Pediatricians are some of the most meticulous clinical professionals, and for good reason. The tiniest error in a NICU or PICU can have massive implications for a patient. This detail-oriented mindset often leads pediatricians to gravitate toward clinical informatics, as the same outlook applies to the world of electronic health records (EHR) and clinical data. In today’s data-driven environment, nothing exists outside the EHR, so how can pediatricians and other clinicians think more like informaticians to improve outcomes and streamline operations?
On today’s episode of In Network's Designing for Health podcast feature, Nordic Chief Medical Officer Craig Joseph, MD, chats with two doctors from Cincinnati Children’s Hospital, neonatologist and biomedical informaticist Kevin Dufendach, MD, and CMIO Andy Spooner, MD. Their discussion centers around their early days in the NICU, why so many pediatricians end up also working as informaticists, and the role of AI in reducing documentation.
Listen here:
In Network's Designing for Health podcast feature is available on all major podcasting platforms, including Apple Podcasts, Amazon Music, Google, iHeart, Pandora, Spotify, and more. Search for 'In Network' and subscribe for updates on future episodes. Like what you hear? Make sure to leave a 5-star rating and write a review to help others find the podcast.
Want to hear more from Dr. Joseph? Order a copy of his book, Designing for Health.
Show Notes:
[00:00] Intros
[05:30] The early days of clinical informatics
[09:02] The exhaustion discount
[16:20] The future of progress notes
[21:32] The promise and perils of AI in healthcare
[32:36] The differences between informatics specialists and generalists
[39:14] Why pediatricians think like informaticians
[40:48] Well-designed things that bring Dr. Dufendach and Dr. Spooner joy
Click here for Dr. Spooner’s Pinterest page with his favorite well-designed things
Transcript:
Dr Craig Joseph: Gentlemen, welcome to the pod. Very excited to have you here. Andy, let's get started by you telling us how you got to be in this position of power, this unholy position where you are seemingly in charge of healthcare informatics for all of pediatrics in the United States.
Dr. Andy Spooner: You know, everybody's got a story of some task they tried to automate when they were a resident. And for me, it was calculating in nutritional loads for neonates and got a lot of a positive feedback. Actually, my first clinical informatics project wasn't that I've got to be honest with you. It was creating a scheduling system for medical students to schedule them for their outpatient pediatric rotations, which is something that they used at my medical school for about a decade after I left. But anyway, I've always been looking for computerized solutions for in healthcare. But then when Meaningful Use came along and everybody wanted to a CMIO I just kind of climbed on board that movement and found my way into a wonderful job at Cincinnati Children's Hospital, where I've been for almost 17 years, working with their EHR project, and that’s my story.
Dr. Craig Joseph: So, Kevin, apparently all good informatics careers begin in the NICU. For me I think it was the PICU but for you is the NICU just like with Andy. Tell us how it all started. We want to know everything about the neonatal ICU.
Dr. Kevin Dufendach: Well, for me, it began in the NICU, and apparently, I decided to stay there since I’m still a neonatologist. But one of the things that I love about the NICU is that little things matter. And so, when you're when you talk about things like the volume of IV tubing, if you're giving an adult, you know, three liters of fluid, nobody cares about the half a milliliter of IV tubing. But when you're running something at 0.1 milliliters per hour, that little bit of tubing matters. And so, for me, it was always little things that matter. And that really, I think, really appealed to my mathematical and kind of if this then that computer-type brain. So, one of my first informatics projects and what was in the actually in the newborn nursery, not necessarily the NICU itself, but it was there that I learned the importance of creating resources that impact people's workflow. And when you change someone's workflow for the better, when you when you design for the work that needs to be done, then you know, you can make a lasting impact. So, one of the things that I did was I created an Excel spreadsheet that would you just fill it out with, of course, the birth information. And then every day you'd update it with the number of like number of feeds, the number of voids and stools, and you put the bilirubin and all that stuff in there. And the cool thing was we used that tool both for rounding because we print it off and we'd carry it around as we would go and see the normal newborns with our attending. But then we also, the very last cell in it was one that we didn't print off because that was the progress note for the day. And so basically after you filled everything out, you just took that last cell and copy and paste it into the electronic health record and you could have all your notes written in 5 minutes or so, and I knew it was a success when I talked with my program director three years after I left. And he said they still had about five copies of it on a backup drive to make sure that nobody broke the Excel document, that everyone used to write their notes in the newborn nursery.
Dr Craig Joseph: Somewhere an information security officer is having a headache and they don't know why.
Dr. Kevin Dufendach: But it works because there's nothing in healthcare that happens outside the electronic health record these days.
Dr Craig Joseph: Kevin, you are you are exactly right. And let's just keep going with that, with that attitude. I do want to point out, though, that, you know, you're designing you were designing solutions to workflow problems, often with very, very basic technology, sometimes even on paper. And so it's not so much the technology part of informatics to you know, it often really just comes back to the information and are you collecting it in an easy way and then leveraging it in an easy way. And I've been amazed that talking with clinicians who are involved in informatics, they always, not always, but almost always begin with stories like you've told of, well, I just had this problem and I solved it for me. And then someone next to me saw how well it worked, and then they wanted it, and then it just kind of goes by fire. And you know, you got it. When people are demanding, you know, you've solved a problem. When people are demanding that solution.
Dr. Kevin Dufendach: Yeah, absolutely. There's nothing that says in informatics that it has to be digital. You can start off with, you know, even with, with paper or with magnets or other innovative ideas and start with that as you're going through the design process.
Dr. Andy Spooner: Yeah, my first NICU application was just the set of printed tables so I could do all the computations pre-compute everything and then it would be transcribing the numbers to various places. But it, cut out the whole step of having to get your calculator out. And I made it for myself. And one day the NICU director came over and said, what are you doing? Where did you get those tables? I'm thinking, my goose is cooked. Because he was not known for being a kind man necessarily. He grabbed my tables and made about 200 copies of them and started distributing them to everyone else. I knew that I was onto something. So anyway, and they were they were spread basically spreadsheet printouts.
Dr Craig Joseph: So yeah, I think it’s great. And as I mentioned, I think for me it was in the pediatric intensive care unit, the PICU, and I was trying to order TPN Labs or trying to order TPN every day based on the lab results of the previous day. And I just couldn't figure it out unless I put it in a you know, in a table in a way that was meaningful to me. And I created that for me. And on the weekend, I think on Friday afternoon, I gave it to another resident and boy oh boy when I came back on Monday, everyone had them. And it was it was pretty cool. They took one of my copies and then made a copy and then whited out all the information that was there and then made 200 copies and there you go. That's how clinical informatics begins,
Dr. Kevin Dufendach: I think that just illustrates the importance of having that interface. And that's what we in clinical informatics, so often we just play the role of that interface between the clinical side and the clinical work that needs to be done. And the, the innovation, the electronics, the digital side as well, the computational side. So it's really fun. It's really fun being able to be that that interface right in between with the content and content experts on both sides.
Dr Craig Joseph: Yeah well, it's good to be needed and it sounds like we're going to be needed for a little bit. Andy, one of the things that when we were preparing for this episode, you told me that you're worried about a couple of things. One of those is that all the rules are for clinicians and how they do their work. And you, you mentioned an exhaustion discount and I'm curious about who gets that discount. Is that 10% Like what is that discount and how can I get it?
Dr. Andy Spooner: So, the way things have worked since the mid-nineties is we do things in clinical care and then we document what we do, and the agreement is if we want to get paid a little bit more, we document a little bit more. So, it often seems like we're getting paid to document. I mean it's not really true, but if you, if you think about it in a certain way, it does seem like, you know, the more you write, the more you get paid. And I fully understand that it's about the complexity of the care you’re giving. I get that. But we have this intermediate step of having to document it, which as far as I know, no other profession works that way. But we work that way and we all sort of know that writing a note for a very high level of complexity takes a lot of time, and we often get kind of tired, and we essentially down code quite a bit because we end up with just saying, okay, well this is this is good enough. I'm going to maybe do my coding a little defensively and maybe down code a little bit just to make sure I don't get into trouble and will maybe not take the time to enter that extended service code that we perhaps deserve. Truly, I mean, that's probably the most neglected code there is, is all those extended service codes. But even some of the high-level codes will downplay those because it's just hard, it's just hard to document you don’t have time to document, you need to get home and all that sort of thing. So here comes AI and we have this promise from AI that maybe we don't have to do documentation anymore. Maybe if the computer can pay attention to things that are going on and listening to the conversation and looking at our orders and taking notice of what's going on in the environment, that perhaps the computer could just generate the documentation. And if that's so and that works, and that's a little bit unproven right now. But the results seem promising from all the large language model-based systems that are being pushed now, then perhaps we will be able to code that extra service time and a higher level of service. And perhaps we won't have to worry about how difficult it is to write something because the computer's doing it all for us. And so right away, one of the things that I worry about a little bit is, well, if computers listening and constantly tossing in CPT codes on our behalf and putting together notes, pretty soon that what I call the exhaustion discount, where you’re just too tired to put it all together will go away, and we'll start billing at a higher level of service and submitting more codes and charges. And you know, that's disruptive. It might be in perfect accordance with the rules. It might be exactly analogous to or not analogous, but exactly compliant with the rules. And yet cost more money to payers than what it currently does. And that's going to cause some disruption and some hard questions. So, the payers might say, well, you know, this care is costing more than it used to. It's all coming out in professional charges. What's going on? And probably there's going to be some adjustment of what we get paid and that will dilute the reimbursement somewhat, I'm predicting as we go to this. So, will we end up getting paid more? I mean, everybody likes to get paid more, but is that really what's going to happen? When is it even necessarily the right thing? I mean, do we need to scour every second of our existence to make sure there's something billable out there? Or do we just sort of, have we reached an equilibrium point where we document what we can get to and let the rest ride.
Dr. Kevin Dufendach: Well it seems a little bit to make an analogy, it seems a little bit like grade inflation. And if everybody's getting the B plus, A plus, and so somewhere in there, we need to be able to pull out that you know that curve and you know, come up with some differentiation. Who really is the complex, complex patient versus the patient that just looks complex because, you know, the AI pulled a bunch of new diagnoses or a bunch of new things that we did to reach that next level of complexity.
Dr. Andy Spooner: And if AI was going to be looking over our shoulder anyway, you know, one of the next things to happen is some suggestions like, hey, you know, if you just bring up one more chronic illness and ask the patient how they're doing with this chronic illness that I found in the patient's chart, you could easily make this a level five or something like that. You know, there have been glimmers of this kind of technology for a while now where if you I mean, I used an EMR years and years ago that did something similar because you checked off all the review systems and that sort of thing, and it'll give you suggestions for how to better support your level of service, which is dangerously close to something that sounds a little bit illegitimate. If someone's coaching you to charge more, that has a potential for causing some disruptions, too. So, it's an arms race. I think we all characterize our relationship with our payers as a bit of an arms race where we're trying to use tools to optimize our reimbursement. And as the tools get better than the payers are going to say, well, gosh, maybe, maybe we shouldn't have been paying that much all along. So, it's going to it's going to cause some reshuffling, I think, and we may end up in a period where the haves, the ones that have the AI assistants will be getting paid a little bit more than the have-nots. And there'll be some interesting questions to answer.
Dr. Craig Joseph: Kevin, do you think that we’ll be writing notes in five years or ten years? Where do you think this is going if are we not just going to record every clinical interaction? Maybe not in the hospital necessarily, but certainly in the clinic just record the interaction and keep that recording somewhere and then everyone can have their own version of the note. Then the payer can have one version and the quality folks another and the clinicians a third. And if anyone has a question about what really happened well you just go back to the recording, which we're going to keep in perpetuity, is that going to happen, was that dystopian or utopian?
Dr. Kevin Dufendach: I don't think that the narrative will go away. And in fact, my hope is that the narrative becomes more of a focus. And then the other stuff that right now we're dumping into our note bloated notes, that stuff kind of gets extracted in a way that becomes maybe more interactive. Why do I need to record the current ventilator settings in my NICU progress note? There's really no reason to do that other than to record a snapshot in time, which I will say, even though it might be unpopular, I will say that that is actually useful. And right now, my EHR doesn't give me a great way to get that snapshot of what was it that I was looking at that moment in time to make that decision. But what I would love to do is I would love to remove a lot of the additional fluff, I mean, why do I need to write that? I am going to start Lasix at one mg per kg at twice daily. Can't you see that I just put the order in, just put that, you know, that should just be able to be referenced somewhere. But why did I start Lasix at one milligram per kilogram twice daily? I mean, yeah, maybe AI could be listening to our conversation on rounds, maybe it could transcribe something from that. Maybe that's what goes into a progress note or into the documentation piece. But I would, I hope that the narrative does not go away completely and rather that the narrative and then the objective piece get further defined by what is actually there and what is actually happening that we are able to record in a structured manner in the whole electronic health record.
Dr. Andy Spooner: Yeah, I finished my residency in '91 and the CMS guidelines for documentation came out in 95 and 97. So I was around for a lot of those classes for how to document, you know, and I remember having the thought, why don't we just get a whole bunch of-I think at the time the technology was Betamax. We may not have had VHS quite yet, but I want to just get a whole bunch of tapes and just tape everything and anybody wants to know. They can just, you know, we can send them the cassette, you know? But I truly think that documentation as we know it today can be and should be made illegal. In the sense of its forbidden to waste a physician's time in this way. It's almost immoral to take a rare resource like that and have them sit around and document all kinds of things that already exist. I mean, I agree about Kevin's comment about how you want to maintain a narrative and you want to have yeah, you want to keep some notes for yourself and for your cross coverage and for, you know, for the family to understand what's going on. I mean, that's a far more important objective than a lot of the other things that we write notes for. Everybody talks about well, we write a note because of billing. Well, it's not the only reason we write notes. We also write notes, at least when we're trainees we write notes because we want a script to read from. We want to have every piece of information there so that nobody has to find us looking in the chart for something. That it's all going to be written out right there. So, I can just look at my note and talk and that habit sticks with us and we end up long past our residency just writing these long scripts where we record every little piece of information. And people justify this by saying, Hey, I'm going to do my thinking in this note and I want to bring all the things I need to contemplate into one artifact. And I can certainly understand that sometimes. I mean, there there's complex thinking to be done and don't want to be overloading your head with, you know, where to click next. But if the computer's got to help us here at some point and I think of this, I'm hoping in ten years we'll look back at what we ask ourselves and our trainees and our colleagues to do and just declare that to be a ridiculous waste of time, that the computer can now help us out a lot more with.
Dr. Craig Joseph: Well Andy, speaking of helping us out, there was a recent study that showed that an AI can be more empathetic than a physician from the perception of the patient and where do you think that's going? Are the computers just better listeners and doctors are just bad? I'm not saying that, but I think you are.
Dr. Andy Spooner: No, I'm not. So, first of all, you're referring to the UCSD study where they looked at physician’s replies on Reddit, and I'll say that again, physician replies on Reddit. To things that were generated by a large language model that had been instructed to be empathetic. Okay, you take a large language model, and you tell it please be extremely polite. It will be extremely polite. That's because that's how large language models work, unless they're malfunctioning or something. And I don't know where you'd get physician replies besides Reddit, I guess you get from an EHR, but that would be a little bit too HIPAA risky to do that sort of study. But, you know, you can take these large language models and ask them to be however you want them to be. And if you, anyone who's been on hold or the car rental agency is familiar with the almost obsequious tone of the little chat bot that pops up, and then how profusely apologetic it is about your wait time and all this kind of stuff. And I would just remind people of what happens every year when college acceptance letters come out, every letter that goes out that’s saying you didn't get in starts with the phrase ‘we regret to inform you’. And, you know, we look at that phrase, it's almost invisible at this point because we know that a rejection letter starts that way. But here's the thing. Does anyone actually regret to inform you of that? Is there someone sitting at Yale going, ‘gosh, that kid really should have gotten in, and I'm wracked with regret.’ No, I mean, of course not, they’re just being polite. So, my question is, we hear a lot about how this large language model can create the very empathetic, very, you know, have a nice bedside manner and give you a very nice and polite response. But it's just what it's being told to do. And the question is, will, is anyone really being fooled by that or is there any value in saying, AI is valuable because it's polite? I don't think so. I think AI can be as polite or as rude as you tell it to be. And I don't think it really matters to people. Once I think once they realize that it is a bot, then it probably didn't matter that much anyway. I mean, if you knew that (as everyone does) that letter from the college telling you didn't get in was a form letter. Everybody knows that. And they obviously understand that there is no regret and that's what the letter template says. Nobody actually had that thought. When the same thing is applied to a response, to an AI generated response to a request, and it's extraordinarily empathetic, is that a good thing? I guess. But is it important? I don't think so. So, I mean, I don't see it as a major driver toward rushing toward AI. Just because it's more empathetic than a doctor. I think it depends on the doctor. It depends on AI model.
Dr. Kevin Dufendach: I think that that concept is probably more helpful when it comes to maybe synthesizing or summarizing or at basically coming up with communications for families or others to know what's going on. So, if I write my NICU progress note in a way ... and I do write my NICU progress note in the way I know that I'm communicating with other health care providers so that they know what's going on with that patient. I think that is appropriate. That's what I'm using that note for. I know as well that, you know, 21st Century CURES, absolutely. The parents are probably getting in there and looking at some of that as well and looking at the some of that information. But I don't try to change my tone of my medical progress note for a family. But how great would it be if my progress note was all of, I don't know, two or three sentences on the actual progress for the day? This gets back to what I was saying earlier about synthesizing what I'm actually thinking and then the additional information could be added by AI perhaps in a very compassionate and, you know, just understanding manner and in a way that's written for that family to be able to understand maybe the family even has little sliders they can adjust the amount of medical jargon on there. After they've been here in the NICU for seven months they probably understand a lot more than they did on their seventh day.
Dr. Andy Spooner: You know I've seen functionality where you can highlight a piece of text in an EHR and you right click on it and say, you know, rewrite this at the third-grade level, or rewrite this at the eighth-grade level or something. You know, I mean, sure, why not. I mean to your illustration earlier, Craig, about having a recording someplace and you can output the results of that recording for various use cases. This is, you know, here's the oncology progress note for the family and here's the oncology progress note for the payer. I mean, all the primary data is the same. It just has multiple outputs that look different. Right now, that's a manual process so it's not going to happen really. But as we get better, why not stick to the original data? The audio recording, the video recording, the orders, the log of the user actions. Just keep it simple. That's the primary information that tells the universe that some care got done. And this business about having to write it up in various ways, it just, I hope we can get away from that and just say that's not appropriate. I mean, if you're spending time doing that as a doctor, you could be doing other much more valuable things instead. So, let's do that and not do not do all this writing of stuff. I mean, imagine any other profession that you take your car to an auto shop and you know, they tell you what the parts cost with the labor was and there's your bill. Or instead, we say, no, now I want you to describe exactly which rack the car went up on and what you did first and who you talk to and all this detail. It's like, I get that there's a lack of trust. I mean, it's all based on, you know, very well justified concern about fraud. Okay. But we spend so much time doing all these notes and it's a it's a waste of a precious resource, honestly.
Dr. Craig Joseph: Yeah. So, first of all, very interesting comments. And I think that those are pretty mainstream views. Physicians are burned out, we know this, clinicians, people that work in the hospital. Clinicians or not are burned out and part of the reason is this documentation and kind of detail-oriented writing that is required to get people to, you know, not get sued and to be able to get paid. One thing, though, I saw regarding, you know, empathy from a large language model, I read this discharge summary for a new onset diabetic. I think a nine-year-old, something like that. And they asked the LLM to consume the chart and to write a discharge summary for both the clinical one and also patient instructions. And I was amazed at reading this and this is probably because I'm not a pediatric endocrinologist, but there was a line in there that said you might think that it's your fault that you got diabetes, but this is not something that you can control. You know, to me that was pretty revolutionary because now, again, if I'm a pediatric endocrinologist, just maybe that's a common thing. And kids often wonder if they don't say that they're responsible for their disease. But I thought that was great, that they're like, that's something that I would love for someone to point out to me or something to point out to me like, hey, the patient didn't say this, but you didn't ask. And they statistically are thinking it, so you might as well try to address that elephant in the room. That’s pretty amazing. And that's a service I want.
Dr. Andy Spooner: Yeah, going back to this idea of having primary data somewhere that generates different outputs, it should be, you know, I mean, I guess in some future data it should be query able. If you've got a concern, what could I have done to prevent my child's type one diabetes at age nine? You could I mean, you might ask a doctor about that too, but why not ask the data too. I mean, if we're really getting into this, all this AI stuff shouldn't be just to, you know, generate notes. It should be to help with communication and to actually take data and turn it into information.
Dr. Craig Joseph: Makes sense. Kevin, we like to talk about design and healthcare at a very high level. You have said previously that you consider yourself an informatics generalist, and to me that makes me wonder what's an informatics specialist and you know how you do their work? How do they design their workflows? How do they address technology in a different way? So, what do you mean by informatics generalists?
Dr. Kevin Dufendach: Yeah, I like that distinction. What is an informatics specialist then? Well, I think it gets back to a little bit just me personally, I really love delving into all things informatics and at least having an idea of how things work. So, I can talk about things like machine learning and artificial intelligence and neural networks, and I have a general understanding of of how they work. And I've played around with them a little bit. And I think that's fun. But you know, I don't necessarily consider myself to be a machine learning expert. That's not what I'm doing day in and day out. Over time, I found myself really acting more in that role of the interface between that clinical care and the processes, the workflows that need to happen. And one of my specific interests in programing, my wife calls it my Sudoku. And so, I love getting in there and trying to figure out how to get the computer to do what it is that I want it to do, whether that's creating an algorithm, you know, a deterministic algorithm or having it respond to different user interactions and that kind of thing. So I think it's important we as clinical and informaticians, that most of us are informatics generalists, that we can do things where we can build directly on the electronic health record, we can speak to things like usability with effectiveness, efficiency and satisfaction of use. We can evaluate workflows, and then we have an understanding. I think it's really important that we have that understanding of neural networks, artificial intelligence, machine learning, how do all these things work and how can we respond appropriately and intelligently as as informatics generalists.
Dr. Craig Joseph: I love it. And I think you've commented also previously that only pediatricians should be allowed to be informaticists. Did I get that quote wrong?
Dr. Kevin Dufendach: I think that that's the gist of it in general.
Dr. Craig Joseph: Why? Why what's so special about you pediatricians?
Dr. Kevin Dufendach: Well, I mean, specifically, I think with machine learning, I obviously don't completely think that. But I do think that pediatricians or really doctors in general have an advantage when we start talking about this machine learning concept, because we do have an understanding of how the brain works, how we as humans learn. So, we understand things like neurons and axons and myelin sheaths. And really, if you think about a neural network, a neural network is basically a bunch of linear algebra matrices that all kind of get scrunched together. I realize I just lost everybody. So sorry for ruining your listenership here, but that's basically all you're doing is you're reinforcing the myelin along those linear algebra matrices. And we understand how somebody goes from you know, if we think about somebody learning to ride a bike. But when they first start, you know, there are a lot of scraped knees and, you know, not a lot of bike riding that's exact that's going on. But through continual reinforcement, you can see that the child is gradually learning to develop their balance. And every time that they're doing something, they're getting a little bit of feedback and that keeps on reinforcing. And that's the way that the neural networks work as well just instead of instead of reinforcing myelin sheaths, you're adding a little bit to the numbers here and there so that one input is weighted more than, say, a different input to get to the outcome.
Dr. Craig Joseph: I like that. I agree maybe for a slightly different reason. I've often said that pediatricians are good in the informatics and the CMIO, CHIO type roles because we're used to dealing with children and yeah, there's some smiles there. I'm not saying that physicians act like children often, nor am I calling out various specialties such as surgeons. I'm not doing that. But there are others who have done that and to each their own. Andy, thoughts about this?
Dr. Andy Spooner: Well, I met an informaticist recently who had been a neurosurgeon, and I thought, who in the world would give up a career in neurosurgery to be an informaticist? The answer was he was from the UK. They explained to me that all you know, doctors essentially all earn the same amount of money in the UK, and so it wasn't that big of a loss to him to switch over to informatics. And perhaps the reason why we have so many pediatricians and hospitalists in the informatics ranks is because folks who can earn a lot more money do in their specialty, aren't as tempted to spend time on things like this.
Dr. Craig Joseph: Andy really you're bumming us out here, Kevin and I are bummed out now.
Dr. Andy Spooner: That's my value add.
Dr. Kevin Dufendach: I do think that one of the reasons, though, that you'll find a lot of pediatricians that are thinking in the informatics mindset is that in pediatrics, especially in subspecialty pediatrics we're almost always busy dealing with rare diseases. If you think about it, kids are not supposed to be in the hospital. And so, whenever we are dealing with something, it's not something that would necessarily be dealt with in mass. And frankly, in many cases, it's taken a while for our electronic health record to add some of the pediatric specific or pediatric friendly functionalities into, say, our electronic health record. And I also think oftentimes we have to deal with things like workarounds or using tools or medications in ways that are maybe a little more creative. I already talked about being in the NICU and one of the things that I love about it being that I have to deal with the little, tiny things, the very small changes in, the volume of ventilation, I have to think about things like dead space in the airway. It's those, little things in those edge cases that I think lend themselves to thinking like an informaticist as well.
Dr. Craig Joseph: Typically, at the end of our podcast, we like folks to tell us if there's something that is so well-designed in their lives that it brings them joy and happiness. And sometimes it's technology and sometimes it's not. Kevin, let's start with you. Is there something there? Are there one or two things that just are so well-designed, they bring you joy?
Dr. Kevin Dufendach: Well, the thing that immediately comes to mind to me is I like to focus on and talk about usability often, and with the three pieces of usability being the effectiveness, the efficiency and the satisfaction of use. And when I think about that, one of the best examples is when you mistype your password on a Mac, do you know what happens?
Dr. Craig Joseph: It shakes.
Dr. Kevin Dufendach: That's right! That's right. When you mistyped your password, the entire screen shakes at you. It's like, hey you idiot, you just mistyped your password. And I love that because is it effective? Yeah, it tells me I just typed my password wrong. Is it efficient? Sure doesn't take very long. I don't even have to click on a box to say you got to retype your password. And is it satisfactory? Yeah, I can kind of laugh at myself because my computer just shook his head at me. So I think that's great.
Dr. Craig Joseph: That's awesome. Andy?
Dr. Andy Spooner: Yeah, there are so many things, but since this is a computer-oriented conversation, there's a vector drawing application that I love. I don't know if I can mention the brand.
Dr. Craig Joseph: Please!
Dr. Andy Spooner: It's called OmniGraffle and it's a Mac only product. I don't know. I don't honestly know why they don't make one for Windows, it would work perfectly well on that platform. Maybe someday they will. But you can immediately tell if a person didn't use this program in a diagram, cause it's going to look clunky, it's going to look crude, it's going to look like an eighth grader had a really good try at it. Whereas with OmniGraffle, you can make it look just beautiful with every pixel in place. And I don't know, there's just certain things like that in this world that bring me joy. And since we're, you know, down this path, I'll just continue one more. There is a programing language. The first programing language I learned called APL, and it stands for a programing language and those who know APL know exactly what I'm talking about. But there was a book on APL by Kenneth Iverson that just took you through it, just teaching you the language. And it's still the best textbook I've ever laid my hands on. I still have my APL book that I bought when I was in seventh or eighth grade, and it's just the most amazing exposition of a complex topic that I've ever seen. So anyway, there's a couple of things.
Dr. Craig Joseph: Well, I am joyful because both of you mentioned Apple products or software that runs on an Apple machine, and I'm a big fan boy. So, this is a terrific way to end thank you so much. Really appreciate it was a very interesting conversation and I look forward to continuing it at some point in the future.
Dr. Kevin Dufendach: Well, I do have to say my Microsoft Surface that I still have right here, version three, will blow your Apple anything way out of the water but you know, good luck with that.
Dr. Craig Joseph: I'm crushed. I'm crushed. And I'm giving you 50/50 odds that I'm going to take that out of the podcast. But let's see. Let's see. All right. Thanks again.
Dr. Kevin Dufendach: Thank you.
Dr. Andy Spooner: Thanks.