Why Robot Blood Draws Are Better for Workers and Patients - DTNS WEEKEND

Welcome to the Daily Tech News Show weekend edition gratefully provided to you and thanks for your support of this show focusing in on some stories that we talk about in the news. Last week we mentioned a Dutch company called the test row receiving the first US FDA approval for a standalone robotic device that can draw blood for blood tests. It's called the Alita. It can read a patient's arm, choose a vein and draw blood in 2 and 1/2 minutes on average. I think we said a minute 49 last week. That's the fastest it that they saw it go. 2 and 1/2 minutes seems to be the average. That's according to its clinical trial. Typical human phlebotomist takes approximately 4 to 5 minutes. Human is still needed with this robot. They supervise but they can supervise up to three devices. Human still confirms the proper tube order, the volume, all of the you labeling, all of that and you know, checks on you. Make sure you're doing okay. So I immediately sent this to Dr. Gary Tolbert to be like, all right, what do you think of this? And you were pretty excited about it. I thanks for joining us Dr. Gary. I appreciate it, man. >> Glad to be here. Yeah, it it was incredibly interesting to see the technology that we've been asking for for so long showing up in real life. We see these robots that can help us with a lot of different things. We do robotic surgeries right now. In fact, it's an amazing technology but to see it doing something that is so mundane for us on the physician side is really cool. >> All right. So before we get into what it does, why should we believe you? Like what is your experience with blood draw? I'm sure people could guess cuz you're a doctor but you know, I I think you have some more interesting experience that people might guess. >> It's possible. So back in the day, long time ago, even back before I became a physician, it was commonplace for interns and other first-year medical students and second-year medical students and third-year medical students, they were the ones that did a lot of the blood draws and collections on the the wards. When I came through, we did a little bit of that, but when I got out of medical school and finished residency, I went into practice with my dad doing what's called direct primary care. So, we were the only ones in the office. We didn't really have a ton of folks that were working with us. So, we did quite a bit of the blood draw. And because I was the younger of the two of us, my dad said that I was the one that got to do it. So, I did most of the blood draws in our office. Which gave me a lot of perspective on on how to do it, but it's also something that we do on a much grander scale throughout our training. Uh when I was a resident, we used to do things called central lines, especially at the subclavian central lines. And those are done entirely from anatomical structure. We didn't This was before bedside ultrasound was commonplace. We would have to know what the anatomy was and just blindly stick. Uh and that's putting in a you know, a 30-cm catheter. And I used to tell patients all the time, "If I can put a 30-cm catheter into a vein that's about the size of your pinky finger without knowing where it is, I can absolutely stick a tiny little rubbery vein on your arm when I can feel it." >> Yeah, so so suffice to say, lots of experience in in how to do this and and also what the hassle is Is that the right word for it? >> It's an adequate word for it. They There are a lot of nuance There's a lot of nuance to drawing blood. Because there are going to be people that have done it for a very long time and they understand how to feel, how to look. The most promising visualization when it comes to veins happened when we got a lot better access to bedside ultrasound and infrared near infrared technologies. And so, from that standpoint, over the last 10 years, it we've gotten lots and lots better at at helping even new people get better at drawing blood. But there's no real uh There's no real substitute for experience. >> Yeah, and I I don't get the sense it's something uh too many people are excited to do. >> As a patient? Yeah, absolutely. >> Yeah. Yeah, exactly. So, if we can make it faster, then that that is better. Let's talk a little bit about the Leta. What Tell Explain to us a little more what it can do and and why you got excited when you read about it. >> So, the machine takes a very simple process and it turns it into something that is robotic, for lack of a better way to put it. This is a process that we do on a daily basis multiple times. We order blood testing throughout the day in the office in the outpatient setting as well as in the hospital. And we have to have a dedicated person a lot of times because of the time it takes to select the tubes for that blood because it has to be processed in a specific way. If we have very very specified kind of draws. If there's something where we want to see a a test that we don't always do on a daily basis, we may not always have the right tubes ready and things like that. So, it's it's a process where we have to figure out what tubes to draw and the the phlebotomist usually know that right off the top of their heads cuz that's what they do. Then they draw the blood, which can be very difficult in and of itself like we just talked about in terms of finding veins and making sure we get enough blood out of them. And then making sure that the patient's comfortable throughout that whole process and making sure, you know, if somebody decides that they're going to have a vasovagal response and pass out, you have to be able to respond to that and and help them. So, what the robot's doing is it's saying, well, for the parts of this that just involve selection and location of the vein and doing the venipuncture and even putting on a band-aid, that stuff can be automated because a lot of times with the technology the way it is, we can see the vein better than we can feel the vein now that we have access to the ultrasound and infrared. So, the machine uses those two technologies to find the vein, find a good place where a needle inserted into it is going to go to the right place, and then inserts the the needle and pulls it back out without having to do a lot of digging and searching, which is where a lot of the pain comes from with a blood draw. >> Mhm. >> And it allows the phlebotomist to oversee that process and take care of the patient and actually be hands-on with the person, so you get that person-to-person care, which is what most people really want in that situation. When you're scared or you don't like needles or you don't like being stuck or you have a response. There are a lot of people that when they get their blood drawn, they get, you know, pale and sweaty. We call that a vasovagal response. That's not fun. You need another person there to help you to make sure that you're going to make it through that, okay? And so >> it's it's it's what we've talked about in the past, you and I, about centaurs. You know, you've got a a machine and a human that are working together to create the best of both worlds. They're both doing the best of the things that they can do. They're working at the top of their skill and working together, they work better than either one by themselves. >> Did did this seem like it was better at doing those parts of the things or just as good as or does it even matter in that in that scenario where you're not replacing the human, you're just increasing the capacity, basically? >> It was as good as for the most most of the outcomes in the article. If you look at it closely, the average human is going to be about depending on which study you look at, 80 to 95% effective at at drawing blood. That means that about one out of 20 people, they they can't get their blood. The machine was right around that. Uh it was very very close. And so because of that and that it that is for very specific subgroups. That's the other thing. They decided ahead of time that they wanted to to look at certain specific groups of people. And so the the fascinating thing in this case in the article was that even in subgroups of people that were considered hard sticks, which are people that it's really hard to get blood from, and people that were obese, which it increases the distance between the the veins and the outside world, that can be a difficult stick for for most phlebotomists, even. And so but the machine had a very similar rate. It was very close. It was just a couple of percentage points off compared to just a standard blood draw. >> So so a typical person also has the same amount of problems is what it sounds like. And so, this this is is at least on par, right? At least from what we can tell. Does a good job and is a little bit faster, which helps capacity. But, tell me a little bit about what you thought of the fact that they could supervise up to three of these at once. You know, when you go to the the Venestro site, they kind of show a bank of these things and someone's sitting at one of them. And the idea is the phlebotomist would just be looking at people as they come and sit down. They get them situated and they're keeping an eye on them. Would Do you That would potentially triple the number of blood draws that you could do. How did you feel about that part of it? >> It's excellent. It's a force multiplier. We've been working about 5 years in our office to get a phlebotomist and a dedicated phlebotomist to do that job. The turnover in that field is really high. They have about a 50% turnover annual. >> Oh. >> So, they have to train all of these new people and that is that adds time to not only the the people that are already doing the job, but also for the new folks, it takes time for them to get up to speed on being comfortable doing those things. With three machines and one trained phlebotomist, we don't need as many of those people to do that specific job. We don't We can still focus on training the new folks for the harder sticks or for the sticks that need to be done without the machine cuz there's still going to be a lot of blood that has to be drawn without using these machines. So, it it is a force multiplier. It is the same thing that we talk about with the spreadsheets and locomotives and all of the things that we've talked about with innovation. It is taking things that can be done by somebody other than a human and isolating those and allowing the human to do the jobs that only a human can do. >> Yeah, and I I I don't think we ever talk enough about the number of types of jobs that there aren't people to do, the staffing shortages that you were talking about. So, this kind of thing really helps with that. I wonder, this is not covered in any of these studies, if having the robot do the kind of tedious part of this helps job satisfaction and could reduce that turnover because you're just doing the the patient interaction part, which, you know, depends on your personality type, but for a lot of people that that's the pleasant part, chatting with the patient. >> Absolutely. It could kind of go either way depending on which your favorite part is. >> Yeah, yeah. >> have get very zen about doing their job and that can happen with blood draws. You get into a groove as they say. That can happen, but for the folks that are more interested in the patient facing part of things, this takes some of the pressure off in the same way because you can sit and talk to the patient just like you would if you were drawing their blood. You can still ask them questions. You can make sure that the machine's doing its job, so you're still supervising. You're still doing all of the same kinds of processes, but you have the brain space to be able to do that multiple times over and focus on other things, looking at the patient. If you're looking at the arm trying to hit the vein because it's down in a tiny little crevice and you >> Yeah. >> it is, but you're having to focus, you may miss the fact that that person is getting pale and sweaty and getting ready to puke on you. >> Yeah, which is, you know, something you want to be prepared prepared for for multiple reasons, obviously. >> Yeah. We want them to be safe. We want everybody to be safe. >> I'm looking at this thing now and I'm like, they should they should be putting a streaming video screen on there for me. >> Hopefully in 2 and 1/2 minutes you wouldn't be able to watch much, so it'd be okay. >> Yeah, well, don't you know about micro dramas? Like perfect. This is perfect for that. I get my draw my blood drawn once a year cuz I have a thyroid issue and and they have to test my thyroid levels and I'll tell you that the the phlebotomist I like the best are the ones that just chat while they're going, you know, and give me full confidence that they're not messing messing up the labels or anything, but they're able to just how is your day, you know, did you fast? Oh, how bad was that? It's always the same conversation, but it just takes my mind off it. I'm not somebody who has a vaso vagal response, but I am someone who doesn't really like to look at the needle going into my arm, so anything that distracts me, I think is is a just a you know, a quality of of life thing. And this this you know, being able to give them more of that opportunity, I think is a good thing as well. >> Absolutely. >> Uh what about the patient reactions to this? So, I am someone who's going to look at this and go, great, as long as it works, I'm into it. And the data seems to bear that out. Uh some people may not look at it that way. They may be like, you're going to put me in that thing? You know, I ain't going to do it. Like, what what do you think about that side of it? >> So, they did an entire study in the United States. The original study was done, I believe, in Norway. And the they did an entire study in the United States looking at just people's reaction to the machine. They put them in the machine, they let the the machine put the tourniquet on, find the vein, but there was no needle inserted into the vein. And they were looking at, how does it feel to be, you know, at the mercy of this machine, essentially? And and how what was the reaction to that? The people that had the biggest reactions, so on the negative side and the and the the positive side, both, were very very small component. It was about 5% on either end. The people that didn't really care or had an okay reaction to it or were were not necessarily one way or the other convinced, made up the bulk of people, just as we would expect. And the people that that really surprised me, in a good way, the about 30% 30 to 32% I think was the original number of people said they didn't care. That is a good sign because that tells us you when we talk about medicine, the things that people don't care about, that it's so invisible that it might as well just happen without them knowing, that's the stuff that we want to have. That's where it this can make a huge impact because if people don't care one way or the other, we can we can actually add that force multiplier and it doesn't slow things down, it doesn't scare people away, it doesn't keep people from coming in to get these things done. And the people that want the machine, they have access to it. And the people that don't, the phlebotomist now has time to take care of those people because the machine is taking care of all the people that are are okay with using the machine. >> in that that middle ground there. Uh that's a great point uh to think about and and I think think something people may not have considered here. Uh obviously, as you mentioned, this is an example of using machine learning, using computer vision in a way that augments, not replaces, uh in a way that improves uh I I don't want to, you know, over generalize and say and therefore everything is great, uh but when you look at this, does it change your opinion one way or the other on on using these kinds of tools? >> Not in to a large degree because I have expected this kind of thing to be coming. We know that in radiology, we use similar tools for finding places for the physician to put their focus. And I think this is a similar idea. This is finding the vein. If it finds a vein and it doesn't feel like it can poke it, it tells the human, "I don't think I can do this. I need your help." >> Yeah. >> And so, this is not replacing a human being. It is it is taking the best parts of the human and the best parts of the the robot vision and the the mechanical parts of things that can stabilize things and are very accurate down to a, you know, micrometer. It's using those to assist that human to do their job better. I I have no doubt that there are going to be even better things coming along, that this is going to become more popular as time passes, that it is going to become more ubiquitous because it is so simple, >> Mhm. >> but it is also incredibly effective in the right ways. And so, from that standpoint, no, I'm not afraid. It does need to prove itself. And that I think is one of the biggest things right now is we need bigger trials and we need to have some understanding of if there are certain tubes of blood when we draw them they have to be frozen right away or they have to be processed in a certain way and these machines are not capable of doing that yet. They're very simplistic. So there's going to be a very small subset of people that are actually going to qualify for getting these blood draws done. >> Mhm. >> And so getting more information about what it is capable of across all of those categories as well as this is a big one original studies were done in Norway. Most of the people in Norway look like you and me. They're a little bit more pale than the average person and so people of color or people that have other pigmentation, there there could be differences. Now I one saving grace for this is that they're using ultrasound and it sees through everything. So from that standpoint it may not make as much of an of a deal a big deal. >> make sure. >> But we need to know. This we we've had ourselves we've gotten into problems before by not looking so we need to make sure that we're paying attention and that's true all body types, all sizes, all shapes, you know, what whatever differences and changes that can that can complicate drawing blood, we need to make sure that we're addressing that and so it right now it defaults to if I can't do it, I'm not going to do it. It's not even going to try and so that's exactly what it should be doing so that also gives me hope that they're thinking about this the right way. >> Yeah, I mean that that part right there is I think really important to underline. This is not Google Gemini AI mode. This is a very finely tuned specifically trained for a purpose version of this. It's not going to hallucinate an arm and and the tests are bearing that out. >> a nerve instead of an arm or it has fail safe to make sure it doesn't choose an artery instead of a vein. It is it is absolutely as and I hate to use this word but it is it is the word that comes to mind. It is as absolutely as idiot proof as a machine can be and still be safe. So probably the safest it can be is is a better way to say it but it is it is designed specifically to know its limitations. And I think that is when you talk about Jimena, when you talk about Claude, when you talk about any of these other systems, they don't think they have limitations. >> Right. >> Um and so, that is that is where those hallucinations can kind of get in the way. In something like this, it is designed specifically to say, I can't do it. So, just like a human, it understands it can understand that I don't just have to make something up. I can absolutely say, "No, let's figure this out and we'll do it together." >> Yeah, there there's no reward hacking, you know, for for coming up with with a vein. Uh I think wider trials to make sure, you know, to find out more specifically what that error rate is, you know, what the risks is. There's a risk when you have a human do a blood draw. So, we're not going to eliminate risk, but are we bringing it into the acceptable range? Uh and like you said, you know, making sure that this works on a lot of different types of people. Uh that that is very important. Any other questions that you think need to be answered uh that that people should know about? >> I I think there is some level of concern about the oversight that is needed for this machine in terms of can one person really manage three people at one time, you know, is splitting your your your consciousness that way? Is it feasible? Because they did it in a small-scale test and it worked pretty well, but when you get into a busy office or you get into a busy outpatient setting, is that going to pan out? So, I think that's the only thing that I was also asking about from a statistical standpoint, is that going to be enough? Uh and does it work when that phlebotomist that's doing it is well-trained versus partially trained versus, you know, who should be the one that's in charge of that? And that's a whole other set of things that we have to kind of down the road be be looking at. >> Yeah, I've as soon as you said that, I've thought of a couple of times I've gone to that lab core that's run by one person who's doing the front desk and all the blood draws, and the woman who's complaining that she, you know, she's been waiting too long and the guy who forgot his insurance card and now is upset cuz he doesn't know what to do and she has to deal with that as well as the the draw the blood suddenly supervising those three people is a lot more stressful than than than it is in this laboratory situation. >> But you can also just push the button and then go back to talking to that person in some cases. So there are trade-offs, right? So it is it is there are lots of Futurama episodes that could be written about this for sure in terms of what could go wrong and what could absolutely go right. >> Yeah. But but we should we should test that in the you know in in the real world now that we've got a little confidence up. >> Absolutely. Give one of those machines to your lady, you know, give two of those machines to your >> I'll write the lab coat. Can you please send this to Encino, California? Thank you. >> This poor overworked lady. >> Yeah. I told her she was a champion. I'm like you are a rock star that you've not stabbed anyone. Good good work. >> Especially with as much access she has as she has to sharp material. >> Or not I guess I should have said not stabbed anyone in a way that you weren't supposed to since she you know literally was stabbing. Dr. Gary Talbot, thank you so much for looking into this and and understanding it and and kind of translating it for us. I can't tell you how much I appreciate that. >> I'm glad to help anytime you need it. >> If folks want more of what you do these days other than visiting you in your office I suppose, where should they go? >> You can find me on most socials as either Dr. Talbot or G. Talbot. And Scott and I do Scott Johnson and I from FrogPants do a show called The Medical Show. We're gearing up for season two of that and that is over at the medical.show. >> Any any idea when season two might be coming? >> Nothing solid yet because he's a very busy man and so am I. >> Yeah yeah yeah. All right. Well, I'm looking forward to it. It's a great show. And if you haven't checked it out, go check out season one. It's not like there's there's an expiration date. >> Correct. And they're they're evergreen topics on purpose and and they're available anytime. >> Fantastic. Thanks again, Dr. Gary. >> Always. >> And thank you for supporting the show. dailytechnewsshow.com is where you can find us. patreon.com/dtns is where you can support us directly until Monday when we get back to the news. Have a great rest of your weekend. Thanks, everybody. >> The DTNS family of podcasts helping each other understand. >> Diamond Club hopes you have enjoyed this program.