UW Shoulder and Elbow Academy: Robotics in Shoulder Arthroplasty: ASES Podcast 156.

Robotics in Shoulder Arthroplasty is a timely and informative ASES podcast hosted by Peter Chalmers and Brian Waterman with two guests: Michael Freehill and Vahid Entezari.

Here is a summary of the points the participants made. Where the guests referred to a specific study, a link to the primary source has been added so that readers can consult it directly and form their own view.

The episode presented enthusiasm tempered by candor about the current state of evidence. Both guests described themselves as early adopters, and both acknowledged that outcome data specific to robotic shoulder arthroplasty is not yet available. Their case for the technology rested primarily on precision in executing a preoperative plan, the ability to generate intraoperative data, enhanced capability in complex and deformed anatomy, and anticipated — as yet undemonstrated — long-term benefit. They also included a discussion of the market and workflow factors that shape adoption.

Here are some details.

Dr. Entezari suggested that shoulder arthroplasty has two distinct steps — planning and execution — and that while preoperative planning has advanced considerably, precision execution has historically lacked dedicated tools. He referenced previous work on the accuracy of planning and patient-specific instrumentation (PSI) showing that preoperative planning and PSI tightened the spread of deviation from the preoperative plan in comparison to manual techniques. See Accuracy of 3-Dimensional Planning, Implant Templating, and Patient-Specific Instrumentation in Anatomic Total Shoulder Arthroplasty.

From this he drew two conclusions: that surgeons doing manual-only surgery may underestimate their own deviation (“you don’t know what you don’t know”), and that enabling technologies — navigation, PSI, or robotics — become especially relevant in cases with significant deformity or difficult exposure. These technologies have the potential to reduce substantial deviation from the preoperative plan.

Dr. Freehill built on this from a training perspective, recalling high-volume leaders he trained under questioning their own accuracy: whether their humeral head cuts and glenoid version were as accurate as they desired. He described preoperative planning as a major advance and PSI as a further step for deformed anatomy, while noting that guides still have limitations related to soft tissue and exposure. He also emphasized that most shoulder arthroplasties are performed by lower-volume surgeons — those doing fewer than roughly ten a year — and suggested these surgeons might benefit from tools that help with execution.

Both guests placed the clearest value of robotics in complex cases: significant glenoid deformity (such as B2 and B3 morphology and higher degrees of retroversion), revision situations, and cases requiring bone graft or augment preparation.

Dr. Entezari described a reverse arthroplasty for arthropathy that had developed after a dislocation and fracture, with anterior glenoid bone loss, in which the robot was used to shape a humeral head autograft to fit the glenoid face, offering it as an example of capabilities that could extend to graft and augment work.

The guests pointed to potential new capabilities: a robotically controlled burr that can sculpt bone at angles a saw or reamer cannot achieve, the potential for subscapularis-sparing and minimally invasive approaches. They pointed out that robotics could capture potentially useful data such as how precisely the preoperative plan was carried out.

The hosts pointed to the mixed hip and knee literature, noting meta-analyses that have not shown a long-term benefit in clinical outcomes even though the rate of use of robotics was rapidly increasing.

The guests did not claim superiority in patient reported outcomes for robotics in shoulder arthroplasty. Dr. Entezari suggested that malpositioning may not affect these outcomes but that adherence to the preoperative plan may matter for long-term implant survival.

Dr. Freehill cited NICE’s assessment of robot-assisted orthopaedic surgery. The relevant NICE document is its early value assessment of robot-assisted surgery for hip and knee replacement (HealthTech guidance HTG743) in which patient-reported outcomes and complications were similar between robotic and conventional surgery and implant alignment was consistently more precise with robotics. The revision data were limited, and the committee was uncertain whether more precise alignment yields better clinical outcomes. He pointed to the need for head-to-head studies, while acknowledging such studies are difficult and slow to complete.

Dr. Entezari cited a Monte Carlo simulation whose authors concluded that, when value is judged by tail-risk rather than by mean accuracy, the principal contribution of robotic assistance in shoulder arthroplasty is risk containment — with clinical and economic implications that remain unproven to date — and that further research linking its use to patient outcomes is warranted and may ultimately redefine the technology’s value proposition. See Mako Robotic-Assisted Glenoid Preparation in Reverse Shoulder Arthroplasty: A Tail-Risk Reduction Perspective Compared with Manual and Patient-Specific Guide Techniques.

With respect to the learning curve for robotics, Dr. Freehill cited experience suggesting an inflection point around ten cases, with time efficiency approaching break-even, and drew an analogy to hip and knee arthroplasty, where operative time tends to normalize with experience.

[Parenthetically, we can observe that the inflection point of ten cases is close to the annual case volume of a “low-volume surgeon.”]

Both guests suggested the real barrier to adoption is neither time nor cost but workflow change.

Dr. Freehill discussed how the technology is presented to hospital administrators, whom he characterized as focused on time-zero costs rather than on anticipated ten- to fifteen-year outcomes or on accuracy data. He noted that an administrator is unlikely to be swayed by work comparing robotic glenoid preparation favorably with patient-specific instrumentation and, still more, with manual technique, because their attention stays on the bottom line. He acknowledged the numbers work against robotics: citing the existing literature, he said that breaking even by minimizing revisions would require avoiding roughly one revision in fifteen (about 7%), whereas revision rates in reverse arthroplasty are currently under 2%, so on that basis one is, in his words, already arguing against the cost-benefit case. See Is Robotic-Assisted Reverse Shoulder Arthroplasty Economically Justified? A Break-Even Analysis. He suggested the more persuasive pitch to an administrator combines several elements: that a robot from certain companies is often already in the hospital for hip and knee procedures; that adopting cutting-edge technology is anticipated to improve market share; and that greater accuracy can be framed as serving the patient’s long-term interest.

Dr. Entezari added that hospitals in a region compete for the same patients, and that device companies invest heavily in direct-to-consumer advertising, so patient demand — already seen in hip and knee — becomes part of the rationale. He noted that costs should fall as more surgeons use the technology. He reiterated that outcome data are not yet available and that early adopters have the opportunity to generate the needed data.

The guests contrasted navigation and robotics. Navigation has a smaller footprint, lower cost, and a lower barrier to entry. Robotics offers greater precision, but with a larger and more expensive early-generation factor, and with current systems limited in the procedures they can perform and the range of implants they can accommodate.

Dr. Entezari suggested neither overselling nor underselling the technology — telling patients that it makes the surgery more precise while being candid that its effect on their individual outcome is not known. Dr. Freehill described a mix of curiosity about cutting-edge technology and hesitation about a non-human element in the operation, and said he uses clinic time to explain the surgeon’s central role.

Both guests urged caution about how the technology is learned. They argued that younger surgeons should first become proficient in manual arthroplasty, comparing the situation to skills that have faded elsewhere in orthopaedics, such as freehand pedicle screw placement, manual knee cuts, and open shoulder stabilization.

Dr. Entezari described losing registration during a case after contact with a tracker and having to rely on his manual skills to complete the operation. Dr. Freehill echoed the concern that trainees might become proficient with the technology without ever learning the underlying skills of standard shoulder arthroplasty.

Conclusion: this was a balanced presentation that emphasized the potential role of robotics in transferring a preoperative plan to the patient in the operating room, without claiming that robotic shoulder arthroplasty has been associated with improved patient outcomes.

Which is better?