Study lacks rigour to determine choice of hip resurfacing or arthroplasty

Although Costa and colleagues’ prospective randomised study comparing total hip arthroplasty with resurfacing arthroplasty was laudable in orthopaedics, it lacks the necessary rigour and detail to reach useful conclusions.1

Costa and colleagues do not describe the surgical approach used, and this may have a clinically significant impact on outcome.2 At least one of them favours a trans-trochanteric approach, which could worsen the outcome for that group over the first year.

They do not give data on metal ions, so whether the resurfacings or hip replacements are “good” or “bad” is unknown.3

They do not describe the type of resurfacing implant used. Outcome varies enormously between the Birmingham hip resurfacing implanted in appropriate positions by experienced surgeons and other implants that fare much less well.4

They used traditional and non-discriminating outcome measures to measure function. These tests have a ceiling effect.5 A five point difference in Oxford hip scores is considerable. The reference they cite states that the minimum clinically significant difference is unknown, and whereas a score of 5 might be a good guess, a score of 3 or even 2 may be correct. They also do not give a reference for the minimum clinically significant difference for the Harris hip score, which undermines the power calculation and raises the possibility of a type 2 error.

We have validated a task based functional assessment tool that shows a significant difference between patients with total hip arthroplasty and Birmingham hip resurfacing for tasks that require balance and proprioception.6

Without the corroborative technical, radiographic, and functional data, this dataset and the cost effectiveness data are invalid.