This month’s episode of “News You Can Use” on HealthcareNOWRadio features news from the month of April 2026
News You Can Use with your Hosts Dr Craig Joseph and Dr Nick van Terheyden
The show that gives you a quick insight into the latest news, twists, turns, and debacles going on in healthcare with my friend and co-host Craig Joseph, MD (@CraigJoseph), Chief Medical Officer at Nordic Consulting Partners, and myself, where every diagnosis comes with a side of humor. We hope you stay curious, stay engaged, and keep seeking the truth in healthcare in a world that thrives on information.
Buckle up as we dive into the ER of excitement, the ICU of irrationality, and the waiting room of wacky wisdom in this month’s show that features a review of:
- Doctors are not calculators
- GLP1 resistance (no not that kind)
- Chat-GPT takes on Doximity and Open Evidence
- MCP (Model Context Protocol) Installed 96 Million Times
- Epic shows up on Freakeconomics
This month we discuss the absurdity of modern medicine, starting with the great AI turf war. Apparently, 40% of us physicians are already using “Open Evidence,” but ChatGPT is trying to crash the party by demanding everything up to your credit history just to prove you’re a real doctor. Nothing says cutting-edge healthcare like a digital velvet rope. And to be clear, none of these shiny AI toys are actually connected to our Electronic Health Records yet. So, we’re all just playing with calculators that can talk, while still manually typing into a system that makes a 1995 DOS computer look fast.
Speaking of calculators, can we stop pretending a single cholesterol reading means anything? Craig’s former cardiologist wrote a blog post titled “Doctors Are Not Calculators,” and we agree, because he’s right. We obsess over a random number going from 200 to 187 like we just cured death, when it’s often a calculated guess, not a real measurement. Meanwhile, 20% of people who drop dead from a heart attack look like Olympic athletes. It’s probably inflammation and/or genetics, but sure, let’s keep arguing about your ratio like it’s a horoscope. If you don’t even know your Lp(a) result, stop obsessing about your LDL. We’re dumbing down science so hard it hurts.
Finally, the gym bros have a new enemy: GLP-1 resistance. I literally overheard a guy at the gym whining that people should just “put in the work” instead of using weight-loss drugs. Enough of the moral superiority. As Craiug points out, your pancreas isn’t “lazy” when you get diabetes; it’s biology. And for the 10% of people who don’t respond to the GLP-1 drugs? It’s probably genetic, which means we need precision medicine, not judgment. We already know 95% of us react weirdly to common meds, yet we still carpet-bomb patients with antibiotics.
97 Million Downloads Later, AI Still Can’t Find the EHR Login
And finally, we catalog the future as detailed in the download statistics of “Model Context Protocol”, MCP’s for short. This is connecting your AI to your computer so it can send, amongst other things, emails for you. Yes, that’s a real thing, MCP, and it has been downloaded 97 million times. Just don’t give it full access until we figure out who’s really in charge. Spoiler: It’s probably not the calculator.
We’ve Connected AI to Everything Except Common Sense
We hope you enjoy our take on the latest news and developments in healthcare and want to help you keep untangling the web of information, dodging the sensational pitfalls, and emerging victorious, albeit a little dizzy, on the other side. In the end, the stories we uncover, and the discussions we ignite, all shape the narrative of our shared future. We want to hear from you especially if you have topics covered or questions you’d like answered. You can reach out directly via the contact form on my website, or send a message on LinkedIn to Craig or me.
Until next time keep solving healthcare’s mysteries before they become your emergencies and stay healthy, stay skeptical, and for heaven’s sake—check who’s listed as your emergency contact.
Listen live at 4:00 AM, 12:00 Noon or 8:00 PM ET, Monday through Friday for the next week at HealthcareNOW Radio. After that, you can listen on demand (See podcast information below.) Join the conversation on Twitter at #TheIncrementalist.
Listen along on HealthcareNowRadio or on SoundCloud
Raw Transcript
Nick: Welcome to the month of April. I’m Dr. Nick.
Craig: And I’m Dr. Craig.
Nick: This week we’ll be dissecting the latest healthcare news, unraveling twists and turns and making sense of the debacles that are unfolding.
Craig: Just remember life’s a lot like a breaking news story. Unpredictable, often absurd, and occasionally it leaves you wondering if it’s all just a cosmic prank.
Nick: This week we take a look at doctors as calculators and the resistance to GLP-1s.
Craig: We also dive into freakonomic epic interview, doctrine in Utah and MCP’s uptake.
Nick: But first off this week we thought we’d hit chat GPT who’s taking on open evidence and doximity for ownership of the physician interaction. What are we talking about here, Craig?
Craig: Well, there’s large language models or AIs that are general, and they’re just supposed to be smart with everything. And if you think about that, the first one that everyone knows about is chat GPT, or you can think about claude, or you can think about Gemini. And those are trained on like everything. And so what we’ve tried to do, what some companies have tried to do is create an LLM or an AI that’s specific for clinical use, really specifically for physicians, and train that large language model only on healthcare related information. And so.
Nick: Oh you mean like open evidence did with all the access to the generals?
Craig: That’s exactly what I mean.
Nick: Oh yeah that’s right didn’t they do this first?
Craig: Yeah. Open evidence did it did probably come out first. Soon thereafter came docs gpt doximities answer to the same kind of question.
Nick: Anna they connected to all the journals?
Craig: They claim to have many of the same access. Yes. So I won’t say definitively yes, but I’d say it’s certainly in the ballpark.
Nick: Okay.
Craig: And now a third group has come out, but this is one that everyone’s heard of, which is chat GPT from open AI. So chat GPT.
Nick: So of course all the physicians are giving up their access to open evidence and doximity and saying oh let me move to chat GPT and set up an account.
Craig: Well, that’s the question. Right?
Nick: Isn’t it?
Craig: It just so they just came out last week. And so the question is a will any of those three, and there’s others, but those are the big three right now. Will any of them be significant?
Nick: Going to say any of those three people using it but okay go on.
Craig: No. Over. So apparently 40% of US physicians have an account with open evidence.
Nick: Yeah. Yes they do.
Craig: All right. And that’s with like no real advertising to get them to do that. That’s word of mouth.
Nick: Quick quick quick survey n of two do you have one?
Craig: Of course I do.
Nick: And do you use it?
Craig: I do.
Nick: Yes so that’s a yes from both of us two thumbs up there right what about doximity?
Craig: I. Yeah. Yeah. I have, so I certainly have a doximity account. And again, doximity is this company that’s been around for many years started off as the LinkedIn or social networking for physicians and kind of has built tools on top of that. I’ve used doximity a couple times but have not found it to be superior to open evidence in my few use cases. And so I’m not sure it’s worse. I’m not sure it’s better. It seemed to be about the same. And since I was already kind of in the open evidence world, I kind of stayed there. And so that’s.
Nick: Yep. Exactly first mover advantage.
Craig: The question I think that’ll be interesting. Will doctors move?
Nick: Well I’ve got to say it’s there’s a bigger question in my mind because I tried to get access I thought oh well I might just go have a look oh no it’s only for US qualified physicians and we’re going to validate that you’re a US qualified physician with an MPI number which I don’t possess but open evidence was much more welcoming they said oh certainly we’ll take your UK or maybe they’d have taken my Australian certificate I didn’t try that one and I can’t get in so I’m already I’m just going to say bias declared.
Craig: Well, this is a different, this is does seem to be one small differentiator between. Again, there’s more than these three, but these are the three bigger ones right now is that chat GPT is apparently made the decision that they’re going to be much more selective about.
Nick: No take up. US centric is what I call it but okay.
Craig: Well, it’s not, you know, so for, so to get into doximity, I’m sorry to, well, yeah, I think doximity and or open evidence, you have to at least give them a provider number. Right.
Nick: You have to prove that you’re a physician but they’re more welcome.
Craig: But there are various ways. So, but apparently chat is a little bit more serious. So not only do I have to say, well, I’m a doctor and here’s when I graduated from medical school, and here’s my ID number. They want you to prove your identity. Right. And so I haven’t gone through this process, but I read it. So much in the same way, you know, asking you what your mortgage is or asking you questions that only you would know based on your credit history or, or something else. And so I don’t know why they’re doing that. They seem to be, they clearly.
Nick: To the CME that they are supposedly offering and you know maybe that’s a differentiator but you know quite honestly that’s just another idea to add anyway I.
Craig: Ah. Yes, that could be. It’s more exclusive, Dr. Nick. And so I want to be in the most exclusive club.
Nick: M. Yeah so I.
Craig: Anything you can’t get into. That’s where I want to be. Is that wrong?
Nick: It is I’m just going to say.
Craig: Oh.
Nick: I’m trying to think of the the quote from the comedian. That says I don’t want to be a member of a club that lets me in.
Craig: Yeah.
Nick: Groucho marks Groucho marks that’s it I don’t want to belong to a club that will accept me.
Craig: Fair enough. Fair enough.
Nick: I’ll twiddle my what was it I think it was a mustache with him it’s been so long since I’ve seen the marks brother.
Craig: So here’s what the, here’s the interesting thing. That these, these big three companies, well, none of them.
Nick: There’s more.
Craig: None of them. Right now today are connected meaningfully. To the big electronic health records.
Nick: Yeah. Yeah.
Craig: Right. So if I’m, if I want to query, if I want to ask one of these large language models or a question, I have to get out of my electronic health record pivot to somewhere else. Type in some stuff. It better not be patient specific information. It has to be generic enough that you couldn’t identify.
Nick: No covered entity in there.
Craig: There’s none. Right. And so, so what’s amazing is how popular these things have become despite the fact that they’re not connected to the workflow that we, we live in. Most of us every single day. So when or how that happens might have a significant impact on.
Nick: I will I will say I did get some feedback from a practicing physician who said why would I bother I’ve already got open evidence so anyway. Moving on and appropriately since we’re talking about access to the EMR let’s talk about freakonomics one of my favorite shows I’ve got to say I love Stephen dubner’s take on things I think he does a really good balanced job and he interviewed Judy Faulkner and I’ve got to say my first question was who approached who there was this part of the PR reach out to try and get them in front of other folks do you think what’s your thoughts?
Craig: The. Well. So first of all, I have no idea, but it is, I don’t know definitively.
Nick: Yous don’t know well hey we could stipulate that every time you open your mouth.
Craig: Let us. Wow. You’re upset. You’re upset about that comment I made a few minutes ago by. All right, that’s fine.
Nick: Oh yes.
Craig: I, I can take that. But it’s, I’m not saying you’re not wrong. I’m not saying that. You know, Stephen Dubner did say in the interview that he wanted, he decided he had never heard of judy falconer before. He was interviewing bob walkter. Bob walked her the chief of internal medicine at UCSF and has written a book writer and just wrote a book about AI and healthcare. And in that book and in that interview, when, when governor did that, he mentioned Judy Faulkner and epic. And since he had never heard of it. So he says like, that’s when I decided I reached out and wanted to talk to you. As you know, she hardly ever does these interviews.
Nick: That’s right. Yeah. And a book writer. Right.
Craig: And and I think, as you’ve also pointed out, it seems that epic is doing more of these kinds of reach outs to the outside world. For various and sundry reasons potentially. So yeah, I assume that they reached out to her. But I think that many, many, many people have reached out to her and generally got to know and they got a yes.
Nick: Interesting. Well I have not listened to it I know you have only recently I it’s on my list but I’m backed up with podcasts because I ended up listening to the Hail Mary book which I’d already read but I’d heard that the actual audio book was good and it was to be clear but that was long book so I’m very behind so.
Craig: You’re behind now. Yeah. It’s a, it’s a good interview. It’s about, I’d say it’s about 50 minutes or so. And they’ve also interview one of the lead developers spent a few minutes talking.
Nick: Oh it’s not just it’s interesting.
Craig: It’s, it’s, it’s mostly Judy, but it’s not only Judy. Yeah.
Nick: Right.
Craig: And she, yeah, she throws a few funny quips in there. And talks about how she got in, how she started the company and why she thinks the company is different. And Stephen Dumner did, did note that he talked, he has a lot of friends who are doctors and they’re not crazy. They’re not crazy about epic. And she asked the questions she often asks, which was, well, have they used other electronic health records? And he didn’t think they had. And she said, well, many people who have used other health records decide that they’d rather be with epic, which I think is true, which is again, not to say that it doesn’t, it’s not bad in certain areas. But is it better than its competitors? The answer is usually yes.
Nick: Yeah. No no. Right. Well it’s on my list maybe that’s something we end up talking about or maybe I just listen to it like everybody else but certainly it’s fun to have that centered in moving in moving on not moving in doctors are not calculators this was all about cholesterol calculations and changing values I’ve got to say when I looked at that my first reaction was exactly what I have every time somebody tells me and they give me some lab value or even wait let’s be clear if you just oh my God I’m X kilos sorry pounds. And pounds in this country no stone or pounds here and I’m my reaction is the same every time it’s a data point on its own it’s almost mean I I don’t think it’s completely meaningless but it has very little value and this was more about this sort of you know the reaction to these levels that you get which are a single point in time much like our blood pressure when we go to the physician’s office.
Craig: No stones. Yeah, well, this was a, this is a blog written by, and I have to say this is the only time I’ll ever say these words. Written by my former cardiologist, Jim Stein from the University of Wisconsin. And, you know, he was just my doctor. And then all of a sudden I stumbled maybe six months or nine months ago that on to he’s writing a blog post and he’s, he’s pretty, he’s, you know, pretty frequent blogger and he doesn’t, he’s not doing superficial stuff like his essays are well thought out and deep. Yeah. And this one, like you mentioned, it’s, you know, talking about, I think it’s titled doctors are not calculators. And basically makes the point that you just did with some specifics really talking about cholesterol. But it could be any test that, hey, you know, most doctors are not very familiar with how these tests are run or what they’re, you know, so you think, well, hemoglobin’s a hemoglobin and it’s, it’s going to be the same. And if I take your, your blood and I run the same test at different labs over the, over the course of one day, I’ll get the same results exactly the same results and you won’t. And, but they’ll be close enough. And so yeah, his major point was exactly what you said, which is, well, first of all, one data point by itself doesn’t, doesn’t mean anything. You need the context. And so we, you know, it’s something that I think a lot of us who are doing this day in and day out just tend to forget because we love to try to be scientific. And, and it’s easy to say, well your cholesterol was 200 and now it’s 187. So clearly we are doing something amazing because it’s gone down 13 points in, in one year. And well, it’s not gone up, that’s for sure. And it’s not gone down by half, but it’s going in the right direction. So it’s that it’s a point in time and it’s helpful, but it doesn’t, it doesn’t mean that you need to, you should change your theory about what’s going on with the patient. If they’re 190 versus 187.
Nick: All right and and hobby horse engaged for a second just let me jump on this you know this whole cholesterol thing and these ridiculous values that we continue to push out one of which is calculated so it’s not even an actual measure and we have only just started to even talk about Lp or apo b which is on the spectrum of these lipid molecules and far more significant and we’re focusing on these broad categories that give you a ratio and now oh my god my ratios it just it really bothers me that we’re dumbing this down so much when we have so much more value and understanding. Do you know what your Lp result is?
Craig: I, I know it’s, I’m one of those lucky genetic folks who has a very, has a high LPA. And by lucky.
Nick: Sorry to hear I genuinely am but the good news is.
Craig: By lucky, I mean unlucky. Yeah, yeah. So as you know, probably most people don’t know, but it’s, you know, LPA is not is a, it’s a genetic, it comes from your genes, right? So yeah, you can be, you can eat as healthy as you want and exercise and do all the things that your cardiologist wants you to do. You’re still at higher risk of having a heart attack.
Nick: 20% 20%
Craig: Because you have that LPA.
Nick: Right and and to be clear we haven’t got time to talk about this but the the recent article that was posted by Eric topple on this you know starts out with the data point that I think most people. Don’t even know and that is. At least 20 25% of people that die of a heart attack look like they’re in great shape cardiovascularly and we don’t fully understand why but there is now some evidence to suggest that it is about inflammatory processes and that’s why they’re using the old gout drug now as a potential although results are mixed not advocating that and none of this is medical advice blah blah blah but I will link to it in the post that goes out worthy of reading there’s so much more that we can offer but yes physicians are not calculators.
Craig: We agree.
Nick: For once all right moving on we have GLP one resistance is is this a movement?
Craig: Are coming out against GLP once.
Nick: I think that’s probably true.
Craig: I don’t know if that’s happening.
Nick: You know I hear it occasionally I do I have honest to God I have heard this in the gym I think X is using a glp one I wish people would you know put the work honest to God that’s I I heard that in the gym.
Craig: Yeah, that’s, that, that’s similar to like, you know, that guy’s pancreas is really just too wussy.
Nick: Yeah.
Craig: You know, don’t give me that diabetes crap. Like just work harder on your pancreas.
Nick: Yeah come on. I gotta say I was a bit shocked I didn’t react I was a good boy.
Craig: Well. This is, that’s good. Actually, that’s good for you.
Nick: For me it is yes.
Craig: And obviously you’ve been working on that and I see you’re going in the right direction. So that’s, I’m happy for you. This GLP resistance is dissimilar, dissimilar to what you were just describing. Some, some research has come out that says that about 10% of of people are not losing the weight. That they expect to lose. And they think it’s due to genetic variants. And so just that they have some sort of gene that are a little bit different in terms of the ability for the drugs to help you regulate your, your blood sugar. And again, what I think this is moving us towards is just another aspect of kind of precision medicine where we, and what we mean by precision medicine is not treating everyone the same way. You have you, oh, you have a bladder infection. Well, then you need this antibiotic.
Nick: Right. Yes let’s carpet bomb the hell out of you.
Craig: Right? Yeah. So it might be.
Nick: Oh my God did that upset your tummy oh dear.
Craig: Right. So we’ve never had the ability to differentiate, especially, you know, beforehand.
Nick: We’ve never had we used not to have I think is probably more accurate now.
Craig: In the past, we did not have the ability to, you know, cheaply and quickly.
Nick: Right correct. We do now.
Craig: We do now for some for some diseases and some treatments. Right. So we do especially if you think about oncology, we see it a lot where, you know,
Nick: I think we’ve got some great success there right.
Craig: Absolutely where, you know, hey, instead of saying you have this tumor, this is the chemo. It might be based on the gene, your, your genetic makeup or the makeup of the tumor itself that, oh wow, actually that is not going to be as effective as this other. Or the side effects for this medication are going to be far greater for you than for the average person. And if we can predict that ahead of time, then it’s, it’s good for everyone. And so this might be, as you’re pointing out, as it gets cheaper for us to be able to make some of these associations and predictions and to screen people, then we will be able to do a lot of things in a better way. And so this might be one of the things coming in the future is, hey, before you get your GLP one, we are going to do a blood test to see if you’ll react the way we expect you to.
Nick: And to be clear I think that’s the case or it should be the case we’re just not doing it as effectively as we ought to I mean to be of the 20 commonest prescribed drugs. Well over 95% of people are resistant. Or super sensitive to at least one and typically more and I can tell you in my particular profile I’ve got three of them where I’m a fast metabolizer and, you know, but we don’t test unfortunately so yep it’s it’s going to be more precision well it needs to be more precision medicine all right so moving on we mentioned MCP. Would you care to explain and what the metric was that we saw in in the news?
Craig: Yeah. So MCP stands for model context protocol. And it’s basically just the connection. It’s a system by which it was developed by. Why am I losing my words? It was developed by the claud.
Nick: Anthrop.
Craig: Company, anthropic. Thank you so much. As a way to say like, hey, we’ve got this AI and it can, we know it can generate text and it can make pictures. That’s awesome. But can it control your computer? Can it talk to other applications? So if you wanted to write and send an email, can it get into your email system instead of just creating the text and then you having to copy it and then you having to open up your email and paste it in? Can it just do that? And that’s what MCP allows it to do.
Nick: Olog. Ist could I just say. Come with me if you want to live.
Craig: You can say that.
Nick: I’m sorry this feels real sky nutty.
Craig: But now. But now we know why you can’t get into some of those, you know, open evidence and other things, right? Because you say things like that and people don’t want you poking around their AIs. I, you know, it’s, you know, certainly there was something, there is something called open claw where, so the concept of an MCP is a connector, but with some with some levels of safety precautions. Right. So for instance, we were talking about connecting your AI to your mail program. You can say things like, oh, well, I’m going to, I’m going to configure this to allow you to read my email.
Nick: Yeah.
Craig: But you may not send email if I don’t want to do that.
Nick: And and fair do is you can control it I think the big problem with open claw if I recall was that it was sort of a fully automated here give me access oh my god I’ve got everything and people with no programming experience were getting into a little bit of trouble let’s just say but I want you to share the stat how many times this was this this connector or whatever was downloaded I’m not sure.
Craig: Correct. Yep. Yep. And it can do anything. Yep. Yes. $97 million.
Nick: 97 billion times.
Craig: $97 million times this, yes, the MCP is used somewhere. And so.
Nick: Wow. And that’s just one of the providers right the others all have their connective versions.
Craig: Pretty much this is kind of becoming this. This is now because it was first, we talk about that first advantage. It came out first and. And it seems to work. And so, yeah, pretty much everyone’s doing it now and following the mcp concept.
Nick: This is across the standard okay.
Craig: So yeah, it’s pretty exciting. So.
Nick: Are you connected?
Craig: Oh, oh, for sure. For sure.
Nick: What about the chat GPT.
Craig: But. No, so I’m using, I’m using claude. But yeah,
Nick: Oh oh oh oh sorry okay.
Craig: I play with it. I play with it. That doesn’t, it can’t write emails for me or do lots of other things. But I am exploring my world, this great new world that you fear.
Nick: Last time I heard that was when somebody oh dude I’m spurring my world man.
Craig: I don’t even know what you’re referencing at this point. This, this, this episode’s gone off the rails.
Nick: Oh completely.
Craig: Dr. Nick. This is why we’re so popular.
Nick: I always remember being introduced by an old friend of mine and you know I was explaining that the last time I’d met him I was actually high on drugs because I had a really bad back injury and I was, you know, given some opioids and muscle relaxants and that was his response he said oh my god it was back in the 60s the last time I was introduced this way.
Craig: That way. That’s hilarious.
Nick: Oh well anyway all right so exciting times connect everything and suddenly just make sure you don’t turn it all on is what I would say.
Craig: I think that’s right.
Nick: But as we always do we find ourselves at the end of another episode exploring healthcare’s mysteries before they became your emergencies until next time I’m Dr. Nick.
Craig: And I’m Dr. Craig.
