BIG T TVR

Welcome back to Behind the Knife. This is Patrick Georgoff, trauma surgeon at Duke University, and I'm joined by Dr. Jason Brill, trauma medical Director for the US Indo-Pacific Command. And today is another installation of the Big T Trauma Series. The Big T Trauma series offers clinically oriented material that focuses on how best to care for the traumatic, injured and critically ill patients.

What if we had the ability to revisit trauma resuscitations in real time, not just through memory or documentation, but with high quality video and audio? Trauma video review or TVR offers that possibility and it's being adopted by an increasing number of trauma centers across the country, and that's what we're gonna focus on today.

As part of this big T trauma series episode, we're exploring how TVR is being used to better understand clinical performance. Team dynamics and communication under pressure. We'll talk about how it's been shown to outperform EMRs and even in-person observation when it comes to capturing detailed

resuscitation data.

Yeah, thanks Patrick. But we'll also acknowledge the complexity here. So there are a number of issues to go through like consent, privacy, cost, . It certainly isn't a silver bullet and its success depends on thoughtful implementation and institutional buy-in. We'll review the data, discuss emerging cases, and hear how teams are using video to reflect, improve, and sometimes even challenge their own assumptions about trauma care.

So whether you're considering TVR at your institution or you're just curious about its impact, this episode is for you. That's right. . Today we're joined by doctors Ryan Duma and Michael Vela, and you may recall them being on a previous episode, a big teeth Throwdown, in which Ryan and Michael went head to head with Bilal Joseph on the use of intraosseous catheters and trauma.

So they are veterans of the program. Welcome back to both of you. Thanks. Thanks for having us. Thanks. Good to be back. As Dr. Michael Velas, an associate professor of surgery in the division of acute care Surgery and trauma at the University of Rochester

Medical Center in Rochester, New York, and he is the trauma medical director at the Kessler Level one trauma center.

He currently serves as chair of the New York State Committee on Trauma, and he has a clinical and research interest in trauma video review, particularly as it relates to trauma team dynamics and initial resuscitation. Dr. Ryan Dema is an associate professor of surgery at Baylor College of Medicine in Houston, Texas, where he serves as a section Chief of Acute Care Surgery.

Dr. Dema, has extensive research portfolio that focuses on video technology to capture and review trauma resuscitations. He has helped develop and run several trauma video review programs across the country and is the PI for a large multicenter collaborative study with over 40 centers that are looking at how we can best use TVR.

Dr. Dema also provides consultation services to surgical safety technologies, a company that develops trauma video review technology. Alright, Ryan, let's start with you. How about the utilization? Where are we at in terms of who is using trauma video review?

So really we were interested in looking

at this from the get go. I first got in interested in TVR probably about a decade ago. And one of the first things that we tried to do was get some research going.

But one of the first questions was who was actually using it? So at the time it's probably data that's over 10 years old now. It was published in 2020. We surveyed members of East and asked them basically, what is their familiarity with TVR? Have they used it at their center?

Do they planning on use on using it? Have they had a program that had to shut down? And basically the long and short of it is at that time. About 30% of level one, level two centers that responded used video review actively, interestingly, and importantly another 20% of responders said that they were planning on using video review in the future or starting a program.

I think that's what's really key, and that's what Mike and I have certainly seen a lot of o over the last 10 years is really a dramatic uptick in the utilization as cloud-based technology. Audio, visual data and storage gets easier and it simply becomes such a big part of our lives to have videographic data

in almost every aspect of our life.

Yeah, it's interesting statistics, but what about more of the subjective data? Mike, can you answer maybe some perceptions amongst programs and how it's been received? Yeah, there's at least. Three studies that I'm aware of, mostly survey-based data that look at perceptions among providers.

One was actually Ryan's group a couple years ago, and then we from our center here published one within the last year or two. And the studies have either looked at programs that have historically had a video review program, so it's surveying active users. Who are pretty well aware of the technology. And then the study that we did more recently actually was a before and after survey looking at perceptions from.

Ed staffs, so nurses, techs, et cetera, as well as physicians on the resident and faculty side of things before and after us starting our new program about five or six years ago now, and in both those, so whether you're an active user. Whether it's before and after

starting a program, the perception among providers is overwhelmingly positive.

Particularly when individuals have the opportunity to use and see the power of video review. It's pretty easy for people to get excited about it. And so as you would expect, some of the concerns that people had before we started the program were things like being video recorded at work.

Is it gonna be used for, punitive purposes, et cetera, et cetera. And when people get experience with the program, those concerns essentially go away. I mean, people realize the power of it. I always say there's an art.

To, to using video review. We're very careful when we do these that we're not using it for punitive purposes or formal job review or he said, she said, type situation. We're very careful, at least in our programs. Not to do that. And I think that's part of the reason that our programs have been very successful.

And so there is a little bit of an arc to it but to the short answer to your question is it's overwhelmingly positively perceived among users. And I can tell you that having helped to start our new program here again

five or six years ago there's only one instance where someone came to me with a very minor concern related to the video re so one time in six years.

Yeah. And we'll drill down into these specific issues in regards to HIPAA compliance, discoverability, legal aspects of all this, which is extremely important. But for now, Ryan, let's stick with the rationale for starting these programs. And really the number one reason that people want this is to get that granular resuscitation data that they're currently missing.

We all know that the current process misses out on a lot, and that is raw data in terms of timestamps and what actually was and was not performed. In addition to things like team dynamics and tempo and tone and trauma. Video review gives us an opportunity to do that. Ryan, you wanna expand on that ?

Yeah. For me, pat, this was really the impetus for performance improvement because I felt like it was impossible to recreate. Trauma, sation timelines from the

electronic, my career, because we teach, for example, A TLS in a vertical fashion, right? A, B, C, D, E. But in reality, it happens horizontally with multiple things happening together, and a good team leader should orchestrate that effort.

But the problem is. Is that universally, even with exceptional nursing documentation , these metrics aren't captured and the EMR only captures time metrics, maybe if you're lucky, blood administration and intubation and those kind of vital things. But it certainly doesn't capture any of the non-technical skills.

How many people were in the room? What was the volume in the room? What was the movement in and out of the room? Was there a clear identifiable team leader? And how do they command the room? And so we've done some studies looking at this, for example, we used er thoracotomy as a model, and we simply just benchmark the procedure.

So for example, like insertion of the retractor and decompression of the right chest. And we timed all those things and we, those time stamps to the EMR. Over 50% of the data

for all, sometimes even more. Is missing from the electronic medical record.

It's just not there. And then we took it a step further and we actually had trained, paid in-person researchers 24 7 whose only job was to mark time points for thoracotomies. And still there was a lot of missing data. Because simply you can't pause, you can't pass fast forward. You can't rewind right in a vi in a video review.

When you have a video, you can literally stop zoom in depending on the application you have and get very accurate timestamp points. So Ryan, what's some of the most surprising things that you've seen when it comes to sitting down and looking at a video In terms of the information pulled from it compared to a standard chart review.

Yeah. It's really the non-technical skills is a big piece. And the other thing, which I say often, which always, you know generates some conversation is the bottom line, the overarching thing. We've all done this for a long time. We're not as good as we think we are, no matter what. Things happen faster.

They, we miss things, right? And so that's

the key in all our data for chest tubes, thoracotomies, vascular access. We are not as good intubations, right? Number of first pass attempts. Like we, we think we, we pat ourselves on the back and we're really, we're pros but we're not because what we do is really hard.

'cause we make very difficult decisions in limited amounts of time. And so that's really what's been the most enlightening to me. That and the non-technical skills. Yeah, I will say when I'm doing a chart review, the first thing I am likely to do as the EMR gets pulled up in front of me is ask the people who are actually in the room and of course the subjective recollection of what happened and when very rarely is the same from person to person.

So those are great points about w why TVR. Let's move on to how, then just briefly can you both talk about some of the platforms that are out there? Like how does this happen? In other words I know there are probably some listeners that are thinking this is just like a webcam in the corner and it goes to a video loop.

And some of the early ones, I know we're like that but what is

the technology like now? What is it capable of? What are some limitations yeah, I mean, I think it's interesting 'cause I think Ryan and I at our different centers have used.

Different technology with success. So again, I don't work for these companies. I don't, there are many great brands out there, so I'm certainly not pushing one or the other. The one we currently use is a Avigilon, so it's basically we have two fish eye type cameras, one at the head of the bed.

One at the foot of the bed, they're on the ceiling. You have to be, you have to plan for these things when you're building new ERs and there's booms and wiring and things like that. But for most beige you can make this type of setup work. So one at the head, one at the foot of the bed, two fisheye cameras, and then we have a single microphone.

And so that's one example of one setup. Again, there are many different types, many different brands, fisheye type cameras or otherwise. Some institutions may use security cameras whose, purposes for security, but they may already be in place in or around the trauma room. So there's many different, combinations of setup.

But I would say the most common is one or two. Video cameras,

particularly the one at the head of the bed. We use that one 98% of the time. And then usually a microphone. I would say the majority of centers. I can only think of a few that are doing video alone. Most do audio and visual. That's the typical setup.

And then what it allows you to capture, again, you're getting an audio visual feed from the trauma room, and then some of the more advanced capabilities allow you to maybe pipe in a. Vital Sign Monitor, for example. So in one little part of the screen, when you're going back and watching a recording of these videos, not only can you look at the actual video, and not only can you hear the audio, but you can actually see what the monitor is showing in real time.

Some of them allow you to have access to recordings of a laryngoscope screen, an ultrasound machine, all those kind of things, just depending on what modules you have available and what technology you're using. And then it becomes even. Much more advanced. And I know this is an interest of Ryan looking at AI and movement detection and, this technology is advancing pretty rapidly.

Yeah. Ryan, could you talk about that a little bit? Some emerging technologies and how that might influence what the systems look like currently? Yeah, I think it's a really important point and that's really why we're seeing a dramatic uptick, I think too nationally. Is because really now everything is cloud-based from our phone backups to, all our lives, right?

Our emails are black backed up in the cloud. And so the same way with Videographic technology now, right? We have the capability for automatic automated 24 7 capture. And these files auto delete in a lot of these platforms. And those are settings that can be adjusted by the administrators. But they auto delete.

Some platforms are even de-identified. Where the voices and faces of the individuals on the video are de-identified using AI algorithms. And so that again, makes the review that much more robust if there's a level of anonymity. So I think the technology has dramatically changed and will only continue to get better, which is I think why it's important to understand and

harness it now.

How about cost? Do you have some raw numbers you can forward? Oh, yeah. So in terms of getting started and, yeah, I think, so an important consideration is cost. You can do this as kind of a Cadillac with multiple cameras, four cameras per room, a vital feed, for example, de-identification technology, and then it's gonna be more expensive, but you can get a video review system going for probably about $10,000

and then there's obviously some maintenance and some biome engineering that may need to happen every now and then. But critical what is often overlooked is the human cost in the sense that it takes time to review these videos. It's a lot of data and, it's very easy to, it's almost easier or certainly easier to install the hardware and get it all up and running.

But then the longevity end of the program, is gonna be dependent on the champions that are really pushing and the performance improvement at your institution? From a cost standpoint, I would say, again, it's, it is probably, I would say 10,000, per base startup cost.

And then, for most of these, some of them have subscriptions, but at a

minimum it's gonna have sort of a yearly maintenance cost. And it's probably. A thousand a year or so. Maintenance cost for a lot of these. The other thing I'll say for those interested in starting programs is there is absolutely grant funding available for these things.

We use some grant funding through the emergency department at our institution, so I encourage those who are interested to reach out and see if that kind of funding is available to you. Yeah, that's a great point about institutional funding. MCW just started a program. They used grant funding. I know that Duke has started a program.

Orlando Health started a program, Arizona, and a lot of them have used institutional funding. Perfect.

Let's focus on the legal side of these things when it comes to consent and privacy, HIPAA compliance, whether or not these audio visual feeds are discoverable in the court of law, et cetera. Yeah, so I, I would say that this is no matter what study you read, looking at video review, when the study has looked at what the concerns are or why

places are not interested in starting a program, it's overwhelmingly either the top one or two answer, usually the top answer or the medical legal implications.

And so when the survey study that we talked about upfront that we did. A couple years ago now it was only two to 3% of people surveyed in that study had ever even heard of any center having a medical-legal issue revolving around trauma video review. So the perceived or the concerns that individuals have don't seem to be born out in the actual medical legal world, but like any.

Performance improvement endeavor, whether it's trauma, video review, or morbidity, mortality conference, it's protected to the extent that it's protected. I can tell you that I've seen at least one. Medical legal case not involving our institution where the resident's m and m presentation was actually a part of the medical records.

So despite, and again, I'm not a lawyer, but when we think these things aren't discover discoverable, the reality is that it's a much more complex

situation. But like any performance improvement endeavor, these are protected by.

Those sort of generalized standards and when we've talked to our lawyers here and we actually published a paper within the last year looking at trauma video review through the legal lens. We actually had some law students from the University of Buffalo do a deep dive into the medical legal literature.

And as far as we could tell, there were only two cases involving TVR at all in the medical legal literature. So it seems to be a very safe practice. And when you talk, specifically to the lawyers that we've talked to is when you make the argument that video review is a patient safety.

Performance improvement tool, which it absolutely is. You, when you make that argument to a lawyer, at least to the ones we've talked to, they feel like that outweighs the medical legal risk of the videos being discovered. So it's not risk free.

But it's very safe.

There are very few cases involving video review. And , when you make the patient safety quality argument, it's a very good legal argument from the lawyer perspective, if that makes, yeah. So Ryan if we'd link it to the quality assurance type activities are we protected by the standard peer review statutes in your opinion?

And are You aware of any state specific laws that protect , this data from being discoverable? I'm not like Mike, I'll make the same disclaimer that I'm not a lawyer either, but certainly I think I mean, Mike's right, I mean, there everything is discoverable in the sense that, is there a scenario in which, you know, a a lawyer asked an institution to retain a video.

Knowing that video review is ongoing, I suppose. Yes. But I think what I think the take and the takeaway that I wanna emphasize with what Mike said from having worked with those lawyers is that what we're really looking to do is change culture and improve performance improvement. And that's what mitigates future lawsuits.

And so the ability of trauma video review to help teams perform at higher

levels. Ultimately in the eyes of lawyers that have looked at this mitigate potential future lawsuits. So the only things that are worth considering as far as consenting and that I know has come up in some states is that there are consents that require two party consent.

For audio visual recording, 38 states in the union are one party con consent states, which means that only one person has to consent to, record a conversation.

Yeah. And I would say the biggest sort of disclaimer, I think one of the more, most important things that we can say as it relates to the medical legal implications of video review is that before you start a program, one of the first things, if not the first thing is you gotta talk to your legal and your compliance departments and get all those things because there are differences , between states, there are differences with institutional regulations and things like that.

So it's critical that you iron out those details. With your local legal department and patient compliance before you start a video review program. I think that's, yeah. And those programs and policies

abound now. Mike and I share policies. Penn has policies, MCW all these centers that I work, that we all both work frequently with, have all developed very robust policies to guide the utilization of trauma video review.

Yeah, and so we used to get consent. The joint commission requirements historically had said that audio visual recording, even if for PI purposes, required consent. And so the way we used the, our institution is that the audio visual recording would obviously happen in real time. But we, before we would watch those videos in any capacity, we got consent from either the patient or next of kin to watch those videos.

The Joint Commission has softened those regulations a little bit to say that it's now sort of suggested, I don't remember the exact verbiage, but if they've softened those recommendations of those requirements, and so we now do not explicitly consent for video review.

But again there's a lot of variability. Some places don't consent at all, some consent specifically for it, and some consent as part of the general hospital admission. that's a great segue into how the

videos are reviewed.

Let's say you're going to lead a TVR review. You're gonna walk through some group of people. Can you describe especially from the non-technical skills. That you're evaluating? Who's in the room when you do it, how do you set it up? What are your wickets that you want to go through?

Is it the same depending on the environment? Mike, if we can start with you since you mentioned it how do you do A-A-T-V-R? Yeah, so there are many different uses for video review and I think we're gonna get through a lot of them. So the research piece, using it for research is one arm.

Using it for coaching. Coaching and education is an important arm. But what you're alluding to is using it for quality improvement, performance improvement, those sort of things. Again, there's not a right answer to this.

There's many different ways to do it. The way we do it our, at our institution and that I've seen done at other institutions is that once a month or once every six weeks. We get together as a group. It's a pretty wide invite in the sense that, the trauma surgery group, the emergency medicine nursing, really anybody involved in the care of the trauma patients is invited to this conference.

It's very

much like an m and m type conference. And we sit down and we choose videos. There, there can be many filters. You can watch videos based on identified performance improvement issues at your institution. For example,

one of the things we realized at our institution is that we weren't as fast at setting up the rapid infuser device as we thought we were. So we perceived this delay in getting blood. So from the time that the blood was ordered to the time that the blood was getting infused. Was about four to five minutes, which is a long period of time.

So we said, okay, as a trauma program, we're gonna video review every single patient that gets blood in the trauma bay. And so we started doing that. We started timing the metrics, time to ordering blood time till the blood was transfusing, the tubing. And we realized that there was a little bit of a delay there.

And so we did a lot of education, a number of different things, and we would watch these cases in that multidisciplinary conference with emergency medicine, nursing, trauma, et cetera. And through that process of watching it, talking about Nvidia review, doing directed education, we actually took our, median

time from four to five minutes down to one to two minutes, oftentimes even less than a minute.

So that's one example. There are many different filters you can use. So you can say, all right, we as a group, we're gonna watch, videos related to. Issue X or issue Y based on identified opportunities for improvement in our trauma program. You may watch a video because maybe a resident came up to you and said, Hey we did this resuscitation last week.

I think as a team there are opportunities for improvement. And then we say, all right, let's, watch this. Or it may be a, high acuity, low frequency procedure like a CRI or thoracotomy that we say, you know what? We as a group haven't seen one of these in a while. Let's watch this as a group.

So there are may different reasons or types of videos you may wanna watch. But the, common theme is getting a multidisciplinary group together, watching the video from start to finish, having the ability to stop, rewind, fast forward, pause for discussion, introduce literature, all that kind of stuff.

And then again, there's an R to it. You're not singling out individuals you're, talking about things that went well, but maybe opportunities from

a group standpoint, not an individual standpoint. I think that's really key to emphasize. But I've been at institutions where they invite EMS crews helicopter crews, other specialists.

To keep it focused on opportunities. For the group, not as individuals, and not calling out individuals by rule or by name

yeah. Thanks, Mike. That really good style points there. Ryan, how do you do it? Do you have any differences? Do you have a case example that you can give us as well? Yeah I think Mike's points were all excellent. I think the non-technical is where it really shines. I mean, because you watch a video and sometimes you only watch a minute and 30 seconds and it's actually what you watch is before the patient even arrived and you say, Hey, watch to this team guys.

Watch them sit around in a circle, introduce themselves to come up with a plan. 'cause they know the patient's hypotensive, assign roles. And once a team of MDs, residents. Nurses watches that it's like it clicks and all of a sudden, next time I've had residents that we've watched a video in the following week, they'd be like, yeah, we did the pre-arrival time timeout

exactly like we saw the prior week.

And I think the perfect examples there too are pre-arrival timeouts, EMS timeouts. We've published on EMS timeouts and the amount of time a physician interrupts our EMS colleagues is remarkable. And we actually wanna step further, and we showed that a well performed EMS timeout is a surrogate for a high functioning team throughout the entire resuscitation.

So basically if we can teach tight timeouts, then that translates throughout the resuscitation. So I think the non-technical skills, so pre-arrival, timeout, EMS, timeout leadership is a big one. We just had the opportunity to publish our work in Journal of Trauma, looking at the styles of leaders.

So we looked at over a hundred videos and we identified the leader by positioning and by their role in the room. And I'll tell you, most of them are passive. And that to me is an opportunity. 'cause I think if you have a hypotensive trauma patient, the opportunity there is to be transformational or transactional in leadership to be very clear and

direct that team.

Yeah, it's remarkable.

If you have the opportunity to undergo a video review process or see yourself for the first time it's absolutely fascinating, right? So everyone has room for improvement. All systems have room for improvement, and the trauma video review system gives you the opportunity to look at the whole. And from that comes.

Just countless different ways to look at a single scenario dig down to the smaller components of it and work towards in improving those individual issues. There's no limit really to each case. You go through okay, we're gonna pull this case just sticking with the ED thoracotomy theme.

From that, then you look at the handoff and then the blood transfusion then did look at iOS and central lines and time to the OR and all these really. Important metrics that are part of the complexity and the nuance of every trauma resuscitation. And there's a lot of value to that.

How is TVR currently being used to either evaluate technical skills and or teach technical skills in the

Bay specifically? Yeah that's a really one where it also shines, obviously. I think we can. Really see what Mike already alluded to is the the low frequency, high impact procedures.

That's what's always made me interested into technology too, there's some chiefs who graduate maybe have not in done a thoracotomy, right? Depending on what center you're in, and being able to watch one and watch one where the room is quiet and there's a clear tempo and things are progressing like a well tuned F1 racing team is really helpful. And like I said earlier our research has been shining a light on what we do. 'cause prior to video review, we didn't really know. The takeaway is that there's a ton of room for opportunity.

The time it takes to get a chest tube in and a hypotensive patient from time of a patient arrives to hit to pleural decompression and the chest tube is in, is almost 10 minutes. And we've shown them that's a really long time. And so now we have target. That we can train to and teach to.

Vascular access is probably where we've had the most

impact. The success rate of a central venous catheter and a hypotensive trauma patient is about 60%. And that we have to think about ways to increase proficiency for our learners and for our teams. . Let's take this one step into the future. So if this is what we're doing with TVR currently what's next? What data is being gathered in terms of next steps what trials or studies are ongoing? What are we gonna do in the next few years with TDR?

For me, and Mike, really we've had the opportunity to work with centers across the nation now, and that's where it really becomes scalable. And one of our first priorities was how are we gonna collect data in a uniform fashion? So we know that centers are using it, we know that centers are increasingly using it.

So now we have to be able to come together and define really what a resuscitation looks like. The optimum times to, to next phase of care, the optimum time to IV placement, optimum time to completion of the primary survey. And so we did

exactly that. So after we published our MCT, we had a lot of other centers come to us.

Our original MCT had 19 trauma centers, and now we have over 40. In this collaborative that I'll all use video review. And so we spent a long time really developing a universal data collection tool.

And that way when we collate data, it's clean. The definitions are identical. 'cause that's really hard, right? You might define the completion of an IV access differently . And so we really drill down onto getting accurate definitions for all these metrics. And so I think now we have an opportunity to leverage this network.

And we meet pretty regularly to discuss projects interests. And Mike mentioned coaching. That's certainly an evolving area where we can really use these videos for both group and individualized coaching. From a research standpoint, answering important resource questions, I mean, besides the trauma team dynamics.

That we already talked about. I mean, what really does TVR give you? That a medical record can't and two of the really important things

are one time to decision making. How long it's taking you to make a decision, which I think is really important not only for QIPI, but for research.

It, and also it immediate changes in patient condition, right? So if you're looking. Just throwing out a research question, does in, is innovation of hypotensive trauma, patients associated with cardiac arrest, whatever these, people looked at this in prospective, observational fashions or whatever, but you can see vital sign changes immediately before and after procedures, right?

Or re OA, what are the vital sign changes immediately after. Putting a rebo a what is the time to decision making? What are the, nuances and decision making that led someone to do, this procedure at this time? And those are things I think that , the TVR is gonna help us answer those important research questions.

It's remarkable isn't it, that, we always talk about moving fast in the trauma bay. It's all about speed. We're trauma surgeons, we got our victories from. Stopping hemorrhage. That's how we can save lives. And we want to be able to move quickly. What is quick then?

What is quick when it comes to IV access, when it comes to blood, when it comes to timed

operating? I think we all can probably answer something that's roughly in the same window or certainly an idealized number, but that's amazing that it's not standardized. And so I'm excited about this multicenter trial that has created a universal data collection tool.

And we'll shed some light on the seconds and minutes that matter in the trauma bay. And that'll be extremely informative for setting standards and looking at individual centers and for individuals moving forward to say this is what you're benchmarked against.

Yeah. Now one of the downsides to trauma video review is getting this data out, you still have a whole video feed and someone for the most part currently still has to sit and watch it annotate, pull the data out. That's extremely time consuming. And one of the questions I have, how good is AI currently, or AI systems, I should say in general at timestamping or pulling out key moments in a trauma resuscitation that'll decrease that amount of time that we are burning, reviewing videos.

Yeah, you really nailed the

Achilles Heel of video review, and that's exactly it. It's labor intensive and it takes time. And naturally As we've talked a lot about technology we certainly can't forget the mention of ai. And so the short answer to your question is that it's not very good.

Unfortunately, AI is always touted. But we've made some progress. We, what we've done we've just submitted some work where we have trained, so computer vision is the, one of the overarching terms that's used. Basically it's computer models that that analyze hundreds of videos and identify patterns and identify when things are occurring.

And using computer vision with some collaborators in Stanford. We successfully trained a computer model by uploading over I think about 200 videos into the model to identify patient arrival, EMS, timeout, acute resuscitation, and departure. Within 97% predictive value or predictability, it can accurately predict those phases.

But that's it. I mean, so that's like bare bones. So yes, though, in

a not so distant future, we wanna develop models that can detect x-rays happening. Fast is happening, chest tubes are being placed, and we did that to some degree, but the performance degrades pretty quickly. Here's why, is because.

A patient arrival is something that a computer can see and interpret very quickly, right? But a chest tube, there's more nuances, ? The limitation is the data source, and that's the videos. And for our collaboratives, that's one of our priorities is trying to identify a central repository of videos.

But of course that really starts getting into HIPAA and how we use these videos. But really that's the future is being able to develop computer models to help us identify these time points across the spectrum of care trauma, and I think it should be said, certainly computer vision is at its worst today.

It'll only get better. Exactly. But what you mentioned though, in terms of the data, that's the footage itself. We've all been in a busy trauma bay where there are countless individuals hovering around the patient doing the work that needs to be done.

And oftentimes you can't see. What's

happening? Sometimes you can infer pairing that with audio is helpful, but that's a very significant limitation in terms of, okay, let's say you had , the perfect computer vision program trained up. It still has to see what's going on. That's by definition, it has to see what's going on.

And so , that's the limitation. it's a limitation for all of open surgery too, just trying to get good quality video to, study is challenging and certainly we've thought about this a lot when it comes to video atlases and trying to get out good content there.

It's very hard to see what's happening in open surgery. So with that, let's try to steal, man this argument, especially for an individual or a center who are interested in starting video review. Is it worth the cost? Is it worth the time and the energy to deploy and really appropriately. Utilize

i, I think the culture change that I have seen in institutions that use video review is what's critical because I think when you

start watching your teams perform just the way athletes do, right? Nobody's playing in the Super Bowl without watching hours of tape. And they're watching the offense.

They're watching the defense, and we are in the business of saving lives. And the way you do that and the way you get better is by watching yourself. You mentioned yourself, Patrick, how enlightening it is to watch yourself perform. And every one of us who've, who's done that and seen us on seen ourselves on the TV VR camera has that same experience and immediately.

You have a lot of insight into the way you, as a provider performs, but also your teams. So I think that's, to me, where it's really hard to argue that we shouldn't be doing this in my opinion, because we have such an opportunity in modern day society to leverage videographic technology, which is ubiquitous.

The moment you walk, you park your car in any hospital, you're on video camera. It is just, the truth is in the lobby, up into the lounge, you're everywhere is you're on camera. And so we have the

data, we have increasingly usable technology. We have the resource that shows that there's a lot of benefit from it.

So I think for me, that's the way I tie it all together, that high performing teams have to review their performance to become the best possible teams. Yeah. I think that's well said, Ryan. I mean, I remember a year ago I was watching my then 8-year-old at a tennis practice watching videotape of him serving, right?

This is done in a lot of different fields, whether it's race, cars, sports, et cetera. As Ryan mentioned I think it's critically important. But I think for me, one of the most profound thing is the cultural change, as Ryan mentioned.

I completely agree with that. When you have a resident or a fellow come to you and say, Hey I, can you watch this video with me? I think I have some opportunities here to do this better. That's very powerful, right? And so one of the things we do at our institution I sort of call it coaching.

It's a very informal process, but we have the third years or the chiefs of our trauma service during the day. Once or twice a rotation. I meet with the

30 years I have 'em pick a resuscitation they wanna watch with me and we watch it and we talk about things that went well. We talk about opportunities and that we've been doing that for three or so years now and that's become part of our culture here.

This idea of we wanna get better. And I can tell you I have some data, hopefully we will publish this at some point, but I can tell you that the things we talk about one day, the next day on service, you'll see those things borne out in the trauma resuscitation. It's a very powerful thing and to me, there's no better way to improve that.

All right. Thanks guys. I have one final question, which we don't have time to answer today, and maybe we're not ready to answer this question, but as Patrick sometimes allows me to do, I'll toss a grenade in a filled room and then walk away.

So should this, if it's that important and it deals with all of the performance improvement metrics that you've mentioned. Should TVR be part of trauma verification? Yeah. Should it be part of the standards? Should it be a requirement? And now that I've done that, walk away slowly.

Walk well real quickly.

Alright, Ryan and Mike, go ahead and sign us off.

Dominate. Oh my God. Thought I was terrible. We to a bloopers reel.