Preoperative Hip Extension Strength Is an Independent Predictor of Achieving Clinically Significant Outcomes After Hip Arthroscopy for Femoroacetabular Impingement Syndrome

Preoperative Hip Extension Strength Is an Independent Predictor of Achieving Clinically Significant Outcomes After Hip Arthroscopy for Femoroacetabular Impingement Syndrome


The main aims are clearly set out in the introduction:

  1. To identify whether a correlation exists between pre-operative isometric hip strength and patient-reported outcome measures.
  2. To determine whether isometric hip strength is predictive of achieving the minimal clinically important difference (MCID) and patient acceptable symptomatic state (PASS).

A deviation from these aims as they are outlined however is the assessment also of the strength of the unaffected, unoperated hip which has also been given an equal weight in the analysis as reported in the results.


The visual reporting of the examination techniques supplied complements their methodology reported and is useful for a clearer interpretation by the reader.

It would appear that this was a focused study, prospectively collected data over a 6-month period however there may be a degree of selection bias which is not reported. It is not clear whether all patients undergoing hip arthroscopy for femoroacetabular impingement during this period has strength testing carried out or whether only those who had their strength testing assessed could be considered for inclusion. A flow chart with actual number of patients coming through the clinic may have been useful here, specifying the reasons for exclusion at each stage. The inclusion and exclusion criteria as reported are appropriate.

The 6 month follow-up is very short, and within the hip arthroscopy literature minimum 2 year post-operative data is generally required as improvements can be seen up to this point. The authors do correctly highlight this as a limitation but the rationale for choosing 6 months is not provided in the methodology. Particularly as this study was published almost 2 years following the completion of the data collection it is unclear why longer term outcomes were not evaluated.

While two trained investigators carried out the examinations, this raises some possible issues with the consistency of the measured readings. Although a handheld dynamometer has been validated for use there is no references provided by the authors for pre-existing literature of its use in studies. Further, there was no inter-rater reliability testing was carried out between these two investigators and as the limitations do state, significant variation in measurements may have arisen where the researcher is holding the dynamometer. Indeed as can be seen in the results the only instances where statistically significant correlations were observed was where the examiners were not holding the dynamometer.

Overall patient positioning for the assessment of the various ranges of hip strength testing could have been more robust. During each of the measurements there is no stabilisation of the hip which introduces scope for error. Focusing on the technique for measuring extension (which forms the basis of the paper) it was felt that the technique employed was not a true test for extension. Without a static structure in front of the patient to limit any forward movement, as the foot is pushed backwards toward the wall as instructed this will undoubtedly force the patient forward or else not perform their maximum extension ability. Simply providing a pole for stabilisation was felt to be insufficient.

The surgical technique and post-operative rehab protocol is appropriate as described.

Considering this is a study on patients diagnosed with FAI and undergoing hip arthroscopy as a surgical correction it would have been useful if the authors supplied data on the extent of the structural bony morphology (i.e. measurement of alpha angles to represent the cam deformities and centre-edge angle to represent the pincer deformities). This would make any results reported from this cohort more generalisable.

The PROMs chosen are appropriate for assessing post-operative outcomes for this demographic. As there is no evidence that any one PROM is superior for this patient group the use of multiple outcome measures in this study is appropriate.


92 patients were included with an 80% follow-up at 6 months. Comparisons between the affected and unaffected limb strengths are reported, showing a difference for flexion, extension, abduction and external rotation. No differences for internal rotation were found.

A major shortcoming from this study however is that the authors do not report the differences between pre- and post-op strengths of the affected limb. Considering this was a prospective study assessing hip strength among surgically treated FAI patients, it is surprising that no follow-up evaluating post-operative hip strength was carried out, or reported.

Additionally, the authors did also not evaluate or report whether there was any pain present during any of the strength tests. Pain will undoubtedly be a limiting factor in the ability to perform these tests. It is therefore unclear whether the deficiency in strength between the affected limb and unaffected limb was due to pain or generalised limited strength in the presence of symptomatic FAI.

Although both limbs measured were statistically significantly different at the time of initial measurement, both limbs, despite these differences, were shown to be predictors of outcome at 6 months. Considering the aims of this study was to identify if 'pre-operative isometric hip strength' correlated with PROs at 6 months, the inclusion of the unaffected limb in the analysis seems unnecessary as there was no direct intervention to this limb and therefore its assessment relative to post-operative outcomes is limited. It would perhaps have been more valuable from a clinical standpoint to attempt to show the extent to which either limbs affect achieving the MCID/PASS/6-month outcomes.

In the main body of the text the MCID and PASS achievability is reported as a percentage of the study group achieving this metric in any one of the three PROs used to assess outcome (78.8% and 74.3% respectively). This formatted reporting can be misleading as the actual proportion of cases achieving MCID and PASS within each of the separate PROs in isolation was less than these values. While this is reported in Table 5 it may have been better placed within the main text.

It should also be noted from Table 6, that all statistically significant correlations with respect to extension and post-op outcomes/MCID/PASS are considered weak in their strength (<0.4) apart from achieving the MCID for HOS-ADL (which had a correlation strength of 0.447, weak-moderate).

Within the regression analysis results of the text the authors report "hip extension measures of the affected side to be the only strength measurement significantly predictive of achieving either MCID or PASS" however this contradicts the results are reported in Table 8 where it is shown that extension of the affected side predicts achieving PASS only. While a statistical significant finding is true here, the borderline magnitude of this significance (p=0.049) should be noted.

Review Group Summary

  • The authors introduce an subject area with an research question which was relatively well planned
  • The methodology used and the way in which the results are interpreted do not support the conclusion for maximising pre-operative rehabilitation and hip muscle strength as a means to optimize post-operative outcome
  • The unaffected limb is just as much a predictor and it is questionable how relevant this finding is for patients undergoing hip arthroscopy for FAI
  • The manner and technique with which the tests were performed can result in variability and inaccuracies and the specificity of these techniques used to measure different hip strengths is questionable for its accuracy.
  • Pain was not highlighted as a potential limiting factor during hip strength examination.
  • The strength of the correlations and clinically important predictor variable identified are weak for the majority.
  • The Hip and Groin Clinic
  • International Society for Hip Arthroscopy
  • Whitfield Clinic
  • Queen's University Belfast
  • Royal College of Physicians and Surgeons of Glasgow
  • RCSI Royal College of Surgeons in Ireland