement in mHHS and iHOT-12 scores along with a high likelihood of achieving the MCID after surgical hip dislocation as a primary or revision procedure for FAI. This suggests the surgical hip dislocation approach is effective in improving pain and function for patients with a combination of intra-articular and extra-articular impingement symptoms that are not treatable by hip arthroscopy. Factors associated with not meeting the MCID underline the importance of postoperative acetabular coverage, as described by Steppacher et al. [35, 36]. A higher LCEA and ACEA should cause the surgeon to consider rim trimming, and if undercorrected, could lead to persistent pincer impingement. A lower acetabular version on CT may indicate pincer impingement caused by retroversion in addition to deficient posterior coverage. Preoperative radiographs and 3D CT images can assist the surgical team in delineating symptom etiology between impingement and instability to determine intraoperative rim trimming or acetabular correction through concurrent periacetabular osteotomy. The use of image analysis software intraoperatively may be useful to quantify the extent of rim resection to achieve the optimal acetabular coverage and impingement-free ROM [7]. The association between the absence of a labral tear and not achieving the MCID agrees with Naal et al. [21], suggesting that those with labral injury benefit more from surgical hip dislocation than those without. Open management of FAI facilitates the treatment of labral injury, including refixation, which has resulted in positive short-term patient-reported outcomes [8]. The mHHS and iHOT-12 have been shown to demonstrate a ceiling effect, suggesting that higher preoperative scores have a decreased potential for outcome change [12, 30, 37]. In the current study, a lower preoperative score was the only factor associated with not achieving the PASS. Intuitively, patients with a lower iHOT-12 score will require a larger outcome change to meet the PASS. Furthermore, the lack of association between the mHHS and achievement of the PASS suggests the iHOT-12 might be a more accurate measure of satisfaction in this patient population.

The proportions of patients achieving the MCID across the cohort and achievement of PASS in the primary surgery subgroup were similar to those seen in patients undergoing primary hip arthroscopy for FAI [23]. Achievement of the PASS in the revision subgroup was lower, however. At a mean of 5 years of follow-up, Naal et al. [21] reported that 82% of patients were satisfied after surgical hip dislocation, using subjective satisfaction questions. Additionally, in a cohort undergoing revision hip arthroscopy, fewer than 65% of patients achieved the MCID for the mHHS and iHOT-33, which was less than in the revision surgery subgroup observed in the current study. Although this may further support the use of surgical hip dislocation over hip arthroscopy in a revision scenario, evidence detailing patient-reported outcomes after revision hip arthroscopy is sparse [34]. Defining the MCID and PASS using PROMs in this particular patient population would provide a better understanding of clinical improvement and satisfaction after surgical hip dislocation. Patients with more-complex femoroacetabular morphologies or surgical histories undergoing surgical hip dislocation may experience increased pain severity because they may have multiple sources of impingement. Thus, clinical thresholds that consider patient-specific preoperative expectations and baseline pain and functional status should be explicitly defined.

Survivorship Free From Reoperation

As a primary procedure for FAI, surgical hip dislocation demonstrated excellent survivorship (90% [95% CI 82% to 98%]) free from reoperation at 5.8 years after the index procedure. As a revision procedure for FAI, survivorship was slightly lower (75% [95% CI 60% to 95%]). Consistent with previous evidence [2, 21, 29], no preoperative variables or intraoperative procedures were associated with reoperation. Future research using larger patient samples is needed to elucidate factors predisposing the need for reoperation in this specific cohort. The distribution of revision procedures suggests that hip arthroscopy and periacetabular osteotomy are used to treat the most common persistent sources of hip pain after surgical hip dislocation. Previous research has implicated inadequate bony correction as one of the most common causes of revision after hip preservation surgery [6, 31]. In the current study, residual bony impingement was an indication for revision hip arthroscopy. Five patients younger than 35 years at the time of surgical hip dislocation underwent subsequent periacetabular osteotomy for three instances of iatrogenic instability after arthroscopic rim resection, idiopathic acetabular retroversion, and trauma-induced anterior instability. Unrecognized or untreated dysplasia has been described as a risk factor for further procedures after surgical hip dislocation [20]. In the current study, however, no patients undergoing reoperation had radiographic evidence of acetabular dysplasia severe enough to warrant concurrent periacetabular osteotomy. For young, active patients with prior rim alterations or concurrent borderline dysplasia, preoperative counseling before surgical hip dislocation might include the possibility of subsequent periacetabular osteotomy if hip pain persists.

Survivorship Free From Conversion to THA

As both a primary and revision procedure, surgical hip dislocation for FAI demonstrated excellent survivorship free from conversion arthroplasty at 5.8 years after the index procedure. The only preoperative factor associated with conversion arthroplasty was Tönnis Grade 2 changes, reiterating the notion that hips with this magnitude of radiographic arthritic change at the time of surgery are at the greatest risk of undergoing arthroplasty [4, 20, 35, 36]. For patients with degenerative changes and a prior hip surgery, surgical hip dislocation could be recommended to delay rather than entirely prevent a conversion arthroplasty procedure. Although no patient in the primary subgroup underwent a subsequent arthroplasty, longer follow-up will be needed to determine the extent to which surgical hip dislocation as a primary procedure for FAI prevents or delays conversion arthroplasty.

Conclusion

Improvements in patient-reported outcomes along with high survivorship free from reoperation and conversion arthroplasty were observed at a mean of 6 years after primary surgical hip dislocation for FAI. Surgical hip dislocation as a revision procedure led to similar improvements in patient-reported pain and functional outcomes, with slightly lower survivorship. Factors associated with not meeting the MCID emphasize the importance of sufficient acetabular coverage in managing postoperative pain and function when deciding on osseous correction and concurrent intraoperative procedures. Realistic patient expectations of survivorship after surgical hip dislocation as a revision procedure should be set, because revision is a possibility in these patients. In the era of hip arthroscopy, surgical hip dislocation is an effective approach for complex intra-articular and extra-articular FAI that is unsuitable for arthroscopic intervention and as a revision preservation procedure. Future studies should evaluate long-term survivorship and PROMs in larger patient cohorts, as well as identify factors associated with reoperation and conversion to arthroplasty.