 Our next speaker is Dr. Alex Langerman. Alex is assistant professor of surgery and a junior faculty scholar at the Bucksbaum Institute for Clinical Excellence here at the University of Chicago. Alex is an otolaryngologist trained in head and neck surgical oncology and microvascular reconstruction. His ethics related research surrounds decision making for head and neck cancer, surgeon communication with patients and families and colleagues. Dr. Langerman completed both medical school and his surgical residency here at the University of Chicago. I have to say I have known Alex for many years. Back before he ever went to medical school he was a research assistant who did some work with me and we published a paper in 2000 that he probably doesn't even remember but I found on my CV that is opinions and use of advanced directors by physicians at a tertiary care hospital. So Dr. Langerman's been interested in these topics for a very long time and has really developed a tremendously creative area of research. So Alex. Peter, thank you. And that was actually my first academic publication and in academics as in love you never forget your first time. Thanks everyone for being here this morning. I appreciate it and to the McLean family, the center, Mark, Peter, everyone for getting up early this morning. Thank you for being here. So let's see if this works. Great. So the operating room. Fun place to study. It's sort of unique in that there's a highly technical component to it. There's lots of expensive technology and the assistance we use, you know the very cutting edge of medical knowledge but at the same time it's fundamentally a manual skill and so there's some interesting dichotomy there. So the surgeon meets the patient in clinic or in the ER, wherever they are, and then they shepherd them through the operative sort of life cycle into the operating room and out. And because they're the one that sort of the patient has created this compact with, I believe, is really the fundamental reason why surgeons have a certain primacy in the operating room as far as decision-making goes. And again, Charles Bost talked about the concept that there's this quasi-normative culture here so one surgeon wants to do it this way, another surgeon wants to do it that way and that's okay, even if it's the same surgery. Looking at our surgeons here performing one particular complex operation called a Whipple Procedure or Pancreaticode Duodenectomy, you can see that here's the textbook up top and then A, B, and C are three different surgeons all ostensibly doing the same operation here at the University of Chicago and all of them do it differently. Different orders, different steps, some include some things, some include other things. And so you say, well, gosh, maybe a Whipple isn't a Whipple, isn't a Whipple. And so all of those studies saying, well, and patients in arm A got a Whipple and arm B got chemo-radiation or something along those lines might not be getting the same intervention. What do we have as a record of those interventions? Well, we have the operative report and the operative report is inherently subjective so they're often incomplete and poor quality. They will often not have a high sensitivity or specificity for certain important events in the operating room. And in fact, if you combine observers in the operating room and then look at the operative report, well, sometimes they gloss over some important issues or maybe some complications. So maybe that isn't what we should be relying on as our only record of what goes on in the operating room. We looked at just neck to section operative reports and found that there was often information missing that would even affect a patient's downstream care as far as the medical or the actual surgical intervention. And we're in good company. Plenty of studies have done this as well. Of course, the surgical, you know, what you're sort of the narrative of the surgical operation isn't the only thing that affects patient care in the operating room. There's all sorts of other factors that go into it. Communication, competing tasks, the experience, whether or not someone's coming in and out of the room, whether there's a phone call, whether it's a hard part of the surgery, whether someone's exhausted when they do that and all that actually affects patient care and theoretically could be measured. This is a wonderful landmark study and this is the third component, which is sort of the actual skill of the operator. So in the Michigan Bariatric Surgery Collaborative, this is a longitudinal study of patients undergoing bariatric surgery and they follow them out afterwards and look at their complication rates and they have over 5,000 cases in this. And so some of the participating surgeons were asked to send in a video of what they felt was a representative video of one of their bariatric surgery cases. And then these videos were edited down to be kind of short in particular steps and then they were rated by an expert panel to say who's gentle, who has a better economy of motion, who has better exposure. And they rated these surgeons basically as like really good surgeons and not so good surgeons. And what they found is that there's a really clear correlation here between surgical skill and complication rate. It goes down as you get a better surgeon. So actually that's surgical skill, complication rate. There we go. This is a video I took in fellowship of my mentor and one of the residents he was operating with. It just sort of gives you a visual thing and this is one of the things. Putting this into sort of numbers is kind of hard because our eyes and brains can see it and I hope you'll be able to see that in this video here. So this would be the novice. It seems like everything they touch sort of bleeds and doesn't quite, you know, it's hesitant. You get the idea, okay. So that's, you know, that's that. And then here's the master surgeon, my fellowship mentor. You see, he's not even using it like a knife. He's using it like a retractor and then he's going to, you know, cause, you know, get some better tissue exposure with this, you know, cuts a little bit, finds the perfect tissue plane, pulls it aside, this thing can hold it. And you can see the flow of this operation just naturally to our eyes is better. And you wonder how you might be able to capture this. So this is important information, information about how good people are at operating, information about, you know, the communication in the team or how it works or other factors that may go into patient care. It's important. It's a national healthcare priority to find information to make healthcare better and actually getting this data in some way out of the OR, therefore, I would say is a national healthcare priority. So how do other industries collect data for quality improvement, air prevention? Molly nicely set this up. Of course, we have the aviation flight recorder, which has two components. There's the flight data recorder. So these are all the technical things about the plane as well as some sensor data, gyroscopes, wind, speed. And, you know, they keep that for a long time. That's actually just a lot of numbers. So it's sort of not that hard to save. And then the cockpit voice recorder, you could certainly record more of it, but, you know, after some negotiations, it was decided through by pilots that, you know, well, they'd be comfortable having the last half hour, you know, recorded. And it's a continuous loop and it eats its tail. And the idea is if there's a crash, okay, you have the last half hour before that. The essence of that, all their personal conversations, you know, go into the ether. But there's no opt out. It's always on. It's always being recorded. The two ways that this data is used, so pilots, if they have a near miss or airline companies, if they notice that there's a problem, can voluntarily submit data to analyze by the FAA. If there's a crash, the National Transportation Safety Board takes over and, you know, no one gets to negotiate that data is automatically taken out. There's actually another potential analogy. I think that those of us who've been following the news know that there's all sorts of fuss about police uses of force and efforts to try to curtail that and police are now wearing cameras. And they've been doing that for a few years now, but it's increasing. And actually, it's mostly for the police officer's protection. It can ensure that, you know, despite any complaints of an arrestee that the, you know, arrest was conducted appropriately. These are officer control, though. So they choose when to turn them on and turn them off sort of as it stands right now, although there's plenty of debate about that. And then there's the duration of retention. How long do you keep this? And interestingly, you know, the ACLU, who would normally say no more cameras, you know, and I'm sort of paraphrasing, actually say, well, maybe some cameras if we can prevent the use of excessive force. And so there's definitely some tension in this world about wanting to have privacy and security and protection. So we can do this in the operating room. This is David Song, one of my colleagues, and we did some experiments just using GoPros and several of the other competitors in the operating room to see how they work. And it's really fun and great, and they take awesome videos and it's easy to use. You can actually do things with these videos. So this was a GoPros sitting up on a tripod and the heat map that you see overlaid is actually all of the staff activity as they were setting up a room. This is really preliminary data, but the point is we can sort of track where are people, what are people doing around the operating room and begin to sort of look at how you can make it more efficient. Now one of the problems with this data, of course, is this is like HD video. You know, this is like tons and tons of gigabytes to get even a single case. And I mentioned the police officers, well, they're spending, you know, communities are spending millions and millions of dollars just for storage of these videos, you know. It's expensive now. It gets cheaper, keeps getting cheaper, but nevertheless, millions and millions of dollars, even for an individual, medium-sized community. So then we say, well, can we ratchet down that data, sort of get more like the cockpit or the flight data recorder? Just numbers. And so we can use things like the Microsoft Connect, you guys may be familiar with it. It's a video game component where basically if you dance in front of the screen, the character dances in front of the screen. And so we've been trying this out in the operating room. What you see is you go from HD video to a point cloud which is sort of this colored image which doesn't have a lot of detail but gives you some sense of people moving around and is also a little de-identified, which might be nice. And then you go down to a skeleton and you really, you know, not have much data at all that you need, but it's still XYZ coordinates so you can tell when people are moving their arms. I'm not sure what the usefulness of this is yet, but I think that there could be. You just have to do some correlational studies. So we're still working on this technology. You can also just look at audio. You can parse out audio, figure out when instruments are being used, when the door opens, when something drops on the floor, or when people are talking and even remove the content of the voices and just hear, and get some sense of people talking or communicating to each other. So there's lots of opportunities there. And then you can also do myographic recordings and I guess the only question I would ask is, which surgeon do you want operating on you? A or B? And so you get some idea of the data that we might be able to collect. So what can we do with this? You could create better surgical techniques. You could train people to be better surgeons. You could actually investigate when some complication happens or some terrible things happen. You could go back and really get some rich data, not just a recollection of what happened, but actually what happened. You could use that to create early warning systems. You know, Peter, you're operating a little too fast. You seem a little too stressed out. Let's slow it back. You know, everyone calm down or like let's stop the disruptions in this room. We're gonna shut off the phone. We're gonna stop people from coming in and out. We're gonna keep calm right now because this is a high risk moment. And then you could also use this to generate automated medical records. So there's some big questions and in true exploratory ethics format I will not answer any of these today. But I wanna talk a little bit about them. And also there's some other questions I should point out off to the side here. I don't even address, but are also quite your main to this. We don't have all day. So can the surgical team opt out? So this was a somewhat controversial framework to talk about how we can kind of balance the research QI dichotomy. This came out in the Hastings Center report. And one of their arguments, this group of authors was that physicians have the duty to provide the best care. That's sort of acceptable to everybody I think. They're the only ones in the position to improve care. Maybe controversial, maybe true. So therefore they're obligated to participate in programs that improve care. And so you could make that argument. I'm not necessarily gonna go into all that today, but you one could make that argument that physicians should participate this some way. Flight teams can't, oh, sorry. Flight teams can't opt out, so we know that already. And it sounds like police might not be able to opt out either. This is a very recent ruling in Albuquerque. The Justice Department said they're required to wear body cameras to record their encounters. It used to be sort of an optional thing for department up by department. And now we might see that all police officers are gonna be required to wear this. Maxineau, all physicians are gonna be required to wear this. That's entirely theoretically possible. So should patients participate in this? Well, there's an argument again from these authors that they benefit from the common good. If we made the healthcare system better, then everyone benefits. Therefore they should be willing to have this sort of minor data about their vitals or what's going on in the case available. Did we give patients the choice to opt out to not have that monitoring? Or is it their data anyways? Is it really their data? And so the two arguments on both sides of that are, well, if it's a patient's data, it's about their procedure, much like if you had a CT scan or the narrative operative report, that's data about what happened to them. So therefore it's their medical records, their data. And then if it's surgical teams data, it's because it's about their performance in the operating room. It's not really about the patient per se. It's about what everyone was doing in that room and it's about those human beings. So it probably gets a little bit more complex than that. If you have the raw data coming out of the operating room, you can imagine that there's some data about team activity, communication, performance, sort of the things I was showing earlier. This includes, and I separate this out intentionally, but procedural images and video, that's like the richest stuff to collect. You can tell a lot from it. And then there's clearly the patient data stuff. So vitals, procedural data, implants. And so we can see that you might split some of that rich data up in two camps. And how is this regulated? Well, starting in the center there, depending on what the video is, is it identifiable? What's the purpose for it? You could sort of put it in either camp. But team data may be protected under sort of the peer review process. And then patient data, well, that's HIPAA, that's healthcare record standards. Now we could say, well, let's just do this. Let's just take the team data, and we're gonna ship that over to some regulatory body to determine if actions are happening or how to make things better. That's a great way to handle it. Well, the problem is that patient data is actually incredibly germane to team performance. Whether they're operating on a morbidly obese patient or someone who's a Jehovah's Witness and hypo-coagulable might change dynamics and performance in a way that should appropriately sort of risk adjust whatever they're doing. So I'd like to propose today sort of four models of thinking about this, and I hope I, yeah, I got time, good. Four models of thinking about this. So the first is the aggregate model. So we could theoretically take this data in sort of de-identified format and just report it to somebody and they could call through all that data and they could say, okay, well, you know, teams that talk, have better communication or surgeons that are gentler and much like that bariatric study, they simply have better outcomes. And that's a great thing and I'd love to actually do that. I think that's wonderful. Now one of the problems is it doesn't let you go back then and intervene, it just sort of allows you to make broad statements about what you're doing. So something like flights that take off too fast, crash or something, but not going back and telling pilots that are maybe flying a little too fast that they need to change their behavior. Second is a peer review model. So you just simply keep it in the hospital or in the state level or whatever and you report all the data and you say, well, it's a peer review process so we'll allow all that data to go in there and there's nothing wrong with that except it's incredibly effortful and it's also subject to the bias of when you decide to report it. There's a live record model. So that could be everything you collect just becomes part of the patient's record and one could make an argument. Yeah, you like that. So one could make an argument that everything in the operating room is essentially part of the patient's record. But then you gotta think about those of you who don't operate maybe you're in clinic with a patient and the entire recording of all your clinic, everything you're thinking and writing and the conversations you have out in the hallway about them and all that goes into the record might be a tough way to practice but an interesting thought. And then there's this parsed record model which I'm gonna explain a little more and I think this may be an opportunity to move forward. So Alex Pentland is a data scientist at MIT, the IDEA Lab. He spun off all these amazing products and concepts and designs and he created this idea of a personal data store. So what we probably don't think about all the time but probably know happens is that anytime you turn on this multi-sensor apparatus that is our smartphone, it starts sending all sorts of data about us up to our service providers, up to our data companies and that's how they can tell things about us. So for instance, those of you who use the Google Maps function and actually look at the traffic, the traffic is all the other users of those phones, they're sending data continuously up in their service prior and they aggregate that and they say, oh, these phones are moving slower than these phones so therefore there's a traffic jam here. And so that's how that all works. Well, they can also use the rich data that they collect about what stores we're walking into, where we're geographically, what we're paying for, who we're calling, who we're talking to to learn a lot about our behavior and really for their own purposes. And I guess I don't have too much problem with that but maybe I should, maybe I should have a problem with that and people get very worried about the NSA, that's certainly a national topic right now but the reality is plenty of people are monitoring our behavior all the time. So Pentland's idea was to create a digital barrier on these devices that we carry. Essentially you could decide what information goes to who but more importantly in the part that I think really is relevant to what we're thinking about in the operating room is you can create a system so that instead of sending all of the raw rich data out it actually sends only the specific information you need for the question that you're asking of the service. So if you're asking, how do I get from here to here it would send just a brief snapshot of your present GPS location. If you wanted it to tell you, follow you along and you could send it more continuously but it's not gonna track what stores you're walking into or who you're calling or the pictures that you're taking at that time like it otherwise might. So could we create a system where the data coming out of the operating room is actually parsed in a way so that we could ask very specific questions about team behavior, about surgeon performance, about the patient's likely outcomes and that could be sent to a regulatory body or even placed into the patient's medical record in a way that would allow us to analyze this without trying to scoop absolutely everything especially when there's gonna be a lot of data that early on is completely meaningless to people. It's hard to tell what it means when someone barks at somebody or the significance of a sound of something dropping and that could easily be teased at through afterwards if there was an adverse event but if you don't really know if there's a true correlation then it's just gonna muddy the system. Additionally, this kind of parsed data would also require much less storage capacity and thinking about it from a technical standpoint that's an important reality check to this. You can't just keep videos of every single procedure we run out of space very quickly. So a few conclusions. I believe that collecting rich surgical data is a good thing in some way or another. I hope that maybe I've demonstrated that to you today. I think that we need to stay ahead or keep up with the advance of technology because the things that I describe are, many are just a year old and that means that next year there's gonna be something even more amazing. Protections do need to be in place to maintain an annuity but still allow quality improvement. If we just take it all in aggregate, we're not gonna be able to get the same kind of quality improvement that we could if we can look at sort of individual instances or particular adverse events. And then finally, we need to create a system that would encourage physicians and patients to be willing to sort of not opt out that it would be something that everyone would accept as, this is okay if we just run this in the background all the time, because that's the way we're gonna get the best data. So thanks very much, I appreciate it. Hi, thanks for your talk. I'm Steve Sherson from the University of Chicago. I think that I have the privilege of working in an environment right now which may be the only one in the hospital where we currently have 24-hour data collection. We have videos in the NICU trained on the patient and anyone taking care of the patient at all times. And I think this is more of a comment but I think there are some things we've learned or some things to think about with that sort of 24-hour no opt-out recording. And there was a case actually in the last year where a nurse was fired as a result of review of the video from the care that she had provided. So it's an up in the air issue right now, still ongoing but there's no audio, just video. But anytime any procedure is done, including when surgeries are done at the bedside they're recorded. So usually we use the data if something bad happens but it's happening all the time. So I think it's... That's wonderful. Yeah, very similar also to codes, they'll record the codes and our audio goes on and what they do with that data is another question but I'd love to talk afterwards about your experiences if that's okay with you, that'd be great, yeah. Giuliano. Testa from Baylor Dallas, Alex that was a great talk. I have a comment, one is the analogy that we draw all the time between pilots and flying and surgeons, I think it's the wrong one. It's always been wrong from the get go. The pilot that goes from Dallas to Chicago knows exactly the route, knows which altitude is gonna fly and has all the data in front of him and has to follow a path. I would say the analogy would be better with a war pilot as a target name but doesn't know how to get there or if he knows, he doesn't know what is gonna happen when he gets there. I think that's a better analogy for surgery because we do have a route but we don't know what we're gonna find when we open that abdomen or when you open your neck. I think that's a little bit. And the second thing is the pilot, the co-pilot is somebody who's already finished his training. It's not somebody that is trained to become a pilot. That's a big difference with the kind of intercourse or relationship at the cockpit that may occur but to more pertinent to your talk, how it may influence the training of the surgeons in terms of the action that we as the people who train these surgeons should act upon when we notice that these guys are now up to the task because reality is when you really put this to the extreme many people go to pilot school or whatever that's called and we never graduate because it's clear they cannot perform the task and in our job, unfortunately or fortunately for the people who know how to operate you need to know how to operate. You need to know how to use your hands. So how would you implement then knowing that five of the 10 that you are training would should now finish their training while in reality we finish almost everybody? All right, that's a really great question. That's a huge tension in surgery and I was actually talking with a new Milani, I don't know if he's here, but the other night at a reception about this very topic. I'm so glad you brought it up. And the thing is that is something that I know you personally experience, I personally experience when you fred over it when you have trainees and you're like, wow, kind of bad hands. I don't know what's safe or what's right as far as graduating this person. And what do I wonder is, well, is there a way to do this operation that would be easier for that particular person with those particular bad hands or maybe another question is looking at all those Whipple surgeons. They all do it differently. Maybe that performs best in each of those hands and if they had to switch and do it a different way even with practice or training, it wouldn't happen even though they're awesome surgeons. And so I think that you hit it at a concern that we all have, but I think we also ultimately only have our individual perceptions of that person working with us or it might be a personality clash. And also we don't have a way of sort of adjusting surgical procedures that might fit better in certain people's hands than others. We only have our own way of teaching it. And so I think that there's a lot of fruit for that investigation. And it's so great. I love the pilot point is dead on and thanks. Thank you. Thank you.