Abstract
Background: Resurfacing arthroplasty for cam-type deformities, which are a common cause of early osteoarthritis, is a technically demanding operation. Like any other arthroplasty, it requires both accuracy and precision. On the basis of the results of series reported by expert surgeons, we considered it desirable that this operation should be performed within ±10° of the desired angular orientation and ±6 mm of entry-point translation in 95% of hips. Technological aids are now available to help surgeons achieve that level of accuracy. Three models of cam-type hips of increasing severity were used to assess the efficacy of three systems of instrumentation at delivering the required level of accuracy and precision.
Methods: Thirty-two students of surgical technology were instructed in hip resurfacing and shown detailed plans of the desired operative outcome for the three hips with cam-type deformity. They then used conventional instruments, imageless navigation, and computed tomography-based navigation to perform the operation as accurately as possible.
Results: Conventional instrumentation produced an unacceptably wide range of entry-point errors. Imageless navigation was able to deliver adequate accuracy and precision in varus-valgus angulation and superoinferior translation, but was less satisfactory in version and anteroposterior translation. Computed tomography-based navigation enabled novice surgeons to navigate hips that had difficult cam-type deformity with acceptable precision in all four degrees of freedom measured.
Conclusions: Only computed tomography-based navigation appears to be appropriate for delivering both the accuracy and the precision needed by surgeons on the steep part of their learning curve. Neither conventional neck-based instrumentation nor imageless navigation provided enough help for novice surgeons learning to perform this technically challenging operation.
Clinical Relevance: Training with this computed tomography-based navigation system may shorten the learning curve for inexperienced surgeons, leading to a reduction in the prevalence of poor results and revision surgery.
Cam-type deformity of the femoral head has been established as a common cause of hip impingement1. Resurfacing arthroplasty is an operation that has been associated with promising early results for the treatment of hips at later stages of this disease, particularly in younger patients who have higher physical demands2-4. The procedure is considered to be particularly suitable for younger patients because it necessitates the removal of less femoral bone than would be removed during a total hip arthroplasty; however, for some surgeons, this gain is offset by the need to use a larger acetabular component to accommodate a larger femoral head to avoid notching5. Others have managed to perform resurfacing arthroplasty safely without the need for a larger acetabular component6, implying that surgical strategy may vary substantially. It is widely reported that the operation is more technically demanding than total hip arthroplasty, as the surgeon may be responsible for the complication of femoral neck fracture by inadvertently performing the operation inaccurately, that is, by inserting the femoral component either with angular or positional errors that may lead to a varus component position or by notching the femoral neck7-10.
Positioning in terms of translation error has not been quantified at all in any series of conventionally performed hip resurfacings; however, a greater risk of notching has been associated with an inferiorly placed head component in comparison with a superiorly placed one11.
When acquiring a new skill, the time taken to reach expert levels of performance is commonly described as a learning curve. The learning curve in hip resurfacing arthroplasty has been described as both steep12,13 and significant14, with an associated early failure rate of as high as 22%15. The results of examination of retrieved specimens following fracture also implicate learning-curve-associated technical errors in alignment and seating of the prosthesis as being an important cause of failure16. The higher failure rates reported are in part due to the surgeon experiencing problems in achieving the optimal angulation and positioning of both components17.
With increasing ease of image manipulation on personal computers and laptops, image-based preoperative planning is now accessible to ordinary surgeons. We have reported on image-based navigation as a method of shortening the learning curve for resurfacing when the single variable of varus-valgus angulation was considered18. We have also demonstrated the advantages of precise preoperative planning and the use of very low-dose computer-assisted tomography scanning during resurfacing19.
Currently, there are no published guidelines for either the precision or accuracy that should be delivered in resurfacing the femoral head. We have used the data from series published by designer surgeons to generate what we consider to be expert levels of accuracy and precision. In angular terms, this means a standard deviation of <5° in both planes. If surgical errors are distributed normally, then 95% of hips would be resurfaced with an angular error of less than ±10°. Entry-point precision in translation should be delivered with a standard deviation of <3 mm, meaning that 95% of hips would be resurfaced with an entry-point error of ±6 mm. In this study, we assessed whether a navigation or instrumentation system for helping surgeons master the learning curve and aid delivery of a high level of accuracy and precision could help them gain the experience necessary to achieve these levels in the skills laboratory with use of appropriate deformity models.
A randomized controlled trial was designed to test the hypothesis that planning made no difference to the precision with which hip resurfacing could be carried out. A power analysis (with alpha = 0.05 and beta = 0.2), based on the variability of femoral component position with use of a conventional or a navigation system by experienced surgeons, indicated that a minimum of eight participants in each group would be sufficient.
Thirty-two medical students who were studying toward a Bachelor of Science degree in surgical technology were recruited to participate. In their course, they had been taught the principles of hip resurfacing and how to plan the operation but had had no previous experience with either arthroplasty or navigation. They were instructed in both two-dimensional planning, which was based on plain radiographs and acetate templates, and three-dimensional planning, which was based on a computed tomography dataset and a hip planner (Version 1.2.1; Acrobot, London, United Kingdom). By combining the two-dimensional and three-dimensional visualization in the planner, the students were shown that it was possible to make small adjustments to optimize the fit of both components and avoid oversizing, notching, and angular malpositioning.
Models were made on the basis of the computed tomography datasets of three patients, representing worsening types of cam-type deformity—mild, moderate, and severe. The severity of cam-type deformity was assessed by the triangular indices: anteroposterior a angle, lateral a angle, and superior and inferior offset. These indices become more severe in each model. The optimal position and angulation for each type of hip were established with use of the rules established in earlier work19 (Figs. 1-A through 3-F).
The students were given the task of inserting a guidewire into the optimal position within a femoral head with use of three different methods:
Method 1. Conventional instruments supplied by the manufacturer (Corin, Cirencester, United Kingdom).
Method 2. An imageless navigation system (Hip Essential 5.0; BrainLAB, Westchester, Illinois), which allows navigation based on osseous landmarks acquired with use of infrared cameras and bone-mounted arrays of reflective balls, without the need for any diagnostic images. Numerous points are taken from different zones about the hip. After checking and confirming the accuracy of the model, the entry point of the guidewire can be adjusted to the desired angulation and translation in both planes. In this station, the students were asked to undertake three sequential tasks. They had to first attach the optical array to the bone with use of a single trochanteric pin. They then had to acquire landmarks on the dry bone, thereby allowing the system to generate the three-dimensional model and suggest the optimal component size. The students then had to navigate the guidewire to 130° of inclination and neutral version on the neck, avoiding notching by adjusting the entry point.
Method 3. A computed tomography navigation system (Acrobot). This is a low-dose computed tomography-based navigation system that makes use of two digitizing arms instead of optical tracking technology. One digitizing arm is fixed to the bone with use of two pins. It tracks the bone continuously while the other arm tracks whichever instrument is attached to it. Registration of the bone to the three-dimensional model involves acquiring approximately forty points in total from the three regions of interest: femoral head, greater trochanter, and lesser trochanter. After registration, a guidewire is passed with use of the navigated guidewire guide by aligning the crosshairs and circles of the tool to the target position. Students were asked to perform three sequential tasks. They first had to attach the arm to the bone with use of two self-drilling self-tapping pins. Next, they had to register the bone to the model by acquiring the points. Finally, they had to navigate the passing of a guidewire according to the plan by aligning the tool position to the target position shown on their plan.
A custom-designed lower-limb workstation (Medical Models, London, United Kingdom) was used to hold the dry bones. This allowed the proximal femoral inserts to be mounted and fixed to the table in the lateral position, representing the way a femur would appear to the surgeon through a posterior approach.
The participants were randomized into three groups. Each group undertook the task by using the three methods in a different order.
The three-dimensional plans were used independently in the conventional and imageless navigation stations and incorporated within the computed tomography navigation software as the target position. Students were not asked to modify the plan, but were simply given the plan and asked to achieve it with use of the technology provided. No detailed record of time was taken to avoid exerting any time pressure.
After completion of the tasks, all bones from the three stations were reregistered by one of the authors with use of the method described, and the computed tomography-based navigation (Acrobot) system was used to measure the trajectory and offset of the entry point from the target position. The guidewire error was measured in all four degrees of freedom: inclination, version, anteroposterior translation, and superoinferior translation.
A one-way analysis of variance on ranks was used to compare all three groups for errors in all four degrees of freedom with use of the Dunnett post hoc test.
All students completed the tasks required and no bones were eliminated from the analysis. All three systems achieved the accuracy standard set, and both imageless navigation and computed tomography-based navigation achieved the required level of precision in inclination. Only computed tomography-based navigation achieved the required levels of precision in anteroposterior translation and version.
Inclination and Version Accuracy
Both forms of navigation delivered accurate inclination and version (imageless navigation resulted in a mean inclination error of 0° and a mean version error of 0°; computed tomography navigation resulted in a mean inclination error of 1° and a mean version error of -1°). The technique that made use of conventional instruments was acceptable but less accurate (a mean inclination error of 4° and a mean version error of -2°) than either of the navigation methods.
Inclination and Version Precision
Both navigation systems delivered acceptable precision in inclination (imageless navigation resulted in an inclination standard deviation of 2°, and computed tomography-based navigation resulted in an inclination standard deviation of 1°). The conventional instrumentation technique was still acceptable but was significantly less precise (inclination standard deviation of 4°; p < 0.001, for both comparisons). There was no significant difference between the two navigation systems (p = 0.881). Only computed tomography-based navigation delivered acceptable precision in version (version standard deviation of 2°). Conventional instrumentation was just acceptable but significantly less precise (version standard deviation of 5°, p < 0.001), and imageless navigation was significantly less precise (version standard deviation of 9°) than computed tomography navigation (p < 0.001) or even conventional instrumentation (p = 0.015).
Translation Accuracy
Both computed tomography-based navigation and imageless navigation were accurate in translation in the anteroposterior plane, delivering a mean posterior entry-point error of approximately 1 mm, while conventional instruments were less accurate (posterior entry-point error of 3 mm). Computed tomography-based navigation was accurate in the superoinferior plane, delivering an inferior entry-point error of 0 mm, while imageless navigation (inferior entry-point error of 2 mm) and the conventional instrument technique (inferior entry-point error of 2 mm) were both significantly less accurate.
Translation Precision
In the anteroposterior plane, only computed tomography-based navigation delivered acceptable precision (posterior entry-point standard deviation of 1.6 mm). Computed tomography-based navigation was significantly more precise than either of the other two methods (the conventional method resulted in a posterior entry-point standard deviation of 3.4 mm [p < 0.003], and the imageless navigation method resulted in a posterior entry-point standard deviation of 6.2 mm [p < 0.001]). In the superoinferior plane, computed tomography-based navigation delivered acceptable precision, (inferior entry-point standard deviation of 1.1 mm). Imageless navigation was also acceptable (inferior entry-point standard deviation of 2.1 mm) but significantly less precise (p < 0.002). The conventional instrument technique was much less precise (inferior entry-point standard deviation of 4.4 mm) compared with imageless navigation (p < 0.018) or computed tomography navigation (p < 0.001).
When analyzed for significance with use of analysis of variance, neither the sequence of training (p = 0.8) nor the degree of cam deformity (p = 0.4) was found to be significant.
This study was designed to address the difficulties faced by inexperienced surgeons when they encounter a typical cam-type deformity in a young patient with an arthritic hip. The resurfacing devices available today are all based on the premise that the head is on the end of the femoral neck, something that is not true in this type of deformity. Correction of the deformity is not straightforward; indeed, correction is only possible in some hips and compromises must be made in others. This study suggests that computed tomography-based navigation can enable inexperienced surgeons to optimize both the position and orientation of the femoral head. In contrast, neither conventional instrumentation nor the imageless navigation system gave enough assistance to the inexperienced surgeon who was learning to treat this common deformity.
The study has a number of shortcomings: it is a dry-bone study that evaluated the performance of technically inexperienced surgical students. It is far from a genuine test of efficacy in the operating room, and it does not claim to be one. The students were well instructed, however, and they used the technologies available to try to achieve the best result possible. The highly reproducible three-dimensional data that are used in the objective analysis of dry-bone trials allow us to report at a higher level of accuracy and precision than it is possible to achieve in a clinical study. Nonparametric statistical tests were used to determine whether any significant differences could be detected on the basis of ninety-six observations of femoral guidewire insertion in pathological femoral heads. The levels of significance suggest that there are real differences between the methods.
While designing the study, we selected the range of ±10° for angular error and ±6 mm for translational error from the results we could extract from the published expert series with use of radiographic follow-up. A system that could offer that level of accuracy 95% of the time was one that patient groups would understand and that surgeons might find acceptable. A smaller range is certainly desirable, and in our experience can be delivered with computed tomography-based navigation and with use of robotic assistance. As yet, however, there is no clinical evidence that a very narrow range has clinical impact, while extreme outliers are accepted as a cause for concern. Therefore, for this experiment, we believed that to declare that the acceptable range of angular error should be ±5° might raise the bar too high. A larger range (more than 20° angular range and >1 cm in translation) would lack credibility among surgeons and patients when using the words "accurate" and "precise."
To our knowledge, no computed tomography-based clinical results have yet been published; thus, real validation of these ranges is lacking. The complex interaction between angular and translational errors makes it hard to compute a single scale for either accuracy or precision: one cannot simply add degrees to millimeters. Surgeons having made an error of entry-point translation may then accommodate that error with an alteration in angulation, and vice versa, and they may also increase the size of the femoral head, so these are not independent variables. We have not found, nor have we been able so far to devise, a scoring system that combines angular and positional accuracy and femoral head size in a useful quantitative way.
The conventional instrumentation used for this test was femoral neck-based. It was used under instruction of a trained surgeon and exactly as recommended by a major implant manufacturer. It did not provide an unskilled trainee with enough precision with large ranges both in angles and translation. We have not yet compared the different methods of conventional instrumentation but expect them to have their own strengths and weaknesses.
Imageless navigation was able to deliver adequate accuracy and precision in varus-valgus angulation and superoinferior translation even in these difficult hips within the limits proposed. Both inclination and superoinferior translation precision were within the limits proposed. However, version and anteroposterior translation were not delivered with acceptable precision in any of the three cam-type bone models, with a substantial range of angular and offset errors (18° and 12 mm).
Computed tomography-based navigation did deliver both sufficient accuracy and precision as defined before we started the study. Even inexperienced individuals were easily able to achieve accuracy and precision. The use of a detailed plan allowed the inexperienced surgeon to achieve levels of precision that were substantially better than those that would be demanded in terms of both alignment and orientation.
For the trained surgeon, the issues will be different. When faced with a cam-type hip deformity in which the head is deformed posteriorly and inferiorly, he or she will be able to use those transferable skills learned elsewhere in arthroplasty surgery to help achieve the preoperative plan. A skilled arthroplasty surgeon will be accustomed to making judgments of angulation and translation in his or her everyday practice, although there is little objective data to show just how good he or she is at doing so. Resurfacing the cam-type hip is technically difficult. Neither conventional instruments nor imageless navigation provides enough precision for an inexperienced surgeon who is learning the technique. The use of planned navigation does give enough precision to enable an inexperienced surgeon to achieve an accurate and a precise result with 95% confidence in the skills laboratory. This technology does appear fit for its purpose: using it, the inexperienced surgeon can gain valuable experience with the deformities that he or she will face in clinical practice, in the safety of the skills laboratory, with timely feedback regarding real accuracy and precision. Armed with this experience, the inexperienced surgeon can start higher up on the learning curve. 
Wagner S, Hofstetter W, Chiquet M, Mainil-Varlet P, Stauffer E, Ganz R, Siebenrock KA. Early osteoarthritic changes of human femoral head cartilage subsequent to femoro-acetabular impingement. Osteoarthritis Cartilage.2003;11:508-18.11508
2003
[PubMed][CrossRef]
Daniel J, Pynsent PB, McMinn DJ. Metal-on-metal resurfacing of the hip in patients under the age of 55 years with osteoarthritis. J Bone Joint Surg Br.2004;86:177-84.86177
2004
[CrossRef]
Vale L, Wyness L, McCormack K, McKenzie L, Brazzelli M, Stearns SC. A systematic review of the effectiveness and cost-effectiveness of metal-on-metal hip resurfacing arthroplasty for treatment of hip disease. Health Technol Assess.2002;6:1-109.61
2002
McMinn D, Treacy R, Lin K, Pynsent P. Metal on metal surface replacement of the hip. Experience of the McMinn prosthesis. Clin Orthop Relat Res.1996;329 Suppl:S89-98.329 SupplS89
1996
Loughead JM, Starks I, Chesney D, Matthews JN, McCaskie AW, Holland JP. Removal of acetabular bone in resurfacing arthroplasty of the hip: a comparison with hybrid total hip arthroplasty. J Bone Joint Surg Br.2006;88:31-4.8831
2006
[CrossRef]
Vendittoli PA, Lavigne M, Girard J, Roy AG. A randomised study comparing resection of acetabular bone at resurfacing and total hip replacement. J Bone Joint Surg Br.2006;88:997-1002.88997
2006
[CrossRef]
Shimmin AJ, Bare J, Back DL. Complications associated with hip resurfacing arthroplasty. Orthop Clin North Am.2005;36:187-93, ix.36187
2005
[CrossRef]
Mont MA, Ragland PS, Marker D. Resurfacing hip arthroplasty: comparison of a minimally invasive versus standard approach. Clin Orthop Relat Res.2005;441:125-31.441125
2005
[CrossRef]
Bowman NK, Bucher TA, Bassily AA. Fracture of the stem of the femoral component after resurfacing arthroplasty of the hip. J Bone Joint Surg Br.2006;88:1652-3.881652
2006
[CrossRef]
Shimmin AJ, Back D. Femoral neck fractures following Birmingham hip resurfacing: a national review of 50 cases. J Bone Joint Surg Br.2005;87:463-4.87463
2005
[CrossRef]
Beaulé PE, Dorey FJ, LeDuff M, Gruen T, Amstutz HC. Risk factors affecting outcome of metal-on-metal surface arthroplasty of the hip. Clin Orthop Relat Res.2004;418:87-93.41887
2004
[CrossRef]
McMinn DJ, Daniel J, Pynsent PB, Pradhan C. Mini-incision resurfacing arthroplasty of hip through the posterior approach. Clin Orthop Relat Res.2005;441:91-8.44191
2005
[CrossRef]
Cuckler JM. The optimal metal-metal arthroplasty is still a total hip arthroplasty: in the affirmative. J Arthroplasty.2006;21(4 Suppl 1):74-6.2174
2006
[CrossRef]
Schnurr C, Michael JW, Eysel P, König DP. Imageless navigation of hip resurfacing arthroplasty increases the implant accuracy. Int Orthop. 2007 Dec 22. [Epub ahead of print].
2007
Marker DR, Seyler TM, Jinnah RH, Delanois RE, Ulrich SD, Mont MA. Femoral neck fractures after metal-on-metal total hip resurfacing: a prospective cohort study. J Arthroplasty.2007;22(7 Suppl 3):66-71.2266
2007
[CrossRef]
Morlock MM, Bishop N, Rüther W, Delling G, Hahn M. Biomechanical, morphological, and histological analysis of early failures in hip resurfacing arthroplasty. Proc Inst Mech Eng [H].2006;220:333-44.220333
2006
Beaulé PE, Lee JL, Le Duff MJ, Amstutz HC, Ebramzadeh E. Orientation of the femoral component in surface arthroplasty of the hip. A biomechanical and clinical analysis. J Bone Joint Surg Am.2004;86:2015-21.862015
2004
Cobb JP, Kannan V, Brust K, Thevendran G. Navigation reduces the learning curve in resurfacing total hip arthroplasty. Clin Orthop Relat Res.2007;463:90-7.46390
2007
Barrett AR, Davies BL, Gomes MP, Harris SJ, Henckel J, Jakopec M, Rodriguez y Baena FM, Cobb JP. Preoperative planning and intraoperative guidance for accurate computer-assisted minimally invasive hip resurfacing surgery. Proc Inst Mech Eng [H].2006;220:759-73.220759
2006