“Digital OR” wasn’t a well-known term back in 2001 when a surgeon conducted the first transatlantic gallbladder procedure. It was also the last transatlantic gallbladder procedure, due to technical limitations that prevented safe and practical replication.
We’ve come a long way since then, said surgical robotics innovator Dr. Fred Moll, co-founder of Intuitive Surgical and chief development officer of Johnson & Johnson’s Auris Health.
He recently discussed advances in surgical robotics, AI and technologies enabling cutting-edge treatment in the digital operating room with Avail Medsystems founder and CEO Daniel Hawkins.
Hawkins launched the marketing and sales department for Intuitive’s Da Vinci surgical robot, later co-founded and led Shockwave Medical as CEO, and left to start Avail Medsystems, which makes an operating room telepresence system that lets surgeons collaborate in real-time with experts around the globe.
“We’re still not doing telesurgery the way that I think it can be done and will be done,” Moll said. “It really has demanded the development of a lot of capability and technology that just wasn’t around. And I think one of the things that Avail represents is the ability to take that step toward reliable communication that can safely integrate itself into doctor-to-doctor communication whether they’re six feet away or 600 miles away.”
Here are some highlights from the conversation, which we’ve lightly edited for space and clarity.
Hawkins: The evolution of Intuitive started with a robot and over time that got more and more into clinical practice. And then there’s this realization that that robot now has computing power you can bring your technologies into, and then there’s starting to be this ecosystem around that robot. How do you think about the evolution of robotics, the ecosystem, and where we’re ultimately headed in terms of digital enablement?
Moll: It’s interesting to think about the progression of the technology. Obviously, the technology has two pieces — hardware and software — and it’s easy to say it’s not about the hardware anymore. It’s really about the software and enabling the software to do more things. In 1998, the Intuitive robot was not an intelligent device. It was a capable device because it gave the clinician the ability to do things in a way that they were unable to do in a minimally invasive fashion. But the opportunity to create true intelligence of a robot is something that obviously has everything to do with the software, but there’s a necessary component of the hardware as well. The progression is a very interesting sort of dance between what’s the minimum hardware capability that you need in the OR to then advance the intelligence, the software that really takes the capability of that hardware to a whole different level. The sky’s the limit. When you start talking about AI and ML and the future of those capabilities in surgery, it gets down to, OK, what is the surgeon going to do and what is the machine going to do? And where is that bright line past which you don’t want to cross, because at some point it isn’t about capability. It’s about knowledge and wisdom that make sure that you’re doing the right thing on the right patient for the right result. We’re entering that phase. Colonoscopy is a very simple procedure that absolute positively could be automated today. The very real data around why you would want to do that has to do with the software programs that tell you something about a manual colonoscopy and how well you did, what spots you missed, what you could have focused better on, what you might want to go back and look at. That’s extraordinary. That is a teaching tool that can be real-time. And if you combine that with a robotic technique, you really have the ability potentially to automate a procedure that doesn’t require — and this is kind of where you have to obviously start — doesn’t require an enormous amount of fine techniques, but can lead to a situation where you have dumbed down the technique enough and you’ve increased the intelligence enough so that you have a procedure that potentially could be handed in a diagnostic format to a physician’s assistant with oversight from a gastroenterologist. That’s a huge societal change that maybe we’re not ready for, but it’s certainly the technology is knocking on the door of that capability. It only gets more interesting from there.
MassDevice: How technology is changing the operating room
Hawkins: As technology leaps forward, the need to bring that technology into clinical practice then requires an enormous amount of technical support, clinical support, workflow change. And as an industry, we can push tech forward, but the reality is it isn’t going to get taken up unless we support the devil ouf ot that. Those people don’t grow on trees. It’s really, really hard to be able to scale those things, those kinds of capabilities and bring clinicians from “That’s really cool” to “I’m using it every day.”
Moll: In the early days at Intuitive — little known fact — we said prostatectomy rather than heart surgery might be a pretty interesting target. When we started in procedure development, the first prostatectomies were done locally in the East Bay, California, by a little-known urologist who was in private practice and thought it was a good idea to try and use the robot to do, one of the procedures he was doing in an open format. He got extraordinarily excited about it, but he was a guy in local practice that had no ability to communicate what he’d done, why it was a reasonable thing to do and how he did the procedure. In other words, there was no ability to export he knowledge that he had gained as to what’s possible with a robot. How do you tell the world? You can write a paper and wait for it to be published, or you can talk about it at society meetings, but what you can’t do is instantly communicate a technique in real-time to another clinician. It goes directly to what are the possibilities with a system like Avail and what’s possible in communication now, and how does that drive new techniques and propagate them throughout the surgical community.
Hawkins: Fundamentally, there are limitations inside of the medtech organizations that are launching these products to be able to scale because you don’t have access to the people, they’re mismatched, the geography, and very often you need some folks in the room that are able to run the equipment, and separately, you need reps in the room to be able to sell whatever widget it is, if it’s an implant or what have you. And that turns out to be a scalability problem and an economic problem that we sought to solve. That’s why I left Shockwave to start Avail Medsystems.
Moll: A great example of that is I was on the board of Mako from the beginning, creation of it through commercialization, and the reason why Mako is owned by Stryker today is although we were having great success clinically, when it got down to how do we make money on this robotic technique and these procedures, we could never get around the fact that the procedure itself and the technology required a rep in the operating room for every procedure. The assessment of the team was we’re not getting around that anytime soon. It was a board meeting I’ll never forget, where everybody looked at each other decided we need to sell the company because we don’t have a pathway to making money here. It’s a very good example of the cost associated with trying to bring new, complex technology into the operating room and make it stick, make it safe, make it effective and propagate and build a business. An Avail-type platform gives the ability to solve that problem differently. Our solution at Mako was to give it to a large organization that can better afford and be more efficient about rep deployment in very technical procedures.
Hawkins: There used to be 5,500 hospitals and 250 ambulatory surgical centers (ASCs) at the beginning of my career. Now there’s about 5,500 hospitals and 12,000 ASCs by some counts. How do you commercially scale and how do you operate in that environment when average sales prices are going down, procedure volumes are barely going up, the middle of your income statement is stacked full of a whole bunch of non-moveable assets called people, and the number of places they need to go sell just tripled or nearly tripled. It’s an impossibility. Now, add technology on top of all of that, everything Fred’s describing of the need to be in the room to digitally enable those procedures. And it’s a mess. It’s an absolute mess. As a medical device industry, how do we enable the clinical benefit of the technology we’re creating if we can’t actually scale it?
Moll: Every technology and every procedure and every specialty is a little different with regard to need. And so if you look for a poster child for the digital OR, I think you tumble pretty quickly to the example of the treatment of stroke because we all know time is brain, right? The communication associated with it, what that patient needs, how is that patient going to get appropriate treatment, and how are you going to do it in a timeframe that makes it relevant, because if you wait too long, doesn’t matter what you do. Better communication and better transmission of technique — however you do it from one place to another — are going to really revolutionize the treatment of stroke. Thrombectomy has been an enormous success. It still has a problem of who’s going to do it and how are you going to get appropriate capability to that patient in time. The revolution is going to happen, but it demands not only the right technique that has been developed, but the right communication capability, the right technology to enable a remote capability that the world is comfortable with. And then, I think it’s going to involve robotics. … Technology will play an increasing role in the OR. How do you connect clinicians, and in many ways really merge that connection with digital information and the use of AI and the capability of a combined planning, decisionmaking and learning cycle that really begins in the collection of data that’s turned into information, that’s turned into knowledge, that’s turned into wisdom? And the cycle of gathering the information and recycling it to affect the patient journey is an obvious halfway. Everybody wants to talk about AI, but I do think it is one of the keys to unlock that merging of individual knowledge, communication and betterment of technique and procedural learning that is going to continue forever.
Hawkins: We’re all familiar with digitally enabled knee implants. Now you’re watching what happens in the patient afterwards. Let’s take that data and let’s bring it back through the cycle to the next time you do it. It’s better, right? We have opportunity to advance logarithmically from a digital enablement capability. Some of the things that I’ve been exposed to, I look at and say, wow, what we can do clinically is going to be 1,000 times better in a decade. How are we actually going to implement? How’s the industry itself going to enable that technology to be commercializable and then commercialized and integrated into workflow? How are we going to pay for it? How are hospitals going be able to absorb all of that expense? All of that goes to, as an industry, can we solve the implementation problems that bring out some of the cost structures so we can actually bring those technologies all the way forward. Ultimately, it becomes an integrated, connected kind of world. Clinically, what we can do will dizzy us in a handful of years compared to where we are today. But in the middle of all of that, we need to put in the business realities of how the heck are we actually going to get all of that done. And I think we need to be very responsible as an industry. Fred alluded to this a little bit earlier. The line between what the machine does and the clinical team does, where’s that line drawn? It’ll keep on going closer and closer to the machine, but at some point there’s a question: Can you go all the way over? Just because we can, should we?
Moll: It’s not perfect, but I sometimes use the analogy to the progression in technology of autonomous vehicles. You see man and machine making decisions together, the driver taking input from what are now smart cars, to a progression to autonomous driving. I think everyone thought at the beginning of the journey of how do we take the driver out of driving that there was sort of a linear pathway to technology getting better and better and it’s going to be a no-brainer to put the passengers in the back seat and let the car drive itself. I think the learning now is the last 10% of that journey is really hard because, again, it gets down to judgment in situations that are not predictable and quite frankly, you can’t plan for. Is that going stop that technology in its tracks? I don’t think so. But I think it is analogous to you get to a certain point — yes, you can do a colonoscopy with your eyes closed or with a physician’s assistant or somebody calling in to check up on how it went and what else might be done for that patient. But you start progressing into more complex capability and it gets to the point where judgment and wisdom are at a premium to make sure you have the appropriate balance between technology and human input. The future of using information to change a clinician’s technique is going to accelerate. … I think it’s really important, obviously, for surgeons to continue to be classically trained in a way that makes some of the best clinicians and technicians that they can be in this period of time. But they no longer have the luxury of saying, “I’ve learned enough. I’m gonna keep doing what I’m doing.” They need to be technically enabled. They need to be robotically enabled, and familiar with how to keep up with what I think will be an accelerating need to understand what’s new, what’s next.