In the Hands of a Robot

Orthopedic surgeons make a living with their hands. Their eye hand coordination skills and their ability to cut, grasp or ream with preternatural three-dimensional precision is often the difference between a patient whose life is defined by pain and one that is not.

With those hands the average orthopedic surgeon generates, on average, about $6 million for their hospitals/clinics/implant suppliers and….at the end of the pipeline…themselves. Considering the value of returning workers to their jobs orthopedic surgeons probably generate ten times that number annually, on average, in pure economic value.

How much of this great value is due to the physicality of their profession and how much is due to its intellectuality?

New data indicates that the physical demands of being an orthopedic surgeon is both more variable and, ultimately, more important to determining outcomes than perhaps many realized.

Robot-Assisted Surgeons Significantly More Precise

At the upcoming Computer Assisted Orthopedic Surgery annual meeting, five investigators from Scotland plan to report that human surgeons can increase their precision significantly when assisted by robots.

The study authors—Mark Blyth, M.D., Bryn Jones, M.D., Angus MacLean, M.D., Iain Anthony and Phil Rowe of the Glasgow Royal Infirmary and the University of Strathclyde in the UK—used the unicompartmental knee arthroplasty (UKA) procedure to test the incremental effectiveness of robots. They randomized 100 patients to either a robotic-assisted UKA or a non-robot-assisted UKA (n=50 in each group).

The UKA procedure is renowned for its technical difficulty. In a 2010 review of the literature comparing high tibial osteotomy to UKA, the authors stated: “Aspects which most likely affect the outcome of UKA are prosthetic design, alignment and stability. Robertsson et al. stated that surgeon’s confidence with this technically demanding procedure markedly influenced the outcome. They showed that hospitals performing less than 23 UKAs per year reported a revision rate 1.6 times higher compared to that of departments performing a higher volume of UKAs.”

Later in that same review, the authors noted: “Barrett and Scott reported on 29 failed first generation UKAs that were converted to TKA [total knee arthroplasty]. The mechanism of failure was loosening in 55% of the cases and progression to degenerative disease in the opposite compartment in 31% of cases. The average time to revision was 47 months (range 4 – 113 months). In 66%of the failures, the authors identified technical errors at the time of the initial UKA or poor patient selection.”

50 Versus 50

In this corner, humans. In the opposite corner, humans + robot. Each has 50 patients and the patients are randomized to receive either the robot assist or the pure, unadulterated human.

In this study three surgeons performed the UKA surgeries. Each one contributed equally to each patient group.

We would also make the point at this stage that, in addition to the “robot” assist device, the modern operating room is chocked full of electronic and mechanical assist devices. Here is a photo of the modern OR. One might be tempted to ask if, in the midst of the other surgical assist equipment; is any room left for the operative team?

So, for perspective, it is instructive to understand that these devices are part of a long line of the surgical assist technologies.

The 50 Non-Robot-Assisted UKA Patients

Three months post-operatively, the 50 non-robot-assist patients underwent CT scanning of their knees. The CT scans were used by the researchers to calculate varus/valgus, flexion/extension and internal/external rotation parameters. Those calculations, then, were compared against the patient specific target values that the surgeons set pre-operatively.

The 50 non-robot-assisted patients were, it seems, generally pleased with their surgery at three months. We say that because the AKSS score (the AKSS score is the American Knee Society Score, which is a widely used functional outcome score for knee arthroplasty) was 99 before surgery and 142 after surgery, a 43% improvement in just three months. An AKSS score of 142 is considered to be good. In fact 26% of the non-robot-assist group reported an “excellent” AKSS score (160 or better) at, again, just three months post-operatively.

The 50 Robot-Assisted Patients

The 50 robot-assisted patients, however, posted up better scores than their counterparts. The robot-assisted AKSS scores averaged out to be 166 at three months post-operatively. That score is about 24 points higher than the non-robot-assist score (+17%). And when compared to their AKSS score pre-operatively, the robot-assisted group reported a 63% improvement (versus 43% for the non-robot-assisted group).

It’s only three months, of course. Eventually, like over two years, the two group scores could come closer together. But…when the researchers dug into the precision of the two types of surgeries…they found that the robot-assisted surgeons were 48% more precise in their implant placement and bone reaming.

Did that higher level of precision create the strong outcome differences between the two groups as early as 90 days?

“’It is not clear at this point the reason for the differences in early pain scores and outcomes between the robotic and non-robotic groups. Whilst greater precision is one potential cause, there are other factors such as absence of intra-medullary instrumentation and limited depth of tibial resection that may favour the robotic group, ” said lead investigator, Dr. Blyth. “We are continuing to follow these patients in the hope that these short term differences remain at longer term follow-up.”

Could such improved precision deliver better long term results?

“Yes, precision has been shown to deliver long term benefits in term of survivorship of the implant, ” said Dr. Blyth, “But demonstrating improved function in the short or longer term hasn’t been demonstrated using the CT methodology that we used. We hope it will in the long term but can’t say that right now.”

The Data

Source:  Blyth et al.; Accuracy of Unicompartmental Knee Arthroplasty (UKA) implant positioning and early clinical outcomes in a Randomised Controlled Trial comparing Robotic Assisted and Manual Surgery

Which Assist Device Did They Use?

The investigators used MAKO’s RIO haptic robot where the surgeon controls an interactive robot arm. Half the patients in the study received Biomet’s Oxford Unicompartmental Knee manually. The other half received MAKOplasty. (The following graph is a comparison of MAKO and Manual Precision.)

MAKOplasty is the name for robotic arm assisted partial knee resurfacing. It selectively targets those parts of the knee that have been damaged by osteoarthritis. MAKOplasty can be performed through a four to six inch incision with additional small incisions in both the femur and tibia.

In addition to precision, the MAKOplasty procedure preserves more of the patient’s natural bone and tissue.

MAKOplasty is also an increasingly valuable marketing vehicle for hospitals and physicians. Hospitals with MAKO market it fairly aggressively to patients saying, in effect, that by using MAKOplasty their staff is more innovative and precise than non-MAKOplasty hospitals. Their hope, of course, is to attract more new patients.

MAKOplasty is indicated for patients with early to midstage osteoarthritis in the medial, patellofemoral or lateral compartments of the knee. Powered by the RIO Robotic Arm Interactive Orthopedic system, the surgeon, as we saw in this study, can consistently improve his/her precision.

Also increasingly, because of MAKOplasty, patients are able to be treated on an outpatient basis.

Handmade Precision

Ultimately medical decisions are driven by patient outcomes and economics. Much as IBM’s robot has finally mastered the greatest chess minds in the world, so to, it appears, computer driven, haptic controlled robotic assist devices appear to be claiming a key role in the surgery center because, by being more precise than the surgeon’s hands alone, they are improving outcomes and, ultimately, the economic value of such interventions.