Javascript not enabled
HipFurther Opinion

Imageless computer navigation without pre-operative templating may lead to malpreparation of the femoral head in hip resurfacing

M Olsen, E Davis, M Chiu, P Gamble, N Tumia, R Boyle, E Schemitsch

J Bone Joint Surg [Br] 2009;91-B:1281-6.


Contemporary hip resurfacing arthroplasty has recently gained popularity, and is considered as a viable alternative to conventional hip replacement in the young and active adult with osteoarthritis. Presumed advantages are preservation of native anatomy and bone stock of the hip, and good joint stability due to the use of large heads. Some encouraging early to mid term results have been published in the literature, sometimes comparable to those of THR. However, specific complications have been also reported, including the risk of neck fracture, neck narrowing and femoral implant loosening, and more recently the adverse effects of metal on metal bearings using big heads. Risk factors for premature failure are now well known, related to patient characteristics (female, obesity, poor bone quality, diagnosis), and to surgical technique (superior cortex notching, varus malalignment, improper sizing). Usually, the surgeon uses pre-operative planning on standard AP radiographs, then during the procedure, conventional jigs and/or templates, then visual corrections if necessary. The correct position of the femoral component is supposed to be 5 to 10° of valgus between the neck-shaft angle (NSA) and the stem-shaft angle (SSA). The accuracy of this method is questionable, depending on the underlying pathology, the conditions of the preoperative radiograph (position of the hip), and the experience of the surgeon. Moreover, little information is available about the version and translation of the femoral implant. A more modern technique is to use computer assisted navigation, either CT based or imageless. In the literature, there are contrasting results on the increased accuracy of computer navigation compared to conventional instrumentation in hip resurfacing. Overall, the superiority of navigation seems to be demonstrated in frontal plan but not fully in the others, particularly with the use of image free navigation.

The hypothesis of the authors was that imageless navigation was not a powerful tool in the placement of the femoral component in hip resurfacing. They compared the NSA and the size of the femoral implant in 100 consecutive patients with osteoarthritis, measured first on pre-operative planning and then on imageless navigation. They found considerable differences between the methods, with a mean absolute difference of 16° (ranging from 42° of varus to 52° of valgus). 60% computed NSAs were valgus, 38% varus and 2% equal to the planned NSA. Only 16% of the computed NSAs were within 3° of the measured NSAs. The computed size of the femoral component was larger in 76%, smaller in 11% and equal in only 13%. The same evaluations were made in a cadaver study and gave similar conclusions with smaller differences (mean absolute NSA difference = 5.4°, ranging from 22° of varus to 24° of valgus, 61 % of oversizing) and low level of repeatability.

Based on this experience, the authors do not condemn the imageless navigation in hip resurfacing but stress the risk of malposition of the femoral component using only this method. They insist on the importance of simultaneously using conventional pre-operative planning.

One agrees with this conclusion, considering navigation as a tool, only helping the surgeon, but not supervising him. Hip resurfacing is a demanding procedure, probably more so than conventional THR. Hip resurfacing implies that the native anatomy of the hip should be restored. The question is to know what is native anatomy in each patient? In others words, what is the optimal angle, length and version of the femoral component for a given patient, particularly if the patient has bilateral hip disease, destructive lesions or severe bone deformity? The authors do not answer to this question and focus mainly on the placement in the coronal plane. Their methodology might be open to criticism because they do not report the rate of malposition and inappropriate sizing in their series, and only compare mean values, whereas large inter individual differences might be expected. It is therefore difficult for the reader to conclude on the accuracy of both methods. Errors are not strictly related to the method itself, but rather to the conditions of registration of the data, in preoperative planning and in navigation as well. Both show multiple bias of registration and a rather low intra- and inter-observer reliability, even with experienced surgeons.

Bearing in mind that the correct position and adequate sizing of the femoral implant is crucial to the success of the procedure and considering the inability of the tools to achieve it, this study might be thought to illustrate the limitations of hip resurfacing itself.


Bizot P, Professor, Chief of the Department of Orthopaedic Surgery

CHU Angers, University of Angers, Angers, France

E-mail :