Clinical outcome scores for arthroscopic femoral osteochondroplasty in femoroacetabular impingement: a quantitative systematic review

Abstract

Background and aims

Femoroacetabular impingement is the abnormal contact of the proximal femur and acetabulum during motion. It causes hip pain and joint degeneration in young patients. This systematic review aims to clarify the clinical effect of arthroscopic femoral osteochondroplasty for cam lesions and to review the available literature for the general medical readership, including providers of primary and secondary care.

Methods and results

Electronic databases were searched for studies of arthroscopic femoral osteochondroplasty in primary femoroacetabular impingement. A total of 2618 article titles, 242 abstracts and 33 full text articles were considered. Ultimately nine studies with clinical outcome scores met the inclusion criteria and were included in the qualitative systematic review. Six studies were suitable for meta-analysis using an inverse variance, random effects model (RevMan software).

In the nine studies, improvements were seen in Western Ontario and McMaster Universities Osteoarthritis index, Non-arthritic Hip Score and Modified Harris Hip Scores. Across the six studies suitable for meta-analysis (537 patients), a 24-point weighted mean improvement in Non-arthritic hip score was seen. This yielded a large overall effect size of 1.6.

Conclusion

Arthroscopic femoral osteochondroplasty appears to be a beneficial treatment for primary femoroacetabular impingement, with a large effect size seen across six eligible studies.

Keywords

Arthroscopy osteochondroplasty femoroplasty femoroacetabular hip

Introduction

Primary femoroacetabular impingement (FAI) was described by Ganz as a cause of hip pain and osteoarthritis.¹ It results from abnormal contact/impingement of the femoral neck and acetabular rim. This reduced joint clearance can be a consequence of increased non-sphericity of the femoral head–neck junction (cam) and/or acetabular over-coverage (pincer).^1,2 These are described by alpha angle of ≥55° (cam lesion)^3–5 and a centre edge angel ≥40°(acetabular over-coverage, i.e. pincer impingement).⁶ These pathologies can be combined (mixed FAI). Early reports from a prospective cohort suggest that cam deformity is associated with a significant increase in the risk of end stage osteoarthritis (OA) at 5 years.⁷

Symptoms of the FAI include groin pain associated with sporting activity and prolonged or repetitive hip flexion.^1,2 The clinical impingement sign involves the elicitation of groin/hip pain from forceful flexion and internal rotation of the hip.⁸ Conservative measures can be trialled, e.g. non-steroidal anti-inflammatory drugs and activity modification, but these may mask the underlying joint destruction.²

Before the advent of arthroscopic osteochondroplasty of the hip, Ganz pioneered the use of surgical dislocation via a trochanteric osteotomy to facilitate adequate, safe exposure of the hip joint in treatment of FAI and other disorders.⁹ Due to the trochanteric osteotomy patients’ flexion and weight-bearing status must be restricted for up to eight weeks to allow the osteotomy to heal.² This type of surgery has shown improvements in the Merle d’Aubigne Hip Score,¹⁰ Harris Hip Score¹¹ and in both the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and short-form 12.¹²

There was subsequent interest in the use of therapeutic hip arthroscopy. Stahelin et al.¹³ described resection of the head–neck junction using a high speed burr until no femoroacetabular contact was visible in 110° flexion, 20° internal rotation and 20° adduction. To avoid fracture, it is suggested that not greater than 20% of the neck diameter is resected based on cadaveric biomechanical studies.¹⁴ Various clinical outcome scores are described in the literature and are summarised in Table 1.

Table 1.

Scoring systems in the literature.

Scoring System	Intended Use	Relevance in hip arthroscopy
Modified Harris Hip Score (MHHS)	Modification of the original Harris Hip Score to remove the clinical examination component. Primarily designed for osteoarthritis	Validated in hip arthroscopy.¹⁵ Correlates with patient satisfaction in hip arthroscopy¹⁶ and SF-36¹⁷
Western Ontario and McMaster Universities Osteoarthritis index (WOMAC)	Designed for and validated for use in osteoarthritis of the hip hip/knee^18–20	Not validated in hip arthroscopy. Considered an imperfect tool²¹
Non-arthritic hip score (NAHS)	A modification of the WOMAC which has been validated in 20–40-year-old patients, with normal radiographs	Correlates with Harris Hip Score and SF-12²²

Aim

A critical review of the literature was performed followed by analysis of preoperative and postoperative clinical outcome scores. The hypothesis was that arthroscopic femoral osteochondroplasty for cam lesions resulted in an improvement in clinical outcome scores.²³

Methods

Search process and article selection

Electronic database and hand searches were performed in March 2013. Medline and EMBASE databases were interrogated. Since FAI was first described in 1999,²⁴ the search was confined to 1999 to March 2013.

The following short search strings on the title and abstract: ‘femoroacetabular’; ‘femoro-acetabular’; ‘hip AND impingement’; ‘hip AND debridement’; ‘hip AND arthroscop$’ (where $ is a wildcard); ‘hip AND (osteochondroplasty OR osteoplasty)’; ‘femoroplasty’; ‘cam AND impingement’; ‘pincer AND impingement’. Each search string’s results was added to the project workspace on OVID. This simultaneous search within OVID allows an initial deduplication of articles with the same unique identifier.

PUBMED was interrogated for citations using the equivalent search strategies, specifically looking for articles which are not from Medline. Furthermore, The Cochrane Central Register of Controlled Trials was also searched for trials not indexed in the aforementioned databases.²⁵

To compensate for the lag time from publication to database indexing, a hand search of the table of contents of relevant journals covered the period October 2012–March 2013. Potentially relevant journals were: (1) American Journal of Sports Medicine, (2) British Journal of Sports Medicine, (3) Clinical Orthopaedic Related Research, (4) Journal of Bone and Joint Surgery (Am), (5) Bone and Joint Journal, (6) Operative Techniques in Orthopaedics, (7) Arthroscopy, (8) Hip international, (9) Journal of Arthroplasty, (10) Techniques in Orthopaedics, (11) Orthopaedics & Traumatology: Surgery and Research.

Citations were imported into Zotero reference manager (version 3.0.14) and an initial screening of article titles was made. Thereafter, potential studies were identified for further abstract review with suitable articles undergoing subsequent full text review.

Inclusion criteria:

Studies concern skeletally mature patients with primary FAI;

Cam/mixed impingement undergoing arthroscopic femoral osteochondroplasty ± procedure;

Minimum six months follow-up;

Peer reviewed journals;

Levels of evidence 1–4;²⁶

Have pre and postoperative clinical outcome scores.

Exclusion criteria:

Skeletally immature patients;

Secondary FAI (e.g. hip dysplasia);

Revision surgery;

Small case series < 10 patients;

Conference proceeding abstracts;

Duplicate/overlapping cohorts – the most recent will be used;

Review articles/ systematic reviews / meta-analysis / cadaveric studies.

Figure 1 demonstrates the selection process according to the template recommended by Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Figure 1.

Article selection process.

Data extraction

Details extracted include study design; population/age; timeframe for recruitment; inclusion criteria; surgical procedures performed; follow-up time; loss-to-follow-up rate, preoperative and postoperative clinical outcome scores and where applicable the p value for the difference; definition and rates of failure, e.g. conversion to total hip replacement (THR), revision surgery and complications. An overview is presented to establish similarities and attempt to draw wider statistical conclusions, i.e. meta-analysis.

Statistical analysis

Where possible, for each study, RevMan²⁷ was used to calculate the absolute mean change in clinical outcome scores and standardised mean difference (effect size).²⁸ The Non-Arthritic Hip Score (NAHS) was used across multiple studies. Due to the heterogeneity of the studies, a Random Effects (inverse variance) model was used to calculate the weighted mean change in NAHS and overall effect size.²⁹ An effect size of >0.8 is generally considered large.³⁰

Some studies do not describe the SD in the mean clinical outcome score. The authors were contacted to request these, but at this time some data were not available. Assuming an approximate normal distribution of data, the SD can be estimated for Cochrane Reviews and other meta-analysis using the method described by Walter and Yao.³¹

Results

The data were extracted from the individual studies tabulated to facilitate comparison of study design and patient population (Table 2). The individual study results are combined in Table 3 along with the absolute ‘change in mean’ and the ‘effect size’ and indeed an explanation for those studies which will not be included in the meta-analysis.

Table 2.

Study design and patient population.

Author (year) and reference	Study Type	Number of hips (subgroup)	Gender (total)^a	Type of FAI	Mean age (range)	Mean follow-up (months)	Inclusion criteria	Exclusion criteria	Other noteworthy comments
Bardakos et al.³² (2008)	Retrospective case series (historical control)	24	14 M: 10 F (24)	24 cam	33 (27–41)	12	Pure cam FAI	RA, Perthes, dysplasia, previous surgery, Tönnis grade 2 and above	All were Tönnis 0 or 1
Brunner et al.³³ (2009)	Prospective case series (comparative)	25 (non- navigated) 25 (navigated)	39 M: 11 F (50)	50 cam	43 (17–67)	27	Pure cam FAI, alpha >50°, +impingement sign	Tönnis grade 3, previous surgery
Brunner et al.³⁴ (2009)	Prospective case series	53	41 M: 12 F (53)	31 cam/22 mixed	42 (17–66)	29	Cam/mixed FAI, +impingement sign	Tönnis grade 3, previous surgery, RA, other MSK disease	Tönnis grade: 0(n = 8), 1 (n = 32), 2 (n = 13)
Byrd and Jones³⁵ (2009)	Prospective case series	207	138 M: 62 F (200)	163 cam/44 mixed	33 (range n/a)	16	Cam/mixed FAI	Bone-on-bone contact	OB grade: 0–2 (n = 17), 3 (n = 83), 4 (n = 107); 45% were athletes
Haviv et al.³⁶ (2010)	Retrospective case series	170	132 M: 34 F (166)	170 cam	37 (14–78)	22	Failure of 12 months non-operative treatment, + impingement sign,	Tönnis grade 3	OB grade: 2 (n = 35), 3(n = 83), 4 (n = 52)
Ilizaliturri et al.²¹ (2008)	Prospective case series	18	10 M: 8 F (18)	18 cam	34 (27–43)	24	Pure cam FAI	Dysplasia/SCFE/paediatric disease
Javed and O’Donnell³⁷ (2011)	Retrospective case series	40	26 M: 14 F (40)	40 cam	65 (60–82)	30	Failure of six months non-operative management. >60 years of age, pure cam FAI	Tönnis grade 2 and above, bilateral disease, Perthes, dysplasia, previous surgery	All were Tönnis 0 or 1
Palmeret al.³⁸ (2012)	Prospective case series	201	99 M: 102 F (201)	152 cam/49 mixed	40 (14–87)	46	Cam/mixed FAI, +impingement sign	Tönnis grade 3, dysplasia, synovitis, >1.5 cm chondral lesions	OB grade 0–3 (n = 156), 4 (n = 45); Tönnis grade: 0 (n = 174), 1 (n = 25), 2 (n = 1)
Stahelin et al.¹³ (2008)	Prospective case series	23	15 M: 7 F (22)	23 cam	42 (18–67)	6	Symptomatic, pure cam FAI	Previous surgery, Tönnis Grade 3	Tönnis grade: 0(n = 15), 1 (n = 5), 2 (n = 3)

Total patients (some may be bilateral).

FAI: femoroacetabular impingement; MSK: musculoskeletal; OB: outerbridge cartilage grading; RA: rheumatoid arthritis; SCFE: slipped capital femoral epiphysis.

Table 3.

Results for individual studies.

Author (year) and reference	Scoring System	Number of hips (subgroup)	Arthroscopic Procedures	Pre op Score		Post Op Score		Statistical Test and significance of change	Change in Mean (95% CI)	Effect Size (95% CI)	Include in meta-analysis (Y/N)
Author (year) and reference	Scoring System	Number of hips (subgroup)	Arthroscopic Procedures	Mean	SD	Mean	SD	Statistical Test and significance of change	Change in Mean (95% CI)	Effect Size (95% CI)	Include in meta-analysis (Y/N)
Bardakos et al.³² (2008)	MHHS	24	100% femoral OCP ± labral debridement# ± chondral procedure# ± microfracture#	Median 59 IQ range 52–64		Median 83 IQ range 75–87		Wilcoxon Signed Rank p < 0.001	Insufficient data		N (cannot use medians and IQ ranges)
Brunner et al.³³ (2009)	NAHS	25 (Non- navigated) 25 (navigated)	100% femoral OCP Labral debridement (n = 39) Microfracture (n = 20)	51.9	18.72¥	81.9	13.34¥	Paired t-test p < 0.001	30.00 (20.99, 39.01)	1.82 (1.15, 2.48)	Y
Brunner et al.³³ (2009)	NAHS	25 (Non- navigated) 25 (navigated)		56.6	13.34¥	89.5	10.16¥	Paired t-test p < 0.001	32.90 (26.33, 39.47)	2.73 (1.95, 3.52)	Y
Brunner et al.³⁴ (2009)	NAHS	53	100% femoral OCP Labral debridement (n = 41) Acetabular rim trimming (n = 22)	54.4	14.47¥	85.7	11.92¥	Paired t-test p < 0.001	31.30 (26.25, 36.35)	2.34 (1.85, 2.84)	Y
Byrd and Jones³⁵ (2009)	MHHS	207	100% femoral OCP ‘correction of pincer’ (n = 44) Microfracture (n = 58)	Described a mean score change of 20 but does not give pre and post op scores				No statistical tests	Insufficient data		N (insufficient data)
Havivet al.³⁶ (2010)	NAHS MHHS	170	100% femoral OCP ± chondral procedure# Microfracture (n = 29)	69.8	16.4	84.8	16.3	Paired t-test p < 0.001	15.00 (11.52, 18.48)	0.92 (0.69, 1.14)	Y
Havivet al.³⁶ (2010)	NAHS MHHS	170		70.7	16.6	86.1	16.2	Paired t-test p < 0.001	15.40 (11.91, 18.89)	0.94 (0.71, 1.16)	Y
Ilizaliturri et al.²¹ (2008)	WOMAC	18	100% femoral OCP Labral debridement (n = 15) Ligamentum teres debridement (n = 11)	82	9	89	9	Wilcoxon Signed Rank p = 0.001	7.00 (1.12, 12.88)	0.76 (0.08, 1.14)	N (WOMAC not validated in hip arthroscopy)
Javed and O’Donnell ³⁷ (2011)	NAHS MHHS	40	100% femoral OCP Labral debridement (n = 15)	62.1	13.24	77.2	15.4	Paired t-test p < 0.001	15.10 (8.81, 21.39)	1.04 (0.57, 1.51)	Y
Javed and O’Donnell ³⁷ (2011)	NAHS MHHS	40	100% femoral OCP Labral debridement (n = 15)	60.5	16.25	79.7	16.18	Paired t-test p < 0.001	19.20 (12.09, 26.31)	1.17 (0.70, 1.65)	Y
Palmer et al.³⁸ (2012)	NAHS	201	100% femoral OCP ± labral debridement# Labral repair (n = 11) Acetabular rim resection (n = 46)	56.1	15.9	78.2	15.8	Wilcoxon Signed Rank p < 0.001	22.10 (19.00, 25.20)	1.39 (1.17, 1.61)	Y (cautious of skew)
Stahelin et al.¹³ (2008)	NAHS	23	100% femoral OCP Labral debridement (n = 21) Labral repair (n = 2)	49.5	19.6	74	19.5	Wilcoxon Signed Rank p < 0.05	24.50 (13.20, 35.80)	1.23 (0.60, 1.87)	Y (cautious of skew)

IQ: range interquartile range for non-parametric data; MHHS; Modified Harris Hip Score (0–100); N: no; NAHS; Non-arthritic Hip Score (0-100); OCP: osteochondroplasty; #: number of procedures not detailed; SD: standard deviation; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index (0–96); Y: yes; ¥: calculated using method described by Walter and Yao (see ‘Methods’ section).

Table 4.

Meta-analysis including studies by Palmer and Stahelin.

Author (year) and reference	Number of hips (subgroup)	Pre op NAHS		Post Op NAHS		Weight^a (IV random effects)	Change in Mean (95% CI)	Effect Size (95% CI)
Author (year) and reference	Number of hips (subgroup)	Mean	SD	Mean	SD	Weight^a (IV random effects)	Change in Mean (95% CI)	Effect Size (95% CI)
Brunner et al.³³ (2009)	25 (Non-navigated) 25 (Navigated)	51.9	18.72¥	81.9	13.34¥	12.4%	30.00 (20.99, 39.01)	1.82 (1.15, 2.48)
Brunner et al.³³ (2009)	25 (Non-navigated) 25 (Navigated)	56.6	13.34¥	89.5	10.16¥	11.1%	32.90 (26.33, 39.47)	2.73 (1.95, 3.52)
Brunner et al.³⁴ (2009)	53	54.4	14.47¥	85.7	11.92¥	14.4%	31.30 (26.25, 36.35)	2.34 (1.85, 2.84)
Haviv et al.³⁶ (2010)	170	69.8	16.4	84.8	16.3	17.2%	15.00 (11.52, 18.48)	0.92 (0.69, 1.14)
Javed and O’Donnell³⁷ (2011)	40	62.1	13.24	77.2	15.4	14.8%	15.10 (8.81, 21.39)	1.04 (0.57, 1.51)
Palmer et al.³⁸ (2012)	201	56.1	15.9	78.2	15.8	17.3%	22.10 (19.00, 25.20)	1.39 (1.17, 1.61)
Stahelin et al.¹³ (2008)	23	49.5	19.6	74	19.5	12.8%	24.50 (13.20, 35.80)	1.23 (0.60, 1.87)
Total	537					100%	24.09 (18.42, 29.76)	1.58 (1.16, 2.00)
Weighted mean statistical analysis:	Heterogeneity: Tau²= 47.52; Chi² = 47.03, df = 6 (P < 0.00001); I²= 87% Test for overall effect: Z = 8.32 (P < 0.00001)
Overall size of effect statistical analysis:	Heterogeneity: Tau² = 0.25; Chi²= 45.14, df = 6 (P < 0.00001); I²= 87% Test for overall effect: Z = 7.34 (P < 0.00001)

Weight is illustrated for weighted mean rather than effect size.

df: degrees of freedom; IV: inverse variance; NAHS: Non-arthritic Hip Score (0–100); SD: standard deviation; ¥: calculated using method described by Walter and Yao (see ‘Methods’ section).

The meta-analysis (weighted mean change and effect size) has been performed twice. Table 4 includes the non-parametric data of Palmer et al.³⁸ and Stahelin et al.¹³ (mean change in NAHS 24.1 and effect size 1.6) and then repeated excluding these two studies yielding a mean change in NAHS of 24.6 and an effect size of 1.7 (this separate calculation has not been tabulated).

Using the data from all NAHS studies, a forest plot with each study’s mean change in NAHS, 95% confidence interval and the overall weighted mean change is shown in Figure 2.

Figure 2.

Weighted mean change in Non-arthritic Hip Score.

Discussion

There were no randomised controlled trials which met the inclusion criteria that compared non-operative and operative treatments. Three studies were retrospective,^32,36,37 while six were prospective.^{13,21,33–35,38} One of the papers by Brunner compared computer navigated with conventional arthroscopic surgery³³ but no clinical advantage was noted with navigation.

The NAHS was the most common outcome measure used in six studies,^{13,33,34,36–38} the Modified Harris Hip Score (MHHS) was used in four^32,35–37 and WOMAC in one.²¹ All nine studies reported improvements in the clinical outcome scores used, and those that reported complication rates described these as being low. Variable definitions of failure were used, including decreased clinical outcome score, conversion to THR and revision surgery. Where complications were discussed, generally a <1% rate of dysaesthesia/paraesthesia was reported, however this was as high as 10% in one paper.³³ Across all nine studies (786 patients), only two cases of heterotopic ossification were reported, and no cases of femoral neck fracture or avascular necrosis. There may be underreporting of complications.

All nine studies clearly stated their proportions of cam/mixed impingement and had otherwise similar inclusion/exclusion criteria. Seven studies mentioned Tönnis grade of osteoarthritis with five studies excluding grade 3^{13,33,34,36,38} and two studies excluding grade 2 and above.^32,37 Byrd and Jones³⁵ excluded those with ‘bone-on-bone contact’ but did not use the Tönnis grading system. Ilizaliturri included many patients with osteoarthritis but again, did not use Tönnis grade. Few studies mentioned the specific radiological criteria for inclusion, e.g. femoral alpha angle. The mean ages of the patients ranged from 33 to 65 years. If Javed’s study³⁷ was excluded the range would have been 33–43. Javed had specifically looked at patients over 60 years of age. It could be anticipated that his patients may have had a poorer outcomes based on age and likelihood of osteoarthritis, however all his patients were Tönnis grade 0 or 1. Interestingly, Javed’s study and that of Haviv et al.³⁶ had the poorest mean improvement in NAHS (15.1 and 15.0 points, respectively). Haviv’s patients had a high number of patients with Outerbridge grade 3/4 chondral damage. Javed and O’Donnell³⁷ and Haviv et al.³⁶ used both MHHS and NAHS with similar improvements for both scoring systems. For the purpose of meta-analysis and estimating the overall size of effect, both measures could not be included as the resultant overall size of effect would be disproportionately weighted in favour of these papers. The NAHS was the common system used across all eligible studies; it was therefore the only measure used to calculate the overall size of effect across these studies.

Three studies were excluded from the quantitative meta-analysis.^21,32,35 Byrd and Jones³⁵ described a mean ‘change in score’ of 20 for MHHS without stating the pre and postoperative means and standard deviations. Their data whilst interesting, and appearing in keeping with the other data sets, could not be used to estimate an effect size or weighted mean. Bardakos’ study³² was excluded since it described medians with interquartile ranges and used Wilcoxon’s signed rank test for comparing pre and postoperative MHHS. These data could therefore not be used to calculate an effect size.³⁹ Ilizaliturri was the only author to use the WOMAC scale,²¹ which was designed for osteoarthritis¹⁸ and has not been validated for use in hip arthroscopy. This raised concerns regarding its suitability for hip arthroscopy. Furthermore, the postoperative results had a non-parametric distribution of data. For these reasons, the study is not included in the quantitative meta-analysis.

Three papers used validated outcome measures (MHHS and NAHS) and provided enough data (pre and postoperative means, standard deviations and statistical tests) to be directly included in the meta-analysis.^13,37,38 Haviv provided the SDs for his data³⁶ by email and therefore had his study included. The two papers by Brunner et al.^33,34 had pre and post op means, the NAHS ranges and used parametric statistical tests, suggesting normally distributed data. These papers could therefore reasonably be included in the meta-analysis after having their standard deviations estimated using the method detailed by Walter and Yao.³¹

As discussed in the individual results section, the papers by Palmer et al.³⁸ and Stahelin et al.¹³ were analysed by their authors using Wilcoxon Signed Rank tests suggesting non-parametric data. The authors did, however, describe means and standard deviations. An attempt was made given the limited information available to consider the extent of this potentially abnormal distribution. RevMan is considered by the Cochrane Group to being robust enough to tolerate a small amount of skew if the data were reported as mean and SD.^39,40 The meta-analysis was therefore performed with these non-parametric studies and also without them. The weighted mean improvements in NAHS were 24.1 and 24.6, respectively, and the overall sizes of effects were 1.6 and 1.7, respectively (both considered large effects³⁰). We can therefore see that subjectively little difference is noted with and without their inclusion. Therefore, the full data set is presented in Table 3. This analysis included the data from 537 patients across six studies^{13,33,34,36–38} and gave a weighted mean improvement in NAHS of 24.09 points (95% CI 18.42–29.76) which has been demonstrated in Figure 2.

Table 1 demonstrates the considerable differences in mean age, gender and follow-up. Heterogeneity is considered to be extremely high based on tau²; the ratios of chi² to df (degrees of freedom), and the high I² statistics⁴¹ shown in Table 3. Heterogeneity is further demonstrated in Figure 2, where the 95% confidence intervals do not all overlap. This heterogeneity is due to differences in condition of articular cartilage, age, gender, activity levels, occupation, comorbidity and patient expectation in addition to surgical proficiency or technique. The random effects (inverse variance) model was therefore more appropriate than a fixed effects meta-analysis. Differences in the selection criteria between studies may affect the variance, i.e. SD. More restrictive inclusion criteria may reduce the variance and increase the weight given to that particular study using an inverse variance model. This would influence the meta-analysis result. Furthermore, as described, the high heterogeneity reduces the integrity of the weighted mean/overall size of effect.

Nonetheless, the individual mean changes and individual effect sizes can always be considered by the reader without such inter-study statistics/meta-analysis. Each study has therefore had a separate description of their patient characteristics, interventions and outcomes to allow the reader to make a qualitative assessment based on the narrative review in addition to the attempted quantitative meta-analysis. However, with extremely high heterogeneity, the meta-analysis results should be interpreted with caution.

What this review contributes to our understanding

Despite hundreds of articles having been published on FAI,^42,43 this review includes only nine studies in the qualitative review and six studies in the meta-analysis. It therefore may not represent the true effect seen in other studies that did not meet the inclusion criteria. Previous systematic reviews have covered a wide range and combination of treatments which vary in approach (open/arthroscopic/combined) and technique (femoral/acetabular osteochondroplasty, chondral and labral procedures). Unlike previous general reviews,^44–47 this review focuses on the outcomes of arthroscopic femoral osteochondroplasty, where objective clinical outcome scores are available. The general nature of previous reviews made it difficult for authors to perform inter-study statistics. Kemp et al. reviewed FAI studies in 2010 and calculated the individual sizes of effect.⁴⁸ They included studies that would not fit our inclusion criteria, i.e. not all patients underwent arthroscopic femoral osteochondroplasty, and they included groups of ‘pure pincer’ patients. Our study focuses on the effect and treatment of cam lesions, i.e. pure cam and mixed patients. It would otherwise be difficult to compare and contrast so many different studies. It is appreciated that this review does include studies of varied additional chondral and labral procedures, but this is very typical of FAI treatment in day-to-day practice. It would have been interesting to have focused on pure cam rather than including mixed FAI, but there was extreme paucity of such eligible studies.

Although the studies included in this review are of a ‘low level’ of evidence,⁶ and are particularly prone to selection/reporting/publication/other biases, their cohorts experienced statistically and clinically significant improvements in their MHHS and NAHS. One author has previously highlighted difficulty in getting a highly motivated and opinionated group of patients to agree to randomisation.⁴⁹ Nevertheless, McMaster University is currently recruiting patients for a randomised controlled trial of cam/mixed FAI with two treatment arms: arthroscopic femoral osteochondroplasty versus lavage.⁵⁰ Their results are expected in 2015.

Conclusion

We believe this review is the first to calculate a weighted mean difference in the NAHS but it remains necessary to improve the quality of FAI research. Until then, this systematic review does support a benefit from arthroscopic femoral osteochondroplasty for cam lesions but patient selection and magnitude of benefit remains unclear.

Footnotes

Declaration of conflicting interests

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Ganz

Parvizi

Beck

. Femoroacetabular impingement: a cause for osteoarthritis of the hip. Clin Orthop Relat Res 2003; 417: 112–120.

Lavigne

Parvizi

Beck

. Anterior femoroacetabular impingement: part I. Techniques of joint preserving surgery. Clin Orthop Relat Res 2004; 418: 61–66.

Nötzli

Wyss

Stoecklin

. The contour of the femoral head-neck junction as a predictor for the risk of anterior impingement. J Bone Joint Surg Br 2002; 84-B: 556–560.

Pfirrmann

CWA

Mengiardi

Dora

. Cam and pincer femoroacetabular impingement: characteristic MR arthrographic findings in 50 patients1. Radiology 2006; 240: 778–785.

Neumann

Cui

Siebenrock

. Impingement-free hip motion: the “normal” angle alpha after osteochondroplasty. Clin Orthop Relat Res 2009; 467: 699–703.

Kutty

Schneider

Faris

. Reliability and predictability of the centre-edge angle in the assessment of pincer femoroacetabular impingement. Int Orthop 2012; 36: 505–510.

Agricola

Heijboer

Bierma-Zeinstra

SMA

. Cam impingement causes osteoarthritis of the hip: a nationwide prospective cohort study (CHECK). Ann Rheum Dis 2012; 72: 918–923.

Wagner

Hofstetter

Chiquet

. Early osteoarthritic changes of human femoral head cartilage subsequent to femoro-acetabular impingement. Osteoarthritis Cartilage 2003; 11: 508–518.

Ganz

Gill

Gautier

. Surgical dislocation of the adult hip a technique with full access to the femoral head and acetabulum without the risk of avascular necrosis. J Bone Joint Surg Br 2001; 83: 1119–1124.

10.

Beck

Leunig

Parvizi

. Anterior femoroacetabular impingement: part II. Midterm results of surgical treatment. Clin Orthop Relat Res 2004; 418: 67–73.

11.

Yun

Shon

Yun

. Treatment of femoroacetabular impingement with surgical dislocation. Clin Orthop Surg 2009; 1: 146–154.

12.

Beaule

Le Duff

Zaragoza

. Quality of life following femoral head-neck osteochondroplasty for femoroacetabular impingement. J Bone Joint Surg 2007; 89: 773–779.

13.

Stahelin

Jolles

. Arthroscopic offset restoration in femoroacetabular cam impingement: accuracy and early clinical outcome. Arthroscopy 2008; 24: 51–57.

14.

Mardones

Gonzalez

Chen

. Surgical treatment of femoroacetabular impingement: evaluation of the effect of the size of the resection. J Bone Joint Surg 2005; 87(A): 273–279.

15.

Byrd

Jones

. Prospective analysis of hip arthroscopy with 10-year followup. Clin Orthop Relat Res 2010; 468: 741–746.

16.

Aprato

Jayasekera

Villar

. Does the modified Harris hip score reflect patient satisfaction after hip arthroscopy? Am J Sports Med 2012; 40: 2557–2560.

17.

Potter

Freedman

Andersen

. Correlation of Short Form-36 and disability status with outcomes of arthroscopic acetabular labral debridement. Am J Sports Med 2005; 33: 864–870.

18.

Bellamy

Buchanan

Goldsmith

. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988; 15: 1833–1840.

19.

Bellamy

Buchanan

. A preliminary evaluation of the dimensionality and clinical importance of pain and disability in osteoarthritis of the hip and knee. Clin Rheumatol 1986; 5: 231–241.

20.

Quintana

Escobar

Bilbao

. Responsiveness and clinically important differences for the WOMAC and SF-36 after hip joint replacement. Osteoarthritis Cartilage 2005; 13: 1076–1083.

21.

Ilizaliturri

Jr Orozco-Rodriguez

Acosta-Rodriguez

. Arthroscopic Treatment of cam-type femoroacetabular impingement. preliminary report at 2 years minimum follow-up. J Arthroplast 2008; 23: 226–234.

22.

Christensen

Althausen

Mittleman

. The nonarthritic hip score: reliable and validated. Clin Orthop Relat Res 2003; 406: 75–83.

23.

Safran

Hariri

. Hip arthroscopy assessment tools and outcomes. Oper Tech Orthop 2010; 20: 264–277.

24.

Myers

Eijer

Ganz

. Anterior femoroacetabular impingement after periacetabular osteotomy. Clin Orthop Relat Res 1999; 363: 93–99.

25.

The Cochrane Central Register of Controlled Trials (CENTRAL) | The Cochrane Collaboration [Internet], http://www.cochrane.org/handbook/6212-cochrane-central-register-controlled-trials-central (accessed 9 March 2013).

26.

Wright

Swiontkowski

Heckman

. Introducing levels of evidence to the journal. J Bone Joint Surg 2003; 85: 1–3.

27.

Review Manager (RevMan) [Computer program]. Version 5.2. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012.

28.

The standardized mean difference [Internet], http://handbook.cochrane.org/chapter_9/9_2_3_2_the_standardized_mean_difference.htm (accessed 10 March 2014).

29.

Incorporating heterogeneity into random-effects models [Internet], http://handbook.cochrane.org/chapter_9/9_5_4_incorporating_heterogeneity_into_random_effects_models.htm (accessed 10 March 2014).

30.

Cohen

. Statistical power analysis. Curr Direc Psychol Sci 1992; 1: 98–101.

31.

Walter

Yao

. Effect sizes can be calculated for studies reporting ranges for outcome variables in systematic reviews. J Clin Epidemiol 2007; 60: 849–852.

32.

Bardakos

Vasconcelos

Villar

. Early outcome of hip arthroscopy for femoroacetabular impingement: the role of femoral osteoplasty in symptomatic improvement. J Bone Joint Surg 2008; 90(B): 1570–1575.

33.

Brunner

Horisberger

Herzog

. Evaluation of a computed tomography-based navigation system prototype for hip arthroscopy in the treatment of femoroacetabular cam impingement. Arthroscopy 2009; 25: 382–391.

34.

Brunner

Horisberger

Herzog

. Sports and recreation activity of patients with femoroacetabular impingement before and after arthroscopic osteoplasty. J Sports Med 2009; 37: 917–922.

35.

Byrd

Jones

. Arthroscopic femoroplasty in the management of cam-type femoroacetabular impingement. Clin Orthop Relat Res 2009; 467: 739–746.

36.

Haviv

Singh

Takla

. Arthroscopic femoral osteochondroplasty for cam lesions with isolated acetabular chondral damage. J Bone Joint Surg 2010; 92(B): 629–633.

37.

Javed

O’Donnell

. Arthroscopic femoral osteochondroplasty for cam femoroacetabular impingement in patients over 60 years of age. J Bone Joint Surg 2011; 93(B): 326–331.

38.

Palmer

Ganesh

Comfort

. Midterm outcomes in patients with cam femoroacetabular impingement treated arthroscopically. Arthroscopy 2012; 28: 1671–1681.

39.

The Cochrane Collaboration open learning material [Internet], http://www.cochrane-net.org/openlearning/html/modA1-5.htm (2013, accessed 9 May 2013).

40.

Meta-analysis of skewed data [Internet], http://handbook.cochrane.org/chapter_9/9_4_5_3_meta_analysis_of_skewed_data.htm (2013, accessed 9 May 2013).

41.

Identifying and measuring heterogeneity [Internet], http://handbook.cochrane.org/chapter_9/9_5_2_identifying_and_measuring_heterogeneity.htm (2013, accessed 28 April 2013).

42.

Ayeni

Chan

Al-Asiri

. Sources and quality of literature addressing femoroacetabular impingement. Knee Surg Sports Traumatol Arthrosc 2013; 21: 415–419.

43.

Haviv

Burg

Velkes

. Trends in femoroacetabular impingement research over 11 years. Orthopedics 2011; 34: : 353–: 353.

44.

Matsuda

Carlisle

Arthurs

. Comparative systematic review of the open dislocation, mini-open, and arthroscopic surgeries for femoroacetabular impingement. Arthroscopy 2011; 27: 252–269.

45.

Baldwin

Harrison

Namdari

. Outcomes of hip arthroscopy for treatment of femoroacetabular impingement: a systematic review. Curr Orthop Pract 2009; 20: 669–673.

46.

Botser

Smith

Jr Nasser

. Open surgical dislocation versus arthroscopy for femoroacetabular impingement: a comparison of clinical outcomes. Arthrosc J Arthrosc Relat Surg 2011; 27: 270–278.

47.

Papalia

Del Buono

Franceschi

. Femoroacetabular impingement syndrome management: arthroscopy or open surgery? Int Orthop 2012; 36: 903–914.

48.

Kemp

Collins

Makdissi

. Hip arthroscopy for intra-articular pathology: a systematic review of outcomes with and without femoral osteoplasty. Br J Sports Med 2012; 46: 632–643.

49.

Zingg

Ulbrich

Buehler

. Surgical hip dislocation versus hip arthroscopy for femoroacetabular impingement: clinical and morphological short-term results. Arch Orthop Trauma Surg 2013; 133: 69–79.

50.

Femoroacetabular Impingement Randomized Controlled Trial – Full Text View – ClinicalTrials.gov [Internet], http://clinicaltrials.gov/ct2/show/NCT01623843 (2013, accessed 28 April 2013).