Will Robots Take Their Place in Orthopedics?

Editor's note: At the time of publication (Aug. 28, 2016), a family member of the analyst owned shares in Stryker (SYK) in an account managed by a third party. Morningstar has confirmed that her ownership of Stryker did not influence any ratings or analysis.

While using robots for surgery has been on the periphery for more than two decades, they have become more central to the orthopedic field over the past several years, especially with the purchase of Mako by Stryker SYK and the acquisition of Blue Belt by Smith & Nephew SNN. In the hands of these larger competitors, we expect robotic surgery to expand its application in orthopedics over the next five years. However, ambiguity around the clinical utility of robots and caution among surgeons could keep the adoption curve gradual over the next five to 10 years. There are definitely challenges to the adoption of this technology.

Robotics falls into a larger category of computer-assisted orthopedic procedures, which includes navigation, imaging, as well as cutting and milling of bone. The use of robotics first caught on in orthopedics in conjunction with the introduction of minimally invasive joint replacements. It gave a means of imaging the joint and navigating inside the body through smaller incisions to achieve more precise placement of implants. This is meaningful because it offered surgeons a way to study the joint and formulate a plan for how to prepare it for replacement before the patient entered the operating room. Conventional wisdom held that this kind of upfront planning would lead to smoother and shorter procedures, which would also be less expensive, because each minute in the operating room costs approximately $62.1

More recently, robotic systems have been developed to assist surgeons in the actual cutting and grinding of bone surfaces. Similar to the navigation systems, the concept is that robotic milling would allow for more accurate planning and a better fit of the bone to the implant. Better fit should ultimately translate into a better clinical outcome that minimizes the risk of misalignment or suboptimal placement that can cause pain, leave the patient with differences in limb length, or introduce wear patterns that, over time, cause the implant to loosen and fail prematurely. There is also the belief that the precision that robotic systems offer can help compensate for suboptimal surgeon skill, leading to more predictable and reproducible outcomes. Considering orthopedic surgery remains one specialty where individual surgeon skill wields an outsized effect on outcomes, practitioners are interested in ways to raise the skill level across the board.

There is much controversy around robotics and computer-assistance among surgeons. At first blush, it appears the field is divided fairly evenly between those physicians who believe robotics is critical to the future of orthopedic surgery and those who remain skeptical of the clinical value and significant cost of these systems.

Challenges to Widespread Adoption With little consensus among practitioners, we see considerable hurdles to the widespread adoption of robotics in orthopedics. As with much in orthopedics, the research on clinical outcomes or any related cost savings is thin at this point. Data convincingly suggest that navigation systems do indeed lead to more accurate placement of joint implants. However, researchers have not yet been able to demonstrate how robotics leads to any difference in clinical outcome. That is, we're seeing no difference in functionality or patient satisfaction whether robotic navigation is used or not. The jury is still out on whether robot assistance with bone milling can lead to significantly improved outcomes. It could easily take another five to 15 years to collect enough clinical data to move practitioners to consensus on the value of robotic systems such as Mako.

The absence of compelling clinical data leaves considerable obstacles to establishing reimbursement for special navigation or robotic assistance. The financial aspect of robotic systems can be a tough pill to swallow. For example, the Mako robot (which assists the surgeon in cutting and milling) is priced close to $1 million, with another $100,000 in annual service costs. But the reimbursement for hip and knee replacement does not include anything extra for the use of a robot. We do not expect this to change until there is enough consistent clinical data to sway consensus.

Moreover, the application of robotic technology in other sectors, such as industrial manufacturing, has typically led to efficiencies in terms of lowering labor inputs and lowering the time required to complete the product. Thus far, robotic systems in orthopedics has delivered neither benefit.

First, surgeons are wary of "outsourcing" their own labor to technology. This explains why most of the current robotic systems were developed to only assist the doctor. Unlike Lasik eye surgery, where the machine and its software actually perform the procedure without any active help from the physician (once the machine has been programmed and the patient has been prepped, the ophthalmologist can step out of the room), most orthopedic robotic systems are under the complete control of the surgeon during the entire joint replacement procedure. We imagine there would be significant resistance from surgeons to any innovation in robotics that would allow the machinery to perform autonomously.

Second, there is a large learning curve with robotic systems in orthopedics. Aside from early adopters, many surgeons are reluctant to invest the necessary time in mastering the robot. Finally, the added challenge of involving a robot in the arthroplasty also lengthens the time the procedure takes, thereby making the whole joint replacement more expensive. Studies suggest that robotic joint replacement takes an incremental 15 to 35 minutes. So far, we have seen little evidence that the robot increases time efficiency even after surgeons become familiar with the technology.

The one solution to address all of these hurdles to widespread adoption is high-quality clinical data demonstrating that the robotic system leads to better clinical outcomes and lower costs. Favorable data would convince more surgeons that climbing the learning curve would be worthwhile and persuade payers that reimbursing for the robot is justified. Because this data is still far off, the main drivers of robotic adoption will be a combination of hospital demand fueled by interest in advertising its robotic capabilities to attract more patients and aggressive marketing by Stryker and Smith & Nephew to place robotic equipment. We anticipate these two factors can drive steady adoption through until critical mass of clinical data can weigh in on the value question.

While the robotics field had mainly been populated with smaller companies pioneering these technologies and adoption has been spotty, the market is now poised for accelerated adoption, in our opinion, as major implant makers embrace the technology. Two of the three major competitors in robotic systems are owned by Stryker, after its purchase of Mako, and Smith & Nephew, which acquired Blue Belt Technologies. The two systems are somewhat different, and importantly, Stryker has a significant advantage with its array of indications, in our opinion.

Stryker has made aggressive efforts to expand indications to include total hip replacement and total knee arthroplasty. This means Stryker is in a much stronger position and offers application of its robot across the lion's share of joint replacement procedures. The versatility of Mako is likely to make it more appealing to hospitals weighing the trade-offs between the sizable capital investment, physician interest in acquiring the robot, and the robot's potential for swaying more patients to the hospital.

Studies suggest that total knee replacement performed without a robot can result in bone alignment that is not ideal up to 40% of the time.2 On the other hand, the studies have consistently shown that robots can reduce the number of those mechanical axis outliers. Unfortunately, it is not clear whether achieving ideal alignment translates into superior clinical outcomes. Nonetheless, early adopters of this robotic technology believe that the reduction in outliers, in and of itself, is an important benefit. There is widespread recognition that surgical skill is not consistent and if robots help the less-skilled surgeons improve their outcomes that would be positive for the entire field.

Another proven benefit of the robotic systems is their ability to mill the bone far more precisely, so as to eliminate or minimize gaps between the bone and the implant. This is particularly critical for cementless implants, which need as much bone-to-implant contact as possible for the bone to grow into the implant and stabilize the joint. The more gaps there are, the longer it can take for the bone to grow in and for the patient to reach stabilization, and more opportunities for something to go wrong during the delay in bone growth that could spur instability. Conventional wisdom holds that joint instability is a key factor that can bring about premature failure of the implant. Again, it is not clear whether this difference will ultimately be born out in significantly different clinical outcomes.

Given that clinical studies on outcomes will be slow to emerge, we think adoption of robots can continue and grow a bit stronger at least through the midterm, thanks to expanded indications and Stryker's extensive relationships with hospitals. At a certain point, once there is a critical mass of robots installed, we expect there to be more pressure on the early and late majorities to jump on the bandwagon, if only because their hospitals will not want to be left behind in the competitive dust. As the robotic systems become more widespread, we think the fear of losing the medical arms race could motivate more hospitals to purchase robots in order to fight for patients.

Nevertheless, reimbursement is the one thing that is unlikely to budge before compelling evidence on clinical outcomes is collected. This means that hospitals will need to assess how strong surgeon demand is, the likelihood that the robot will attract new patients, and the higher costs in terms of more imaging and longer procedure time bumping up against limited reimbursement. We suspect the larger hospital systems and research-oriented providers will be more willing to invest in robotics. In the meantime, we're keeping an eye on practitioner perceptions of the technology and Stryker's ability to steadily place more robots over time.

Stryker Positioned to Expand Moat It is not entirely clear what role robotic technology could play in enhancing or detracting from the moats of orthopedic manufacturers. We think it will depend on several factors, including how well the technology evolves to compensate for below-average surgical skills.

We think Mako could help reinforce Stryker's moat as more robots are installed because it is a closed system and works only with Stryker implants. After an initial period of adjustment during the integration of Mako, Stryker has made consistent headway in placing more Mako robots into hospitals. If the clinical evidence weighs in favor of Mako's ability to deliver significantly improved outcomes, surgeons who currently do not use Stryker may seriously consider training on Mako and using the corresponding Stryker implants and instrumentation. However, we recognize this is a big if and that we are a long ways from any definitive indications on the value of Mako's robot. Until we reach that point of consensus on the clinical evidence, we think Mako is largely neutral for Stryker's moat.

For Smith & Nephew, the Blue Belt Navio robot can work with any brand of implant and instrumentation. Assuming the long-term clinical data is favorable, more surgeons may pursue training on the robot and incorporate it into their procedures. But this may not make users any more likely to choose S&N reconstruction devices, especially if they are not stocked widely at hospitals. We doubt the Blue Belt robot holds much potential to enhance Smith & Nephew's narrow moat.

While we view the current robotic systems as generally neutral from a moat perspective, improved robots--including more-autonomous technology--could eventually lower the switching costs that contribute so heavily to orthopedic manufacturers' moats. If robotic technology evolves to such a point that it can easily deliver higher-quality outcomes consistently, thereby compensating for individual surgical skill that is the result of frequency and years of practice on a particular brand system, switching costs may come down (especially if each competitor offers its own robotic system). If it ultimately becomes easier to master the robot than manual use of the instrumentation--and outcomes are better--this could strike a blow against the moats. We suspect the manufacturers are cognizant of this and treading carefully. We anticipate that top-tier competitors will want to make their robots proprietary and maintain a sizable learning curve to discourage greater brand switching.

Next 10 Years of Innovation In addition to robots, other emerging orthopedic technologies could be disruptive if safety and efficacy are established. Although these new developments are unlikely to become influential commercial forces in the near- or mid-term, we expect that some could shift the competitive landscape over the longer term.

Biologics and Regenerative Medicine As the march of science advances, we have begun to see some inklings of the potential that biologics hold for regenerating cartilage. This quest is a holy grail of sorts in medicine, much like creating synthetic blood. We anticipate more progress on regeneration over the next two to five years. If regeneration becomes commercialized over the longer term, we expect that the biologic intervention for acute damage could help more young patients avoid the high risk of premature onset of osteoarthritis.

3-D Printing for Orthopedics We have seen considerable progress in the application of 3-D printing on orthopedics. The major orthopedic competitors have jumped onto the bandwagon and begun exploring how to incorporate the technology into their portfolios. Unlike other organ systems that involve mainly vascularized soft tissue, orthopedics typically uses engineered plastics and metal alloys, which are both more conducive to manipulation through a printer. So far, most of the major orthopedic manufacturers are dipping their toes in the 3-D water by pursuing new products. We have not seen any effort to fully replace existing off-the-shelf implants with customized devices. There are still technical issues that need to be resolved from a quality and regulatory standpoint before such precision-machined and milled implantable products are ready for permanent use. However, with all the device makers working on applications for 3-D printing, we wouldn't be surprised if the field ultimately moves in this direction.

Sensors We are optimistic about the potential applications of sensor technology to orthopedics, especially to help with the early detection of bacterial infection of the implant. Considering the "stickiness" of biofilm that develops around implants after cells irreversibly attach to the implant and each other, the task of ridding bacterial infection that hides in the biofilm is challenging to say the least. The situation is all the more difficult because patients may not experience any symptoms of joint infection, especially if it takes place years after the reconstruction.

Researchers are also working on sensors that can be embedded in the implant to detect early signs of loosening. Loosening of the implanted materials is typically not detected until the later stages when clinical symptoms appear. At that point, there may be substantial bone loss, which makes any revision surgery challenging. It is hoped that sensors can offer enough sensitivity that loosening can be monitored, and intervention can take place before loss of bone presents new risks to the patient.

Clinical Evidence Is Key Robotics should gain ground through the midterm thanks to aggressive marketing by Stryker and Smith & Nephew, as well as adoption by hospitals eager to attract new patients. However, clinical evidence of robots leading to improved clinical outcomes or cost savings remains thin. If the clinical data is ultimately unfavorable, demand for robots could easily fall. We project the installed base of orthopedic robots to grow from 368 in 2015 to more than 1,000 by 2020, with Stryker holding roughly 60% market share. If clinical data establish significant advantages linked to robot usage, there could be upside to our estimates.

Endnotes

1 Macario, A. "What Does One Minute of Operating Room Time Cost?" Journal of Clinical Anesthesia, Vol. 22, 2010.

2 Song EK, Seon JK, Yim JH, Netravali NA, Bargar WL: "Robotic-Assisted TKA Reduces Postoperative Alignment Outliers and Improves Gap Balance Compared to Conventional TKA." Clin Orthop Relat Res 2013;471(1):118-126.

This article originally appeared in the August/September 2016 issue of Morningstar magazine.