If Computer-Assisted Surgery Is More Accurate, Why Isn’t It More Prevalent?

Jefferson Morrison, MD
Southern Joint Replacement Institute

There are two things that all total knee surgeons can agree on when it comes to imageless computer navigation: It is more accurate than mechanical instrumentation in obtaining coronal alignment of the limb, and it is painful to adopt into one’s practice. Countless studies looking at different computer assisted orthopaedic surgery (CAOS) systems have shown improvement in alignment, but full adoption of this technology into surgeon’s operating rooms is uncommon.1 Most of us can easily list the reasons we tried and ultimately gave up on navigation. The cameras take up too much space. The lines of sight to the arrays are hard to keep open with assistants in the sterile field. The work flows are set and not customizable. The interface is outside of the sterile field so the surgeon has to rely on a company representative or nurse to “run” the system. The arrays require extra pins, sometimes outside of the incision. The reflective spheres quit working when they get blood on them or someone gets in between the array and the camera. Registration takes too long. The cutting blocks and instruments are specific to navigation, so if the case starts to go off the rails, converting back to conventional instruments is difficult and time consuming.

We should not, however, ignore the first point. CAOS is more accurate. The question, of course, is does this matter? There is plenty of evidence in the literature that alignment improves longevity.2-4 There is even good registry data that navigated total knees have a lower revision rate in one of the more high-risk demographics.5 There is even some literature to suggest that functional outcomes are better in navigated total knees.6

If we believe that navigation is more accurate, and that accuracy improves longevity and functional outcomes, what will it take to get us to fully adopt navigation into our operating rooms? Assume cost is not the roadblock. The system would have to be smaller and be fully incorporated into the surgical field. The software interface would have to be intuitive, run by the surgeon in the sterile field and customizable in real time. The arrays would have to be placed inside the incision, unaffected by blood and fluid and minimally affected by line of sight issues.

The ExactechGPS® computer assisted surgery system does all of this, and its accuracy has been validated in thousands of cases. I personally have used this system on and off for five years. Its accuracy allows me to do away with the intraoperative x-ray I typically take of my tibia cut. This saves me time, but the registration and “fiddle factor” still adds nine minutes on average to my navigated tourniquet times. While I am very accustomed to the extra pins needed to fix the arrays to the femur and tibia as well as the navigation specific adjustable cutting blocks, I remember that the learning
curve was frustrating. So, when Exactech asked me to work with its team to come up with an easier way to navigate that incorporates the power of CAOS into the simplicity of mechanical instruments without extra pins I was intrigued and agreed to participate.

The initial offering from this work group is ExactechGPS TKA Plus. TKA Plus uses familiar Truliant® mechanical instruments to guide TKA Plus specific cutting guides into an initial position for cutting the distal femur and proximal tibia. The cutting blocks, once fixed to the bone, then become foundations for the arrays that allow bone registration and cut guidance. For the distal femoral cut, an intramedullary guide is still utilized. While this eliminates one of the purported advantages of navigation, it is familiar and will place the block within the range of adjustment to allow for a perfectly planned resection. Intramedullary distal femoral cuts fall outside an acceptable varus/valgus angle up to fourteen percent of the time.7,8 The ability to adjust the cut to the desired valgus angle, flexion, and resection depth will improve alignment. The block offers plus/minus four degrees of varus/valgus adjustment, plus/minus four degrees of flexion/extension, and plus four/minus two millimeters of resection. Registration is six quick points. Only the distal medial and lateral femur require painting, so registration takes about one minute.

The tibial side uses the familiar extramedullary guide. Once pinned to the bone, it takes six quick points to register without any painting. Registration is fast! The intended cut is then verified and adjustments can be made on the block before resection. Much like the femoral block adjustment is plus four/minus four degrees of varus/varus, plus four/minus four degrees of tibial slope, and minus two/plus four millimeters of resection depth.

What I have found is that I’m not that accurate with mechanical instruments. Probably none of us are.

Lab testing has shown that when the blocks are pinned with the appropriate technique (threaded headed pins) for stable, secure fixation that angular play of the block with attempted movement is only 0.2 degrees. So the trackers are stable. In surgery, I have found the system to be very intuitive. Registration and block adjustment only adds three to five minutes compared to my mechanical instruments if I don’t need to recut. Recuts do not happen when I use the TKA Plus navigation so I may save time over many cases.

What I have found is that I’m not that accurate with mechanica instruments. Probably none of us are. In a sawbones study at Stanford in which 36 tibia and 36 distal femoral cuts were made by surgeons with varying levels of training, all cuts required at least one of the three adjustments to get to the intended cut angles and depths. Many times all three parameters were adjusted. I have found a similar trend. Prior to using navigation, I x-rayed all of my tibia cuts to ensure a 90 degree cut the mechanical axis. Twenty three percent of the time I was more than two degrees off and had to recut the tibia. In my TKA Plus cases I have had to adjust at least one parameter two thirds of the time! On the femoral side I have always taken it for granted that my cut was accurate. I was wrong. Studies show that an improperly placed starting point, a femur with medial to lateral bow, or a patulous
intramedullary canal can all lead our cuts to be outside of an acceptable range.7,8 In my TKA Plus cases I have adjusted one of the three parameters twenty five percent of the time!

We all agree that CAOS makes us better. After using TKA Plus, I know it is making me better without disrupting my normal workflow. It adds very little additional time, and if it prevents recuts, will probably save time in the long run. ExactechGPS navigation already gets past many of the hurdles
to the adoption of CAOS. Its accuracy has been validated. TKA Plus takes the next step toward mainstream use of navigation by incorporating it into our standard mechanical instrumentation. Future plans with TKA Plus may incorporate sizing and femoral rotation.


1. Hetaimish BM et al. Meta-analysis of navigation versus conventional total knee arthroplasty. Journal of Arthroplasty, 2012 Jun;27(6):1177-82.
2. Jeffery RS et al. Coronal alignment after total knee replacement. Journal of Bone and Joint Surgery Br, 1991 Sep;73(5):709-14.
3. Berend ME et al. Tibial component failure mechanisms in total knee arthroplasty. Clinical Orthopaedics and Related Research, 2004 Nov;(428):26-34.
4. Collier MB et al. Factors associated with the loss of thickness of polyethylene tibial bearings after knee arthroplasty. Journal of Bone and Joint Surgery Am, 2007 Jun;89(6):1306-14.
5. Australian Orthopedic Association, National Joint Replacement Registry, Annual Report 2013 – Hip and Knee Arthroplasty.
6. Rebal BA et al. Imageless computer navigation in total knee arthroplasty provides superior short term functional outcomes: a meta-analysis. Journal of Arthroplasty, 2014 May;29(5):938-44.
7. Cates HE et al. Intramedullary versus extramedullary femoral alignment systems in total knee arthroplasty. Clinical Orthopaedics and Related Research, 1993 Jan;(286):32-9.
8. Teter KE et al. The efficacy of intramedullary femoral alignment in total knee replacement. Clinical Orthopaedics and Related Research, 1995 Dec;(321):117-21.