Article Text

Download PDFPDF

Femoroacetabular impingement surgery: are we moving too fast and too far beyond the evidence?
  1. Michael P Reiman1,
  2. Kristian Thorborg2,3
  1. 1Department of Orthopaedic Surgery, Duke University Medical Center, Durham, North Carolina, USA
  2. 2Sports Orthopedic Research Centre—Copenhagen (SORC-C), Amager-Hvidovre University Hospital, Copenhagen, Denmark
  3. 3Physical Medicine and Rehabilitation Research—Copenhagen (PMR-C), Amager-Hvidovre University Hospital, Copenhagen, Denmark
  1. Correspondence to Dr Michael P Reiman, Duke University School of Medicine, Department of Orthopaedic Surgery, Doctor of Physical Therapy Division, DUMC 104002, Durham, NC 27710, USA; reiman.michael{at}


Femoroacetabuler impingement (FAI) is becoming increasingly recognised as a potential pathological entity for individuals with hip pain. Surgery described to correct FAI has risen exponentially in the past 10 years with the use of hip arthroscopy. Unfortunately, the strength of evidence supporting both the examination and treatment of FAI does not appear to accommodate this exponential growth. In fact, the direction currently taken for FAI is similar to previously described paths of other orthopaedic and sports medicine pathologies (eg, shoulder impingement, knee meniscus tear) for which we have learned valuable lessons. The time has come for improved terminology, study design, and focus on delineation of successful treatment variables in the interest of those individuals with clinical indications of FAI so that we can appropriately address their needs.

  • Hip
  • Groin
  • MRI
  • Orthopaedics
  • Sporting injuries

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Myers et al1 pioneered surgical procedures for femoroacetabuler impingement (FAI) in 1999—they treated FAI after opening (dislocating) the hip joint in four patients previously treated with periacetabular osteotomy. Later, the arthroscope was used to resect the non-spherical head of FAI (cam lesion).2 Since then the diagnosis and treatment of FAI has become a standard component of orthopaedic and sports medicine practice. It is now one of the most common indications for hip arthroscopy.3–9

There has been an 18-fold increase in the procedure between 1999 and 2009, varying by geographic region in the USA.10 Hip arthroscopy increased by 365% between 2004 and 2009, especially in 20–39-year-olds.11 Furthermore, the FAI-surgery rate has increased by over 600% among newly trained surgeons from 2006 to 2010.12

Thus, FAI diagnosis and surgical correction has raced from medical innovation to institution seemingly based on continually increasing surgeon interest and prevalence of the diagnosis of FAI. Unfortunately, there is an alarming dearth of evidence of its efficacy. In this article we:

  1. Highlight the critical distinction between FAI morphology and pathology

  2. Explore the criteria for diagnosis of FAI

  3. Scrutinise the outcomes reported to date for FAI treatment

Pathology or morphology?

Delineating morphology from pathology is paramount. Morphology, relative to the hip, is the form and structure of the bones of the hip, whereas FAI is the mechanical abutment of the femoral head against the acetabulum.9 Cam refers to the morphology on the femur, pincer refers to morphology on the acetabular side of the hip joint.9 ,13 A misconception is that the presence of cam or pincer morphology automatically implies FAI and hence, pathology.

To more clearly define FAI (pathology), Sankar et al14 utilised the Medical Subject Headings thesaurus of the US National Library of Medicine. The thesaurus definition included five key aspects: (1) abnormal morphology of the femur and/or acetabulum, (2) abnormal contact between these two structures, (3) especially vigorous supraphysiological motion that results in such abnormal contact and collision, (4) repetitive motion resulting in the continuous insult and (5) the presence of soft-tissue damage. In addition to this, Sankar et al14 also emphasised that for the diagnosis of FAI, and thus as an eligibility criterion for clinical trials, individuals should be symptomatic and have a clinical presentation (based on history and physical examination findings) consistent with the diagnosis of FAI.

In agreement with Sankar et al,14 we would further stress that symptoms and reduced function need to be considered key aspects in the definition of FAI as they are the main reasons for patients to seek treatment of musculoskeletal problems. Previous descriptions and definitions of FAI that do not include patients’ pain and reduced function leave clinicians and researchers in doubt as to how potential FAI patients should be diagnostically addressed. It is, therefore, important to include patient history, symptoms, clinical examination, function and supportive diagnostic imaging in future orthopaedic and sports medicine practice when delineating pathological FAI from morphology.

Current FAI diagnosis—an imperfect foundation for treatment decisions

Current examination of patients with FAI pathology is strongly based on radiographic imaging and clinical examination.15–17 Owing to variable interpretations of clinical and radiological criteria, appropriate FAI diagnosis lacks consistency.18–23

The α angle

Radiographic examination for cam morphology often relies on the α angle.24 ,25 Definition of an ‘abnormal’ α angle is variable, with initial reports suggesting one that is greater than 50–55°.24 ,25 An α angle of 55° cut-off with MRI arthrography (MRA) had limited accuracy in establishing FAI. This has led to suggestions to increase the cut-off value for pathology, but also to questioning the use of α angle altogether.26 Further, in a general mixed-gender population, a 95% CI for a normal α angle was found to range widely from 32° to 62°,27 again suggesting large individual variability. More recent findings suggest α angles from 78° 28 and even up to 83° as the cut-off,29 indicating abnormal α angles are inadequately defined and likely much greater than the initial reports. Importantly, the presence of cam morphology on lateral radiographs, as suggested by a positive α angle, does not necessitate a decrease in clearance between the femoral head and acetabular rim.30 An additional concern is the ability of positional changes (supine vs standing) to significantly alter radiographic measures.31 α Angle will only give an indication of the size of the bony morphology and, as previously discussed, the diagnosis of FAI should also include clinical findings.

Delineating pathological findings from normal variants can also complicate the distinction between morphology and pathology. Radiographic findings of cam and pincer morphology, as well as labral pathological changes exist in 10–74% of asymptomatic individuals,32–58 including high-level athletes, none of whom developed persistent hip joint symptoms or missed games, and 77% of whom were playing at the same or higher level within a 4-year follow-up period.59 Clearly, the presence of morphology on radiographic imaging should not solely necessitate the diagnosis of pathology.

Physical examination tests

The most clinically utilised physical examination test for FAI is the flexion-adduction-internal rotation (FADDIR) test. This test, along with most of the other currently described tests for FAI, is generally more sensitive than specific. Hence, these tests serve better as screening rather than diagnostic tests due to a high sensitivity (and low likelihood ratio) assisting in ruling out/screening the potential existence of that pathology.60 ,61 The majority of these tests are purported to replicate the biomechanical abutment of the femoral head against the acetabulum. Unfortunately, the limited quality of diagnostic studies,60–62 as well as the fact that these studies are performed on patients with a biased high pretest probability (the vast majority (84–90%) of the patients in meta-analysis investigation have FAI),61 warrants caution in their clinical utilisation.60 ,61 Definitions60 ,62 and reliability22 ,63 ,64 of the use of these clinical tests between testers is also insufficient, and require better standardisation if they are to be useful as future indicators of hip joint pathology.64 Currently, these tests are only suggested to rule out individuals who do not have intra-articular hip pathology.60 ,61 ,65

Injection for diagnosis

Intra-articular injection has been proposed as a valid assessment method to determine the presence/absence of intra-articular pathology in the hip. Recent data suggest a negative response from an intra-articular injection is more likely to predict a negative outcome from arthroscopic management of FAI than having a positive response predict a favourable outcome from this surgery.66 Intra-articular injection not relieving all intra-articular pathologies shown on MRA suggests that all findings on MRA may not be contributors to the patient's complaints.63 Owing to the current low-quality literature, the diagnostic utility of intra-articular injection for the diagnosis of hip OA has not been given a clinical recommendation.67 Future high-quality research should address this promising and clinically logical approach for determination of intra-articular hip joint pain.

Treatment of FAI—more questions than answers

As mentioned previously, surgical treatment for FAI has progressed from open hip joint dislocation to arthroscopy. Despite robust growth in treatment of FAI, particularly hip arthroscopy,10–12 ,68 questions regarding treatment effectiveness remain.

What are the indications for FAI surgery?

A precise examination procedure to determine surgical intervention remains elusive. Clinical indications for surgical correction of FAI have relied predominantly on the aforementioned clinical and radiological indices. Increased use and interest in arthroscopic surgical correction of FAI must be tempered with proper selection of the ideal candidate. A more recent systematic review showed that indications for FAI surgery are inconsistently reported in the current literature.15 The exact clinical and radiographic indications suggested for arthroscopic management of FAI are also variable.15

This systematic review also showed that 20% of the reviewed studies relied on radiographic indicators only,15 leaving unknown the proper utilisation of clinical examination tests, patient-reported outcome scores and other imaging modalities.65 ,69 Owing to the high prevalence of radiological and clinical signs of FAI in patients with groin pain, we suggest that the decision of surgical intervention in many instances may be tempting and too ‘easy’ to make.

What is the level of evidence for FAI surgery?

The current evidence to answer this question is limited to case series (level IV) studies. While these case series demonstrate promising outcomes post FAI surgery, the outcomes are short term,70–72 dependent on status of chondral damage,73 and are notably less favourable in patients over the age of 40 years.74 Currently, no systematic reviews or randomised controlled trials (RCTs) on the efficacy of FAI-surgery compared to non-surgical or sham treatments exist.75 Recently, several trials on efficacy of FAI surgery have been registered on These trials seek to determine whether the innovation of FAI surgery will stand the test of evidence-based principles. A good lesson to learn from previous findings is that case series for other surgeries (eg, meniscus tear, shoulder impingement) are often favourable, but subsequent RCTs show no additional benefit over non-surgical or sham therapy.76–83

Reported postsurgical complication rates are quite variable.16 ,84 They range from the development of adhesions to deep venous thrombosis and chronic regional pain syndrome.85 Complication rates also appear to be higher than previously reported.12 ,86 ,87 Owing to variability and inconsistency in the current literature, reliable reporting of these complications has, therefore, been suggested as one of the principal requirements for future studies.16

What is the level of evidence for trials on the efficacy of non-surgical options?

Current evidence for non-surgical treatment is also limited to case-series (level IV) studies. The seemingly relentless demand for surgery for FAI patients has limited the opportunity to study efficacy of non-surgical treatment.88

Despite frequent use as a pre-requisite for surgery, non-surgical treatment is heterogeneous and not well reported. More recent reports of physical impairments in patients with FAI89–95 and following FAI surgery96 provide evidence to underpin non-surgical approaches.

A staged physiotherapy approach, emphasising an exercise-based focus and activity modification, led to improved patient-reported outcome in FAI patients.97 ,98 Both of these studies were prospective case series (level IV) with follow-ups from 1298 to 28 months,97 suggesting outcomes were comparable to surgical results.99 ,100 Appropriate non-surgical treatment, rather than being considered counterproductive,101 may therefore be considered an important part of differential decision-making of surgery appropriateness, making sure that sufficient exposure to physiotherapy and an exercise-based strategy has been exhausted. Future well-designed RCTs regarding the efficacy of specific non-surgical measures, such as physiotherapy and exercise, are therefore also essential.

What is the role of surgical correction of FAI for prevention of osteoarthritis?

Hip osteoarthritis (OA) was noted as one of the primary musculoskeletal global burdens of disease pathologies in 2010.102 Few effective medical treatment options are available, and none that have shown consistent slowing of the rate of disease.103 Pushing the agenda in this research field is suggested as a priority,104 and clarity on risk factors for hip OA is necessary.

Although some information exists on OA secondary to grossly visible deformities (acetabular dysplasia and femoral head tilt),105–107 there is almost no information on the natural course of more subtle femoral or acetabular deformities as present in FAI.108 Several studies have shown a lack of association with progression of morphology: α angle specific to the development of OA;54 ,109 acetabular index,54 ,110 ,111 centre edge angle,54 ,109 ,111 ,112 coxa profunda,109 coxa vara54 ,109 ,111 and pistol grip deformity54 ,110 ,113 with respect to development of labral pathology and hip OA.114 In fact, a retrospective analysis of 96 asymptomatic hips with radiographic evidence of hip impingement-related morphology revealed that 82% of these hips remained free of radiographic signs of hip OA for a mean duration of 18.5 years.54 Additionally, the overwhelming majority of hips with cam morphology did not reach end points of total hip arthroplasty or radiographic OA at even up to 19 years of follow-up.41 ,46 ,50 ,54 ,109 ,113 ,115–117 To date, there are no RCTs showing that hip morphology leads to early OA/earlier total hip replacement,118 and a recent systematic review found that only a few radiographic/clinical variables, such as joint space width, bony sclerosis, age and baseline pain seem to be strongly associated with the progression of hip OA.119

Prospective studies are needed to determine the relationship between hip morphology and future OA. Insufficient evidence exists to establish the relationship between pincer morphology and future OA.14 Cam morphology data examining this relationship, however, do exist. A prospective study of 723 asymptomatic Dutch individuals aged 45–65 years with 5-year follow-up revealed a 25% (with an α angle >83°) and 53% (α angle >83° and decreased internal rotation ≤20°) increased risk of developing end-stage OA,120 suggesting that combined radiographic and clinical measurements demonstrate more diagnostic promise than radiographic measurements alone. In the Dutch prospective study, considering α angles >60° alone only demonstrated a positive predictive value of 12% for developing end stage hip OA.120 Other population-based longitudinal studies121 ,122 have also shown increased α angles to be associated with future OA in individuals aged 55 to 80 years121 and 44–67 years.122 It is, however, important to mention that the age of the individuals in all of these studies120–122 is much higher than of the typical athlete and the majority of individuals normally undergoing FAI surgery.11

Prospective RCTs are necessary to completely elucidate the ability of surgical correction to prevent OA. It is also, therefore, too soon to conclude whether progression of OA is delayed with surgical correction.16 ,54 ,84 ,123 Experienced hip surgeons asserting that the link between morphology and OA of the hip has been established suggest that “not infrequently, we have come across patients with cam morphology who stand the test of time with minimal or no symptoms”109 and that “a large number of cam lesions are being excised unnecessarily.”124 RCTs comparing surgical and non-surgical management of FAI (and potential future OA development) are suggested as a future research priority.125

A call for learning from our past

The increase in popularity of FAI diagnosis and surgical intervention parallels the 746% upsurge in shoulder arthroscopy for impingement from 2000 to 2010,78 despite a lack of evidence that arthroscopy is superior to non-surgical treatment.76–80 In each of these conditions, clinical impressions of improvement from treatment often appear favourable, but these impressions can be deceiving when valid outcome measures, spontaneous recovery and/or placebo effects are not accounted for.126 Best practice in orthopaedics and sports medicine must routinely be based on evidence, shared decision-making, and monitoring and analysis of outcomes.126 Currently, interventions in orthopaedics and sports medicine are suggested to be suffering from overestimation of treatment effect and not being based on high level evidence.126

To improve on this with respect to the treatment of FAI, and in defence of the patients with clinical indications of FAI diagnosis, we respectfully suggest the following:

  • Similar to the proposal of shoulder impingement syndrome as a generic term for patients with shoulder pain,127 the term femoroacetabular impingement syndrome is suggested to more accurately encapsulate the nebulous nature of this condition.

  • Future investigation of the examination of FAI should focus on delineating those variables (eg, radiography, physical examination tests, intra-articular injection tests) that determine pathology and future treatment success.

  • There is an urgent need for studies regarding the efficacy of surgical correction, as well as non-surgical measures, in the treatment of patients with FAI. The studies should be RCTs, including placebo treatments. Without well-designed placebo controlled trials, ineffective treatment may continue unchallenged.128

Although surgeons agree in principle that a placebo trial has a scientific basis and is ethically acceptable,129 surgical versus non-surgical, or even placebo intervention type of studies83 ,130 ,131 have been criticised by orthopaedists on the grounds of selection bias.132–134 We appreciate the argument that it is difficult to convince patients to participate in these types of studies,135 ,136 where they fear to miss out on the already orthopaedically accepted ‘gold standard’ treatment.137 The problem of this argument, though, is that it also creates the very selection bias eventually criticised by orthopaedists, leaving the progress of orthopaedic and sports medicine practice in a real paradox, if such arguments are not challenged.

Finally, there is a need to openly discuss accurate and balanced information about intervention benefits and harms with patients. Patients often feel that surgery is their final salvation, without any considerations of the possibility of minimal, no, or even adverse, effects. Recent findings highlight the fact that the majority of patients are likely to overestimate intervention benefits and underestimate their potential harm,138 which should remind us all of the inevitable fact that surgery can only be revised, but never reversed.



  • Contributors MPR contributed to the idea of the manuscript; MPR and KT both contributed to initial drafting, writing, and revision/editing of the manuscript. MPR is responsible for overall content of the manuscript.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.