Latest / Behind The Knife: The Surgery Podcast / Meet your AI Examiner: Introducing Behind the Knife Oral Board Simulator

Transcript

Meet your AI Examiner: Introducing Behind the Knife Oral Board Simulator · Behind The Knife: The Surgery Podcast

• podcast_whole === [00:00:00] Hey, behind the knife listeners, I'm Iman Aya, general Surgery PGY four at Duke Hospital. Today we're introducing an exciting new behind the Knife resource, an AI powered oral board simulator, but beyond this new resource, what's truly remarkable is the story about how it was built by a general surgeon with no experience in computer science, driven by his dedication to his trainees. At the end of this podcast, we'll also give a sneak peek of a complete example case done by me, a clinical starting my PGY three soon, and I'll also give samples of the feedback it provides after completion. Now, before we get any further, let me introduce our hosts and guests. First, we have Dr. George off a trauma surgeon attending of mine at Duke. He's one of the directors behind the knife. And his passion for the education needs. No further introduction. I'm in. This is exciting. This is so ridiculously exciting, uh, for us at Behind The Knife. I can't wait to talk about this ai, uh, simulator. It's, it's pretty impressive. And next we have Dr. Matthew Swenson. He's a general surgeon and associate PD of general surgery residency at [00:01:00] Valley Health System in Las Vegas. He's an Air Force veteran who completed his medical degree from Uniformed Services University and General Surgery Residency at the University of Nevada in Las Vegas. After training, he was stationed in Anchorage, Alaska for seven years. Where he took part in multiple domestic and overseas military exercises. At the completion of his military service, he was recruited to return to Las Vegas as the associate program director for the Valley Health System General Surgery Residency Program. His primary passion is and was ensuring that surgical residents are fully prepared for independent practice upon graduation, and that's what led him to create innovative AI powered study tools for oral board exam prep. He's a creator of the tool we'll talk about today. And on that note, Dr. Swensen, very excited to have you, and can you give a high level overview of what you've created and why? Absolutely. And I am so excited to be here today with you all as well and to, to share my work with your audience. Um, as a surgical educator, I place a big [00:02:00] emphasis on preparing for the oral boards. , The oral board exam is unlike any test that residents have taken previously. There's no multiple choice options. There's no one right answer. Uh, I personally owe my interest and a shout out, I guess to my, uh, to my mentor Dr. Gary Shen. Um, when I was a third year resident, he came to me and said, you're not very good at this, right? We need to practice. And so, almost weekly he took me into practice oral boards. , By the time I took the actual test, the process was very natural to me. So once I began teaching my own residents, I wanted to pay forward what Dr. She had done for me. As part of that process, I began exploring better ways for these residents to prepare for the test. , I was incredibly impressed with, with AI or large language models, um, early in their release. And so I began to wonder if AI systems could be used to actually simulate the. Uh, the oral board experience. So I began working to find out in over hundreds of iterations and different, uh, trials, I was able to create some, [00:03:00] some simulator type systems, uh, to create an oral board experience and grade user performance with, uh, what I feel is pretty extremely high fidelity. Yeah. And, and it's, it's actually an amazing tool and I can't wait to talk more. And, uh, as a testament to that, we have one of his former trainees as well here today and one of the early users. He's now a forge surgeon in Portland, Oregon. And it's Dr. Tala Rafiki. Now one of the first things that I noticed when we went through everything and when we first started to learn about the tool was that it's obvious how many iterations it must have taken to create what you've done. Uh, one, one thing I kind of want to know from you is what part of developing this stands out as the most significant or challenging period or hurdle. Maybe you had some particular problem or behavior that you couldn't, that you couldn't get through, and I'm just curious, how did you tackle that and, um, what did the process look like? Yeah, my concern early on of course was, is this gonna be accurate? Right? I didn't want people taking tests and being told wrong information, or, you know, we've all heard about AI's [00:04:00] tendency to hallucinate or make up information. Um, and so my work early on really focused on making sure that the simulator provided a high quality and accurate simulation experience, uh, without going off script, right. And telling you. What was true, what was not true, right? Following surgical principles. In other words, uh, because of that, it took thousands of tests on the system. But, uh, the results at this point are pretty impressive. But, uh, definitely a challenge to get there. Yeah, I think, I think that that is, um, is definitely what everybody thinks about first. They think, how can AI possibly Yeah. Can I trust it? Yeah, exactly. Like, can I trust it? And, um, what if it says something wrong? Mm-hmm. Uh, and so definitely you need to do thousands of tests to, to see if it does or does not. And, uh, Dr. Rafiki, you, you used it during your training. You're one of the early adopters. So what was your experience like? Um, and, and did you see any of these problems? So the, the first thing is something that we've all heard in training, and that's to have an important foundation. , So studying for the board should be a priority throughout training. Um, [00:05:00] in general surgery, residency, we have the app site every year, so that kind of gives us a benchmark every year. , And for that there's some various, uh, resources, excellent resources including this podcast. Now, Matt shared this AI simulator with me early in its development right around the time I was really getting into oral board studying. , I tend to use few resources, so I added this one onto the one or two that I was using before. This was probably one of the most impactful resources for my board study. I saw it evolve. Um, initially I was being tested on pretty straightforward cases over a tech space client. That was great because it allowed me to work on making my responses flow. I could also knock out a few cases throughout my very busy, uh, fellowship, workday. Then over time it evolved and early on I did, um, I did see like some. Some quirks to the program, but I, I, I had had some experience like navigating the AI scene before, so I was able to work my way around it and, and still make it work for me. Then, [00:06:00] uh, then it continued to evolve and the cases got more complex and they became multilayered. Uh, they started to cover various different subspecialties. Um, I was, I felt like I was being tested on different aspects of general surgery, and then later on that was pretty parallel to my experience on the actual board exam. , Then when it got the ability to use voice, , that was great. Um, and I was able to use voice to text and simulate talking to an examiner. , And with my time constraints and the ease of access, I found myself using this more and more. I could knock out some cases in between or cases, um, and then I was able to share with my colleagues later on as well, and those who are familiar with it have had similar experiences. , The biggest impact this had was helping me keep my responses appropriately detailed, but at the same time, concise. And I feel like that helped me on my board exams. Yeah. Thank you. And you know, I'll, I'll say having done many of these tests in the past week or so, , it is becoming more and [00:07:00] more clear to me how useful this is. Uh, I, I think that just getting the practice of being asked a question that you're not anticipating and having to come up on the spot is something that is very, very valuable. And, um, I have a question for you, Dr. Gird, off, you know, when, when you first heard about this and you, you help us a lot at Duke with oral boards. What did you see as the role and, and function of this type of a, , this type of a program? Well, there's obviously a huge need and a huge opportunity. The, you know what, we've already produced it. Behind The knife is pretty rich in terms of the audio scenarios. Now, video, we have written. Uh, uh, descriptions of operative, uh, different surgeries, but that true interactive component, , is, is missing. And you know, the most important part of studying for the boards is practice. Practicing exactly what Dr. Rafiki just said. You know, being, you know, staying on your toes. Working through a complex, , question that may have more than one correct answer and demonstrating that you're safe and thoughtful and doing so, and that's not easy, [00:08:00] and some people are better at it than others. And so the practice component is massive. So what better way to practice than having literally 24 7 access? To a simulator that has an unlimited number of unique scenarios. And, uh, you know, again, as Dr. Aff, he kind of mentioned this and, and what Dr. Swenson's been working on has matured over time. And in 2025 at this time, AI has matured as well to the point where we have a pretty awesome system. And again, you'll hear that in the, the scenario that you, Dr. Ali do here towards the end. Yeah. And the, the convenience is absolutely incredible. You know, just, just last night before I went to sleep, I just put my phone on my chest and just talked, talked a little bit to, uh, the examiner. It was fantastic. And, and I think that that's a privilege that we have in 2025 that. I, I almost wish people could have had 10, 20 years ago. Um, now one of the most impressive components of this simulator is the richness of its feedback. , Dr. S Swensen, can you tell us more about this? Because I think once we show people and once we release this [00:09:00] tool, that's gonna be something that impresses everyone. Um, sure. Yeah. So the thing that everyone focuses on, which we have been focusing on, is the actual test experience, right? And that is, without doubt, probably the most critical thing, you know, for the simulator. I think one of the biggest values of it though, is once you're done, , finding out how you did. So, I mean, AI is good at many things, but the thing that it excels at is analyzing content. , Most of us, so like what I'm giving an oral board exam, for instance, with, with Dr. Rafiki, , at the end I would pretty much, you know, give him a thumbs up. Good job. That was excellent. Maybe say this little thing different, right? Hey, the margin should be two centimeters, not one centimeter. Little superficial things like that. Right. , The American Board of Surgery lists their criteria for passing the oral board exams. , They have seven essential attributes essentially. So I took that and I created an AI system that thoroughly analyzes your exam performance after the test. It's gonna give you a numerical score, , and kind of an overall score, , as well as commentary on what you did, what needs to be improved, and then the [00:10:00] scores are broken down by those attributes. Also, how did you do on biliary? How did you do on trauma? You know, if there's other multiple layers to the test. , Most residents don't realize this, but if you look at, uh, the documents, uh, for what's testable, they're actually 27 different testable subjects on the general surgery, oral board exam. And each of those have different topics. So when you add that all together, it comes out to 362 different subject and topic combinations that could be on the test, uh, when you take it. Uh, now most of us as humans, certainly, um, you know, residents of course, uh, we tend to focus our practice of what we know. That's why people hire coaches, right? I'm good at one thing. A coach will help me focus on something else. And so the benefit of this system, the benefit of tracking, is you can know what you're good at, but also. What you're not good at, right? You keep doing biliary exams, you need to go do a head and neck exam, right? You need to focus more on procedural steps for a Whipple or whatever else it is. The system really helps you find out where your weaknesses are and focus on those [00:11:00] so that you can make sure that you're completely prepared for any scenario that comes up on test day. Yeah, the, the value that can't be understated. I mean, doc, Dr. Fugue, any comments on just the randomness and the ability to do those scenarios that maybe just challenged you a bit and that you didn't want to necessarily do? Oh yeah, there, there was, the randomness was one of the, one of the key assets of the program because like you would start off a case thinking, it's like, for example, a pretty straightforward vascular case, and then it evolved into some crazy critical care case. So, so it was really nice in that sense because, um, uh, it, it was well equipped to throw you some curve balls, uh, so that you could prepare for the same thing, even though, even though on the. Board exams, like it is relatively straightforward, but they test the breadth of general surgery. And I feel like this program really helped me, , practice for that. Yeah. I'm, I'm getting the same sense. And, uh, Dr. Dr. Juro, just broadly, how does this project fit into the behind the knife mission itself? Uh, [00:12:00] I'm so glad you asked. So at Behind The Knife, our mission is in fact to create innovative surgical education content that's accessible to all. And so clearly AI offers a really massive opportunity to change the way in which we approach medical education. And that's why this simulator and what we're talking about today is so exciting. And we've spoken in the past about how the democratization of technology and this rise of the creator economy has really turned content creation on its head. And I believe that Dr. Swenson's story and the creation of the simulator is the perfect example of this. So here's someone who loves surgery, loves his trainees. And recognizes how impactful large language models can be. And another big part of our mission at BTK is to find these people, to find the Dr. Swenson's of the world, the master educators surgeons who have passion for education and creation. And to be able to partner with them and get them onto the behind the knife platform, which has been made, you know, possible by the contributions of so many different surgeons and educators over the years. Yeah. And we [00:13:00] see, we see in clinical practice as well as new techniques evolve and new devices, uh, evolve to help our patients. Um, and it's, it's. Pretty simple. Why, why would that not apply to education? For example, the robot's a tool to broaden our surgical horizon. , And one application for AI proven in my mind with my personal anecdotal experience is, , that AI is a tool to optimize education. Many of us can't learn as well with textbooks and flashcards. And for my learning style, this was a perfect tool that allowed me to get to the point with my studying. If it was available earlier in my training, I probably would've used it throughout. In conjunction with the BTK podcast and other resources, , I believe that as cutting edge technology becomes accessible, it behooves us to use it to expand our knowledge and remain on the cutting edge of care for our patients. And part of that, of course, is evolving our own education. I'm very excited to see how this tool will change the landscape of surgical education in particular, [00:14:00] especially as an early career surgeon with my own trainees who deserve the best. Yeah, and this is just the beginning. I mean, this is just one example of what large language models specifically can do, let alone other AI technologies. And I also wanna recommend or recognize, excuse me, Dr. Ali's contributions to this project. And the integration of AI at BDK in general. So Iman has been working on this for the better part of a year, and I know that I've learned an absolute ton from him as I kind of dabble in this world of ai. And this is a space where we want to play at BTK, you know, on the edge, uh, on a cutting edge, if I might say, uh, of, uh, surgical education. And to that end, uh, Matt first and then maybe Ayman, what, what have you guys learned when it comes to working on this simulation, uh, simulator and any other kind of pearls of wisdom when it comes to tackling? Uh, pro, uh, uh, projects at the margin, especially using new AI technologies. Uh, sure. I'll go first. Um, well, first off, I, I do wanna compliment, uh, tah, uh, kind of set him up as an example for all residents. , He kept close contact with his [00:15:00] mentor, right? Find a mentor and keep close contact with them. He didn't, you know, Sayara um, at graduation we never talked again. He kept saying, Hey, will you give me oral board exams? Right? Can we do a Zoom link together? Uh, I mean, you can only imagine how excited I was when I talked to him the day after the actual test. And he said, honestly, it was easy, right? He said, I don't know if that means it was easy or not. 'cause he, we didn't know if he had passed or not, but he said it felt easy. It felt natural, right? It was. Just, it was another simulation. Um, and so, uh, I definitely encourage all residents, anyone you know, who's preparing to, I mean, this is a simulator, it's not the real thing, right? Find a mentor, right? Nothing can replace sitting across the table from an actual board certified surgeon. So find a mentor. Um, the second thing, as I remember how intimidated I personally felt when I first envisioned this project. I was sure there was some large company out there that had an army of computer engineers that had already done this. And so I started looking for it. Uh, when it was [00:16:00] clear it wasn't the case that I realized I could either wait around for it to happen 'cause it was clear the technology was there. Um, or I could use that technology to make it happen. , My experience and knowledge of the surgeon actually became a huge advantage. Um, 'cause I was able to take existing technology that was built for other purposes, not for surgical simulation. And I was able to repurpose it for what I knew that we needed as surgeons. And really only a surgeon or only an educator can know that this is what we need. So my advice to other surgeons and trainees is essentially that if you see an unfulfilled need, if it's in your hospital, right, if it's in your clinic, right, if it's in the classroom, whatever else, don't be intimidated by the fact that no one else has done it, or because you think someone else could do it better. Go ahead and, you know, give it a shot. Try to try to make it better. Yeah, and from from, from my perspective, I think probably the biggest lesson that I have learned is I'm constantly humbled by these, um, by, by other folk that have adapted LLMs and applied it to what they're doing. I think [00:17:00] that I have a tendency maybe to underestimate their. Their utility. But then I get very, I very quickly learn that, um, that what, what really limits me is my imagination and how it can be applied sometimes. And I think Dr. Swenson's a fantastic example of that. He knew hi, his surgical background and his background with education gave him a vision on how to apply this project. And with the power of LLMs, he was able to make that happen. , So that's one of the greatest lessons that I've learned. And, and I think that. The other lesson is that the most valuable thing that we have is our background as, as, as surgeons and as, um, in our medical education. And I truly think that a large company may not be able to do this with the success that Dr. Swensen had because they don't understand what we need from it. And the prompt engineering is complex. It's, it's probably not as straightforward as people think and the way to build these tools, although it's easy to do the, the technicalities of how to [00:18:00] construct. The actual thing is very, very difficult. , And so what I will say is that the most valuable thing about that is your background as an educator, and that's how you shape these tools to really be impactful. And so you actually may be the best person to do this, , and not some large tech company. And that's, that's what I've learned the most from this entire experience. I couldn't agree more. I with what you just said. You know, we have a very specific expertise as surgeons and furthermore, if you're a surgeon who loves education and, and may be. You know, a passionate and good at it, man, to be able to marry these two things and, and apply this emerging technology, what an incredible way to approach it. So we want as many folks, uh, as possible to give the simulator a try. It's really, I guess, as we mentioned, pretty cool. Uh, it's, but to be clear, it's not perfect, right? It is far from perfect, but, and this is the launch of our beta version, specifically version 1.0. So it's only gonna get better from here. This is the worst it's ever going to be, and it's pretty darn good. And as you [00:19:00] practice, uh, we, by we, I mean mostly Dr. Ali and Dr. Swenson, the, the, the geniuses here will be able to fine tune the tool and continually make it sharper and sharper. Uh, it's also free to use, so anyone who signs up, we'll get enough tokens to get them through. Two scenarios with feedback. Again, the feedback is pretty impressive in its detail, and if you want to keep going, the resource is affordable. So 20 bucks will get you 10 to 20 more scenarios, again, available 24 7 completely unique scenarios each time. And. I think this is a really exciting addition to our general surgery oral board prep. So right now we have 127 audio scenarios each with expert commentary. So it's a quote unquote perfect five to seven minute scenario. That's part A. Part B is that same scenario and dispersed with important commentary about the actual clinical topic that we're talking about, but also about how you might wanna approach, uh, that specific topic or that answer on the exam itself. We've also just recently added 10 interactive video [00:20:00] scenarios. And 97 operative descriptions. And so the addition of the board, uh, simulator really rounds out this entire, uh, prep package. And we're also very proud of the fact that our prep material not only contains what we really think is, you know, the best content's, the broadest content out there, um, but it's the cheapest compared to the other competitors. That's really important to us. We take that very seriously and we intend to keep this resource, uh, very affordable going forward. Absolutely. Actually one of my goals early on was something that an intern can access, um, because it's 20 bucks, right? It's very easy. Something that you can use throughout your education. You know, you don't have to shell out hundreds or thousands of dollars to access it. , So I mean, that's why I was so excited, you know, to start working with Behind the Knife on this. Yeah. Um, also before we jump into the example with Ali, a couple other things to to point out. Um, so this, there's, uh, two different types of simulators that we have. Uh, one is called the Realtime Examiner. Um, it's, it's the real thing, right? It's [00:21:00] a full conversation, low latency. You speak to it, it speaks back to you. You have an oral board, convers. , I've built in six different examiners. They all have their own attitudes, proclivities voices, , which will represent essentially what you'll get on exam day. , We've also created one that's more text-based, so it's a traditional, you know, chat type of thing. You type in an answer, it responds, goes back and forth. Another cool feature we're trying out and, and to be clear to everyone, we're not sure if this is gonna work. We haven't really been able to test it out much yet, but we are, are adding some, , support for multiple languages. So right now I've added support for Spanish, Arabic, German, and we might add a few others. , We need, we would love your feedback on this. , The beta release, as Patrick mentioned, , now includes, uh, just the practice tests, um, and getting them graded. , Once we've ironed out those kinks, we're gonna add some initial features, things like coaching, um, as well as other specialties. Right now it's just general surgery, but we plan to add additional specialties as well. And I think it's, uh, worth mentioning that, you know, while we often think, and [00:22:00] I think Dr. Rafiki, you mentioned this earlier, we often think of oral board preparation only in advance of the test itself and really kind of kicks into gear in chief years. But I, I, I really think that's a mistake. Uh, it's a lost opportunity for junior and mid-level residents to study. And to grow and to practice in the best way possible. The boards, you know, they're not just the finishing touch on your training, they're actually the, the most legitimate exam you're you're ever gonna take in your entire career. There's no surprises, there's no gotcha moments. It's just clinically relevant scenarios and the complications related to those scenarios that force you to really think. And to show that you're a smart and a safe and a good surgeon. And this is actually what happens literally every single day, uh, when you're in the hospital taking care of patients. And so I think, uh, to Dr. Swenson's point about this being affordable and easy to access, why wait to prepare for the boards until the very end of your residency when you can use a tool like this, you know, throughout your residency. Yeah, I, I completely agree. , I, I personally think that [00:23:00] the, you know, the oral board test is just about the perfect test for assessing clinical competency. It's because in real life, there's no one right answer, right? When you're caring for patients. And so the oral boards reflects this. , One of my favorite questions on the oral boards is when you ask, you know, a resident, what would you do? Right, not what can be done. What are the options, but what are you going to do in this scenario? Right? It forces them to actually think what they would actually do in that situation. This is why I make my interns practice for our boards right alongside my senior residents. Yeah, I agree with, um, with. Pretty much everything that has already been said. Our job in training is to learn as much as we can about what we, what we'll be dealing with. When the buck stops with you and something intimidating and previously unseen to you coming, comes rolling in the door, all you really have is your support system, which, uh, for me was my mentors like Dr. Swenson. And you're training to back you up. So the more clinical scenarios you expose yourself to, the better you'll be prepared to handle what comes your way. [00:24:00] And in, in, in a way it is a numbers game too. , Just like the more patients you see, the more experience you'll become. , Sometimes you do have to adapt and make it up as you go. , Accounting for various comorbidities and anatomical aberrations or whatever. And you never really wanna do that without a solid foundation. And I think that this tool, along with Behind the knife and various other resources are. Or to help you build that foundation. And I just wanna say thank you to everybody. And, uh, the final point that I'll say is this is, the more that I've used this, the more I'm convinced that everybody should be using this. Um, it, it really has helped me quite a lot. And so now with, uh, without further ado, I'm going to play a full example of the oral board simulator. , keep in mind that I will probably say something that is wrong, but I am a, uh, I have finished my clinical two and I'm going to clinical three. And, , it's actually exciting to see how it. And how it responds to that. And I'll also play some expert, , some excerpts from the feedback and it's truly quite fantastic, and definitely things that [00:25:00] I need to work on. So, , enjoy. A 48-year-old man presents to your clinic with a bulge in his lower abdomen that has been gradually increasing in size over the past year. He reports mild discomfort, especially when standing for long periods or lifting heavy objects. He has a history of obesity and underwent an open appendectomy 20 years ago. On examination there is a palpable reducible mass in the lower midline of his abdomen measuring approximately five centimeters in diameter. His vital signs are stable and there are no signs of bowel obstruction or peritonitis. What would you like to do now? At this point I would like to obtain a further history. I would like to know what his comorbid conditions are. Is he diabetic, and if so, what is his last hemoglobin? A1C. I would also like to know if he is a smoker and I would like to know what his BMI is. He does not have a history of diabetes, so there is no recent hemoglobin A1C available. He's a current smoker [00:26:00] averaging about 10 cigarettes per day. His body mass index is 34. Would you like any additional information or do you have a next step in your management? The next step in management for this patient is that I would recommend that we electively repair his hernia, however, in order to optimize him and reduce the likelihood of recurrence. I would like him to quit smoking as well as enroll in a weight loss program. Understood. The patient agrees to attempt smoking cessation and weight loss, but he's eager to proceed with surgery due to persistent symptoms and concern about the increasing size of the bulge. What is your preoperative workup and preparation for this patient in preparation for this patient? Besides the weight loss and smoking cessation, I would also obtain a hemoglobin A1C. However, given his age, I would not do any further cardiac [00:27:00] evaluation, and I would tell him that I would like him to aim for A BMI of less than 30. However, it is more important that he quits smoking. You have recommended optimization and would like to proceed with elective hernia repair. What is your planned operative approach and what are the key steps you would take in this repair? First, I would like to know what is the size of the fascial defect on physical exam. On physical examination, the facial defect is estimated to be about four centimeters in diameter, located in the lower midline, just inferior to the old appendectomy scar. How would you proceed with your operative approach and what are the key steps? I would perform this minimally invasively either robotically or laparoscopically. I would enter the abdomen safely, likely through Palmer's point, as this is far from the open appendectomy site and not likely to have adhesive [00:28:00] disease. After entering the abdomen safely, I would perform a lysis of adhesions, reduce the hernia sac, and then cover the defect with a large piece of mesh. I would ensure that the mesh adequately covers. Uh, with approximately five centimeters of room on each size, in this particular case, I would likely use a 15 by 15 centimeter mesh. The patient is taken to the operating room and you proceed with a minimally invasive approach as planned. During your lysis of adhesions and reduction of the hernia contents, you encounter dense adhesions between loops of small bowel and the hernia sac with one area appearing particularly tethered and difficult to separate. What are your next steps in managing this intraoperative finding? If it is safe to continue minimally invasively, I would do so. However, if it is not, then I would convert to an open procedure [00:29:00] via a midline laparotomy. You proceed carefully, but during alysis, there is a small enterotomy in one loop of small bowel identified immediately the surrounding bowel appears otherwise healthy. How would you manage this intraoperative complication? For a small enterotomy, I would either close transversely or I would resect and perform a side side anastomosis. This is dependent on the size of the enterotomy and whether or not it is feasible to perform a primary repair. If it is small enough to repair, then I would repair it transversely in two layers. You elect to repair the enterotomy primarily in two layers. Please describe the key steps you would take to perform this repair. The key step to performing this repair is to do a layered approach. I would [00:30:00] ensure that for doing so, I capture an adequate amount of submucosa as this is the strength layer of the small bowel anastomosis. The Enterotomy is repaired primarily in two layers. You complete the hernia reduction and proceed to place the mesh given the contamination from the enterotomy and repair. What is your decision regarding mesh use and what are your next steps for abdominal wall closure? If it is a small enterotomy and there is no gross spillage, then I would proceed with a regular Vicryl or other mesh. However, if there is significant contamination, then I would place a biologic mesh instead, and then I would postoperatively counsel the patient that he may have a slightly higher chance of recurrence. However, for enter autotomies that are recognized immediately and closed, there is a very low likelihood of mesh infection. You have completed the [00:31:00] hernia repair with mesh placement and closed the abdominal wall. On postoperative day two, the patient develops increasing abdominal pain, tachycardia, and a low grade fever. His abdomen is moderately distended and tender with mild guarding. What is your differential diagnosis and immediate management? Differential diagnosis at this point includes leak from the enterotomy site, a missed enterotomy, a small bowel infection, a deep infection such as an abscess or a new perforation and unrecognized enterotomy. At this point, I would like to obtain a CT scan with IV and oral contrast. I would also make sure that the patient is started on antibiotics and is receiving adequate fluid resuscitation. What specific findings are you looking for on the CT scan in this scenario? On the CT scan, I'm looking for a differential diagnosis or evidence to support any of the, um, diagnoses on my differential as listed before.[00:32:00] The CT scan of the abdomen and pelvis shows a moderate volume of free fluid with a small pocket of extra luminal air adjacent to a loop of small bowel near the hernia repair site, but no large abscess or organized fluid collection. What are your next steps in management at this point? I'm concerned for a leak, uh, and in this case, given that there was no anastomosis specifically, I'm concerned of a enterotomy. Therefore, I would elect to take the patient back to the operating room for exploration. The patient is taken emergently back to the operating room for re exploration. Upon entering the abdomen, you find a small defect in the previously repaired bowel with localized contamination but no widespread peritonitis. Please describe your operative management at this time. At this time, I would wash out the abdomen and then I would perform a [00:33:00] resection of the piece of small bowel that has an enterotomy, and then a new side to side anastomosis. After resecting the affected segment and performing a new side to side anastomosis, how would you manage the abdominal wall and mesh at this time, considering the contamination in recent repair, I would place a biologic mesh or a, uh, a biologic mesh or a vicral mesh. You complete the abdominal wall closure with biologic mesh postoperatively. The patient recovers slowly, but develops a persistent seroma beneath the incision, which is not infected and is causing discomfort. How would you manage this postoperative seroma? For a postoperative seroma, if it depends, it depends on where it is below the incision. If it is very superficial, then I may open the incision, allow it to drain, and then pack [00:34:00] the wound and allow it to close by secondary intention. However, if it is slightly deeper, then I may aspirate the seroma for symptomatic relief. Understood. The seroma is superficial and not under tension. After opening the incision and allowing it to drain, what postoperative wound management strategies would you employ to promote healing and reduce the risk of infection? I would pack the wound, although if it is larger, then I would consider negative pressure wound therapy. Thank you. Let's say instead that the seroma had become infected with erythema and purulent drainage from the wound, what would be your management approach in this situation? In this situation, I would not consider aspiration and I would just open the wound. In addition, I would give a small course of oral antibiotics. Thank you. [00:35:00] If the infection were to progress and you noted signs of mesh infection, such as persistent drainage and exposed mesh, how would you manage this complication? In that case, the patient would need to go back to the operating room for mesh excision. Your oral board exam is now complete. ​ Okay, so Dr. S Swenson, what did you think of my attempt at this scenario? I, I loved your effort. , It's very clear listening that, uh, that you weren't, first off for all of our listeners, right? You weren't scripted here, right? That you were pausing Not at all. And you had to think, right? So this was not, let's show you how everyone, how smart Iman is. I think you're incredibly smart, but you're also, you know, you've only completed two, uh, graduate years, um, of training. And so I thought overall you did an excellent job. There were certain things you could have focused on, right? The types of mesh to use, right? More detail perhaps on the dissection. , But overall I thought you did a great job. , And when I say that, that is how I [00:36:00] typically would give someone analysis if, you know, if you were Tala, for instance, right? You were preparing for it, right? I would've told Tala general good job if I didn't see any big red flags, any issues, right? Kinda like I discussed earlier, right? Like. You know, thumbs up, good job, study a few things. Right, right. Um, as I look at what the AI said and in our performance evaluation, it's, it's much more thorough than that. , I think you have the results in front of you. Do you wanna share some of the things that, that the AI performance, , said after looking at your exam? Yeah, of course. And, and as you know, it goes over many different details of why it justifies the score that it gives you. , And if you just go to the key procedural elements, it says, the description of my hernia repair was missing several key elements. And I like this. It says, explicit identification of safe entry technique and avoidance of adhesions, , detailed license of adhesions and safe dissections in planes. . It mentioned a max fixation technique. Technique, which I, which I didn't mention at [00:37:00] all. , It told me that I need to, that I omitted a stepwise description of a two layer closure, which is true. I, I, to be honest, I couldn't, I didn't know that's true. How to do a two layer approach. So, uh, it it, and it caught you. And it caught me, yeah. It caught me, and it, it completely picked out that I. Didn't, didn't know what I was talking about when I said two layer closure. , And it, it pointed out a leak test after repair, which I never would've thought I, like, I don't think any attending would've would've pointed that out to me and I didn't think about it. So, um, but it's absolutely something we should all be doing. Absolutely. A quick leak test after closing an enterotomy. Absolutely. Yeah. And, and it, it was great. And you know, it, it also mentioned, , that part of my knowledge gap was that, uh, the exact line in here for, for everyone is, it says management of mesh in contaminated fields was somewhat vague and did not discuss risks of synthetic mesh in contaminated settings. That's true. I like, I, I knew that I needed probably a biologic, but, uh, I, I couldn't really give any further detail than that. And as I said it, I [00:38:00] knew that was a weakness and, uh. I think the LLM caught me, so, yeah. Um, yeah. And then why don't you go over the study plan again? My recommendation to you was, good job, go, you know, look up a few little things. What did the study plan tell you? So study study plan's really great. It, it says, review the operative steps for minimally invasive ventral slash incisional hernia repair, focusing on safe entry, dissection planes, peritoneal closure, and mesh fixation. I think in all of those aspects, I sort of, uh, was very vague. So it, it got, it, it has great points all the way around. Um, and yeah, I think the, the feedback here is, is something that is so comprehensive, it really makes you feel like that was a very worth it scenario. If that, if that makes sense. I thought that, again, I thought you did a good job, but that the analysis here is, is spot on for how you can improve. , and just, I mean, listening to your first few and then this, it's clear you're becoming comfortable with this. And that's the whole point is this is not a test you've ever taken before. It's, [00:39:00] it's not multiple choice. And so, , that's the oral boards are about practicing. That's what made such a difference for me and Dr. Shin. Right. And that's what we've tried to create here is the opportunity where you just practice saying. Two, you know, a two layer closure. I should include that next time. Right. Or just the comfort with pivoting. , And so you've already gotten significantly better. I failed just 14 tests, right? Uh, excuse, well, I guess you said about 20 to 30. Um, and so well done on that improvement, right? And I mean, by the time, you know, you graduate ready for the boards, I suspect it'll be easy for you, right? Because you'll have done this so many times, just like right. A lap. Coley should be easy by the time we graduate, because we've done it so many times. The oral boards should be easy because you've done it so many times, right? And I, I, I truly believe this is. One of the most impactful steps to get there. Um, awesome. And, and the best part is it's accessible. So I think that wraps it up for this. And any last words, Dr. Sunset? No, thank you so much again to Patrick and for [00:40:00] being on, of course, for yourself. , The website is up and running now. , Behind the knife, , oral board simulator.org. Org. , And then of course be on the behind the knife, , website. , This is our beta version, so please go try it out and give us your thoughts and feedbacks and let's make this something that, you know, we can all use and benefit from.