Prior hip arthroscopy impacts long-term outcomes of total hip arthroplasty

Abstract

Aims

The objective of this study was to perform a long-term comparative analysis of patients who underwent total hip arthroplasty (THA) with a history of previous ipsilateral hip arthroscopy (PA) to a propensity-score matched control group of primary THA with no prior hip arthroscopy (NPA).

Methods

Data were analyzed from patients who underwent primary THA for symptomatic hip osteoarthritis between November 2010 and November 2013. Patients included had completed a minimum of ten years of patient-reported outcome measure questionnaires. The PA group was propensity-score matched 1:1 based on age at THA, BMI, sex, robotic assistance, approach, and laterality to the NPA group. Clinical hip arthroplasty outcome thresholds, complications, and revision surgery rates were compared between cohorts. A Kaplan-Meier analysis was performed to assess survivorship.

Results

A total of 108 patients were included, 54 in each group. The groups displayed comparable outcomes at minimum ten-year follow-up, for modified Harris Hip Score (mHHS) (p = 0.370), Harris Hip Score (HHS) (p = 0.370), Hip dysfunction and Osteoarthritis Outcome Score for Joint Replacement (HOOS-JR) (p = 0.380), Forgotten Joint Score (FJS) (p = 0.250), visual analogue scale (VAS; p = 0.150), and patient satisfaction (p = 0.310). The two groups reached Patient Acceptable Symptom State (PASS) for FJS, HHS, and HOOS-JR at similar rates (p > 0.05). The PA group exhibited a significantly higher complication rate, with 11 major complications compared to two in the NPA group, translating to a relative risk of 2.8 (p < 0.033). Among the major complications in the PA group, nine required revision surgery, resulting in a relative risk of 4.5 (p < 0.047).

Conclusion

Patients undergoing primary THA with a history of prior hip arthroscopy achieve similar long-term functional outcomes compared to a propensity-matched control group. However, they face a 2.8-fold increased risk of complications and a 4.5-fold higher risk of major complications requiring revision THA.

Cite this article: Bone Jt Open 2025;6(9):1022–1030.

Introduction

Hip arthroscopy procedures have become increasingly prevalent in the last decade for the treatment of intra- and extra-articular pathologies. Between 2004 and 2009, there was a remarkable 365% increase in the rate of hip arthroscopy observed within the USA.^1,2 This trend has persisted, with the incidence of hip arthroscopy in patients with labral pathology in the USA rising by 85% from 2011 to 2018.³

A well-established association exists between femoroacetabular impingement (FAI) and labral pathology with the development of hip osteoarthritis (OA).^4-6 Arthroscopic treatment for FAI and labral pathology is presumed to have an impact on the progression of hip OA, potentially delaying or preventing the need for future hip arthroplasty.⁷ However, in some cases, the degenerative process may persist, eventually necessitating total hip arthroplasty (THA).⁸ This progression rate appears to be linked to higher age and BMI, as well as the presence of preoperative OA or cartilage damage before the hip arthroscopy.^9-14

With the increasing prevalence of hip arthroscopy and the expected degenerative process, it is likely that the proportion of patients needing a THA with history of previous hip arthroscopy will increase over time. While several studies have yielded conflicting conclusions with some reporting no differences between cohorts in terms of clinical outcomes or complication rates, an emerging body of short- and mid-term literature suggests a higher risk of complications for patients with a history of hip arthroscopy after THA potentially due to the presence of scar tissue from the prior hip arthroscopy, especially when using the anterior approach, and/or a second violation of the capsule when using the posterior approach.^15-24

This study aims to compare the long-term outcomes, complications, and revision rates of primary THA patients with a history of previous ipsilateral hip arthroscopy (PA) to a propensity-matched cohort of primary THA patients with no history of ipsilateral hip surgery (NPA). Our null hypothesis there will be no difference in long-term patient-reported outcomes.

Methods

Participation in the American Hip Institute Registry

All patients included in this study consented to participating in the American Hip Institute Hip Preservation Registry. Although the findings of the present study represent a unique analysis, some patient data may have been reported and used in previous studies.

Patient selection criteria

During the study period of November 2010 to November 2013, data were retrospectively reviewed from our institution’s prospectively maintained database for all hips who underwent primary THA for hip OA with minimum ten-year follow-up. Follow-up was defined as having completed postoperative questionnaires for modified Harris Hip Score (mHHS),²⁵ Harris Hip Score (HHS),²⁶ Hip dysfunction and Osteoarthritis Outcome Score for Joint Replacement (HOOS-JR),²⁷ Forgotten Joint Score (FJS),²⁸ and/or visual analogue scale for pain (VAS) at minimum ten-year follow-up, or having reached an endpoint defined as revision THA during the study time frame. Postoperative questionnaires were collected via email, telephone interview, and during clinic visits. Hips were included in the PA group if they underwent primary THA and had previous ipsilateral hip arthroscopy; hips from patients who did not report a prior ipsilateral hip arthroscopy were included in the control group. Hips were excluded if the patient was unwilling to participate, had workers’ compensation claims, had inflammatory or autoimmune arthropathies, or had previous ipsilateral hip surgery other than hip arthroscopy.^29,30Patient selection is shown in Figure 1. Complications were classified as minor complications if they did not require further surgical intervention and major complications if they needed to be addressed surgically. Furthermore, revision surgery was defined as when at least one of the components (acetabular component, femoral component, head, polyethylene liner) was exchanged.

Fig. 1.

Patient selection flowchart. NPA, no previous ipsilateral hip arthroplasty; PA, previous ipsilateral hip arthroplasty; THA, total hip arthroplasty; WC, workers’ compensation claims.

Patient-reported outcome measures (PROMs) and complications were documented in clinical visits and follow-up questionnaires and were accounted for the survivorship analysis. This study was approved by the institutional review board.

Patient demographic characteristics

A total of 108 hips (103 patients) were included in the study, with 54 hips (91.5%) in the PA group and 54 hips (95.2%) in the NPA control group. (Figure 1). The PA cohort consisted of 34 females (63%) and 20 males (37%) with a mean age at THA of 55.8 years (SD 9.1; 39.7 to 77.6), and mean BMI of 28.5 kg/m² (SD 4.8; 20.5 to 45.3). There were 14 hips (26%) that underwent an anterior approach THA and 40 (74%) that underwent a posterior approach THA. Robotic assistance was used for 30 hips (55.5%). The mean follow-up time was 122.7 months (SD 17.2; 120 to 145.1). The mean time to conversion to THA was 17.4 months (SD 15.6; 1 to 72). There were no significant differences in demographics between the PA and NPA groups, as seen in Table I.

Table I.

Matched patient characteristics of the population.

Characteristics	NPA group	PA group	p-value*
Hips, n	54	54
Sex, n (%)
Male	18 (33.3)	20 (37.0)	0.840
Female	36 (66.7)	34 (63.0)
Side, n (%)
Right	30 (55.6)	27 (50.0)	0.700
Approach, n (%)
Posterior	41 (75.9)	40 (74.1)	> 0.999
Robotic assisted surgery, n (%)
Yes	30 (55.6)	30 (55.6)	> 0.999
Mean age at surgery, yrs (SD; range)	56.6 (9.1; 37.6 to 85.9)	55.8 (9.1; 39.7 to 77.6)	0.630
Mean BMI, kg/m2 (SD; range)	28.5 (4.8; 19.5 to 40.7)	28.8 (4.8; 20.5 to 45.3)	0.840
Mean follow-up time, mnths (SD; range)	122.6 (17.8; 120 to 147.6)	122.7 (17.2; 120 to 145.1)	0.970
Time to conversion to THA, mnths (SD; range)	N/A	17.4 (15.6; 6.1 to 71.4)	N/A

^*For continuous variables, the Shapiro-Wilk test was performed to test for normal distribution, with p > 0.05 indicating normality. The paired two-tailed t-test or Wilcoxon signed-rank test were used to compare continuous data between groups. To detect significant differences between categorical variables, the chi-squared and Fisher's exact tests were used.

N/A, not applicable; NPA, no previous hip arthroscopy; PA, previous arthroscopy.

Surgical indication

Patients who fell under the following criteria were recommended for THA: patients presenting symptomatically with signs of hip OA on provocative examination; patient history; and radiological analysis. Those who failed to improve after a minimum of three months of conservative treatments, including rest, activity modification, nonsteroidal anti-inflammatory drugs, cortisone intra-articular injections, and physical therapy, were indicated for surgery.³¹

Surgical interventions

All THAs in this study were performed by the senior author (BGD). Every patient underwent THA with either spinal or general anesthesia. Intravenous tranexamic acid (10 mg/kg) was administered to each patient before incision, and the operative hip was prepared and draped in conventional sterile fashion. Either the direct anterior or posterior approach was performed. During the study period, the senior author transitioned from the posterior to the direct anterior approach. The final decision on approach was based on patient characteristics, such as BMI, the presence of an anterior overhanging pannus, and patient-doctor counselling preferences. In cases with robotic assistance, a preoperative CT scan-based system, the MAKO robot (Stryker, USA), was used; the system was introduced to our institution in 2011, after this all THAs are done using this technology unless requested otherwise. All patients received a press-fit wedge collarless femoral component, an uncemented acetabular component, and 36 mm ceramic heads.

Propensity-score matching design

Using R v. 4.3.2 (R Foundation for Statistical Computing, Austria), propensity-score matching was employed to control for potentially confounding variables. Hips included in the PA group were propensity-score matched in a 1:1 ratio based on age at THA, BMI, sex, use of robotic assistance, approach, and laterality to a control group of primary NPA. Propensity matching was performed using an optimal pair-matching algorithm.³² Matching was performed until no further matches could be made.

Statistical analysis

An a priori power analysis was used to determine the sample size needed in each group to achieve 80% power. With a mean difference of 10 points in the mean follow-up mHHS between groups and p < 0.05, the power analysis determined that 37 cases were required in each group.³³ For continuous variables, the Shapiro-Wilk test was performed to test for normal distribution, with p > 0.05 indicating normality. The paired two-tailed t-test or Wilcoxon signed-rank test were used to compare continuous data between groups. To detect significant differences between categorical variables, the chi-squared and Fisher's exact tests were used. A Kaplan-Meier analysis was performed to assess THA survivorship free from revision and secondary surgeries. Survival curves were compared to determine significance among the group distributions and relative risk for failure was determined. Descriptive statistics, including means, SDs, proportions, and ranges, were reported when relevant. All statistical analyses were performed using R v. 4.3.2 (R Foundation for Statistical Computing). The level of significance was set at p < 0.05.

Functional hip outcomes thresholds

Patient Acceptable Symptom State (PASS) thresholds for postoperative THA outcomes were used to evaluate clinically relevant outcomes. PASS threshold values from previously published literature was used,^34-36 and the number of patients who met PASS for HOOS-JR (76.7 points), FJS (66.68 points), and HHS (93 points) was reported.

Results

Minimum ten-year outcomes and durability

Comparable mean values at latest follow-up across all PROMs, VAS, and patient satisfaction were found between the groups, as shown in Table II (p > 0.050).

Table II.

Patient-reported outcomes measures of the population at ten-year follow-up.

Mean measure (SD; 95% CI)	NPA group (n = 50)	PA group (n = 44)	p-value*
mHHS	93.4 (10.8; 50 to 100)	91.0 (14.6; 30 to 100)	0.370
HHS	94.0 (9.7; 55 to 100)	91.8 (13.3; 36 to 100)	0.370
FJS	88.7 (14.5; 40 to 100)	87.0 (21.0; 0 to 100)	0.250
VAS	0.9 (1.7; 0 to 7)	1.5 (2.2; 0 to 8)	0.150
Satisfaction	9.2 (1.2; 4 to 10)	8.9 (2.1; 2 to 10)	0.310
HOOS-JR	90.1 (14.1; 56 to 100)	87.3 (16.4; 40 to 100)	0.380
VR12-Mental	62.7 (5.9; 47 to 68)	60.9 (9.1; 27 to 69)	0.250
VR12-Physical	52.6 (7.4; 27 to 59)	51.6 (7.1; 25 to 58)	0.500
SF12-Mental	58.2 (5.6; 42 to 67)	56.5 (8.8; 20 to 67)	0.250
SF12-Physical	51.7 (7.6; 25 to 60)	50.3 (7.3; 27 to 57)	0.380

^*The Shapiro-Wilk test was performed to test for normal distribution, with p > 0.05 indicating normality. The paired two-tailed t-test or Wilcoxon signed-rank test were used to compare continuous data between groups.

FJS, Forgotten Joint Score; HHS, Harris Hip Score; HOOS-JA, Hip dysfunction and Osteoarthritis Outcome Score for Joint Replacement; mHHS, modified Harris Hip Score; NPA, no previous arthroplasty; PA, previous arthroplasty; SF12, short-form 12-item health survey; VAS, visual analogue scale; VR-12, Veterans RAND 12-item health survey.

Figure 2 shows a two-year, five-year, and ≥ ten-year comparison of mHHS, HHS, FJS, VAS, and patient satisfaction to display the long-term durability of PROMs over time. No significant deteriorations (p > 0.050) were detected in either group from two years to ≥ ten years. Hips in the PA group had a significant improvement over time for the FJS (p < 0.050).

Fig. 2.

Durability comparison of patient-reported outcome measures (PROMs) for two-, five-, and ten-year follow-up between no previous ipsilateral hip arthroplasty (NPA), no previous hip arthroscopic surgery, previous ipsilateral hip arthroplasty (PA), and previous hip arthroscopic surgery. Forgotten Joint Score (FJS) for PA showed a significant increased score between two years and ten years (p < 0.050). There were no significant differences in any of the other groups or PROMs. HHS, Harris Hip Score; mHHS, modified Harris Hip Score; VAS, visual analogue scale, and patient satisfaction.

Functional hip outcome thresholds

The two groups demonstrated similar PASS score achievement rates for FJS, HHS, and HOOS-JR with no significant differences at latest follow-up, as shown in Figure 3.

Fig. 3.

Percentage of patients reaching the Patient Acceptable Symptom State (PASS). FJS, Forgotten Joint Score; HHS, Harris Hip Score; HOOS-JR; Hip disability and Osteoarthritis Outcome Score for Joint Replacement (p > 0.050 for all patient-reported outcome measure).

Complications

The PA group reported an overall higher rate of complications at 25.9% compared to 9.3% of the NPA benchmark control group (p < 0.050), with a calculated higher relative risk of 2.8 (95% CI 1.1 to 7.2; p = 0.033). The PA group experienced three (5.5%) minor complications that included two (3.7%) superficial wound infections successfully treated with a short course of antibiotics and one (1.9%) femoral nerve palsy that fully recovered after a period of observation. In addition, 11 (20.4%) experienced major complications, including one (1.9%) iliopsoas tendonitis requiring further arthroscopic surgery for an iliopsoas fractional lengthening, one (1.9%), heterotopic ossification that required further surgery for excision, and nine (16.7%) revision THA surgeries.

In the NPA cohort, there were three (5.5%) minor complications, including one (1.9%) superficial wound infection successfully treated with a short course of antibiotics, one (1.9%) deep vein thrombosis (DVT), and one (1.9%) femoral nerve palsy that fully recovered after a period of observation, with two (3.7%) major complications, both revision THA surgeries (Table III).

Table III.

Complications.

Complication type, n (%)	NPA group	PA group	p-value*
Minor
Superficial wound infection	1 (1.9)	2 (3.7)	0.560
Deep vein thrombosis	1 (1.9)	0 (0.0)	0.320
Femoral nerve palsy	1 (1.9)	1 (1.9)	> 0.999
Major
Iliopsoas tendonitis†	0 (0.0)	1 (1.9)	0.320
Heterotopic ossification‡	0 (0.0)	1 (1.9)	0.320
Complications leading to revision	2 (3.7)	9 (16.7)	< 0.05
Total	5 (9.3)	14 (25.9)	< 0.05

^*Bold indicates statistical significance (p < 0.050). The chi-squared and Fisher's exact tests were utilized.

^†Patients underwent iliopsoas fractional lengthening.

^‡Patient required surgical excision.

NPA, no previous arthroscopy; PA, previous arthroscopy.

Revision hip arthroplasty

There were nine revisions (16.7%) in the PA group with a calculated higher relative risk of 4.5 (95% CI 1.0 to 19.9; p = 0.047). The PA group reported one case (1.9%) of aseptic loosening of the femoral component, which subsequently required revision surgery with a femoral component exchange to a longer diaphyseal fixation component. Additionally, six hips (11.1%) experienced deep periprosthetic joint infections (PJIs), which were treated according to the onset of symptoms. Acute cases were managed with one-stage revisions (one hip), while those with a more latent or chronic onset underwent two-stage revisions (five hips). Additionally, two hips (3.7%) underwent revision surgery secondary to instability, all of which underwent cup revision and conversion to dual-mobility implants. In the NPA group, there were two revisions overall (3.7%), with both hips reporting a deep PJI, with one patient undergoing one-stage revision and one patient a two-stage revision. (Table IV and Figure 4).

Table IV.

Overview of revisions.

Revision type, n (%)	NPA group	PA group	p-value
Aseptic loosening femoral component	0 (0.0)	1 (1.9)	> 0.999
Periprosthetic joint infection	2 (3.7)	6 (11.1)	0.270
One-stage revision	1 (1.9)	5 (9.3)	0.206
Two-stage revision	1 (1.9)	1 (1.9)	> 0.999
Instability	0 (0.0)	2 (3.7)	0.238
Total	2 (3.7)	9 (16.7)	< 0.050

Bold indicates statistical significance (p < 0.05). The chi-squared and Fisher's exact tests were utilized.

NPA, no previous arthroscopy; PA, previous arthroscopy.

Fig. 4.

Kaplan-Meier survival curve for primary total hip arthroscopy with prior hip arthroscopic surgery over a ten-year period. Failure was defined as the need for revision hip arthroplasty. The x-axis represents time to failure in months; the y-axis is the statistical probability for survival (p = 0.019).

Discussion

The primary finding of the study was that hips that underwent primary THA with history of a previous hip arthroscopy showed comparable mean scores across all PROM and VAS scores, and displayed high patient satisfaction, at long-term follow-up. Importantly, these scores were sustained over time, with the PA cohort showing a significant improvement in mean FJS scores from the two-year to the ten-year follow-up. The groups reached PASS for FJS, HHS, and HOOS-JR at similar rates. However, hips with prior hip arthroscopy showed a higher frequency of complications with a relative risk of 2.8 (95% CI 1.1 to 7.2; p < 0.050) and an elevated risk of revision surgery with a relative risk of 4.5 (95% CI 1.0 to 19.9; p < 0.050).

Several investigations have examined and compared outcomes in patients who underwent primary THA with a history of previous ipsilateral hip arthroscopy. Rosinsky et al³⁷ reported mid-term outcomes of 34 PA patients matched to a benchmark control group of primary THA without previous hip arthroscopy. The results showed comparable and significant improvements in HHS, FJS, VAS, and patient satisfaction at a minimum five-year follow-up. Similarly, Haynes et al³⁸ studied a matched cohort of 58 hips that underwent primary THA with a history of previous ipsilateral hip arthroscopy. No statistical differences were found in the mean postoperatory score in the mHHS, the University of California-Los Angeles Activity Score, and three subscales from the Western Ontario and McMaster Universities Osteoarthritis Index (Pain, Stiffness, and Physical Function) between the groups, with a mean follow-up time of 42 months. Our results build upon the results of these previous studies by demonstrating that hips continue to experience favourable and sustainable improvement at long-term follow-up, with the PA group showing significant improvement in the FJS scores over time.

Despite the favourable clinical outcomes observed across all PROMs, THA as a secondary procedure after ipsilateral hip arthroscopy has shown controversial conclusion regarding complications and revision rates. For instance, Jain et al³⁹ reported on a cohort of 18 hips with a median age of 43 that had a previous hip arthroscopic surgery and underwent primary THA matched to a control group; no significant differences were found in complication rates between the cohorts with an average follow-up of 26 months. Similarly, Perets et al⁴⁰ published on 35 patients with a mean age of 53.4 years who underwent primary THA with a history of prior hip arthroplasty matched to control group with a minimum follow-up of two years, showing favourable clinical outcomes but no significant differences in complications or reoperation rate. However, both previously mentioned studies reported on short-term follow-up and were not powered to assess complications or revision surgeries.

Contrary to these prior reports, recent publications have suggested an increased risk for complications in patients who underwent primary THA with a history of ipsilateral hip arthroscopy.^{14,15,17,20,21} In a systematic review and meta-analysis including 15 short- to mid-term studies and 775 PA patients, Arakawa et al⁴¹ reported that PA patients had a higher rate of revision surgery and dislocations, with a calculated relative risk of 2 and 1.6, respectively (p < 0.01). There were no significant differences in clinical outcomes, estimated blood loss, or surgical time compared to primary NPA. Subsequently, Lemme et al¹⁵ reported findings from a propensity-matched cohort comprising 1,940 hips that underwent primary THA with history of prior ipsilateral hip arthroscopy with a minimum one-year follow-up. Patients with prior hip arthroscopy had an overall decreased implant survival within four years of undergoing the THA. Furthermore, a relationship between the time to conversion to THA from the previous hip arthroscopy was reported; patients who converted within one year experienced a higher risk of complications, leading to revision surgery. Similarly in our cohort, most of the major revisions occurred within five years post; the mean time from the arthroscopy to THA was 17.4 months (SD 24.5. Although this did not correlate with an increased risk of complications or revision surgery in our study, it remains an area for further investigation.

The current study aims to fill a gap in the literature by investigating the long-term impact of previous hip arthroscopic surgery in patients undergoing primary THA with a minimum ten-year follow-up, using a propensity-matched benchmark control group. The controlled confounding variables in this study hold significant importance. Sample size surpassed the required number by a priori power analysis, thus minimizing the likelihood of type II errors. The inclusion of multiple validated functional hip outcome scores also enhanced the reliability of the results. Finally, to provide clinical context of the mean PROMs, we reported the proportion of patients who achieved PASS for HHS, HOOS-JR, and FJS.

There are limitations of the current study that must be mentioned. This study was retrospective in design. Preoperative PROMs scores were not available for all patients. The cohort was small and included middle-aged patients who underwent hip arthroscopy, which has been identified as a risk factor for conversion to THA. Furthermore, the study was not powered to evaluate complications and secondary surgeries. Complete intraoperative data from the previous arthroscopic surgery was not available for all patients. Other cofounding variables like surgical time for the THA were not available for all patients. Additionally, given the limited cohort size, an accurate PASS calculation specific for these populations with a reliable area under the curve was not possible. Therefore, we have used previously published PASS scores. Furthermore, because all the procedures were conducted by a single institution, the generalizability of the results between institutions may be limited.

In conclusion, patients undergoing primary THA with a history of prior hip arthroscopy achieve similar long-term functional outcomes compared to a propensity-score match control group. However, they face a 2.8-fold increased risk of complications and a 4.5-fold higher risk of major complications requiring revision THA.

Take home message

- With the increasing prevalence of hip arthroscopy, the incidence of patients needing a total hip arthroplasty (THA) with a history of previous hip arthroscopy is increasing.

- Current studies have yielded conflicting conclusions and there is a gap in the literature on the impact on long-term outcomes.

- The present study found that patients with a history of previous arthroscopy demonstrated comparable long-term functional outcomes; however, they face a 2.8-fold increased risk of complications and a 4.5-fold risk of major complications requiring revision THA.

Data Availability

The datasets generated and analyzed in the current study are not publicly available due to data protection regulations. Access to data is limited to the researchers who have obtained permission for data processing. Further inquiries can be made to the corresponding author.