Science, Politics, and Personality

Harlan Krumholz: Welcome to Health & Veritas. I’m Harlan Krumholz. ‌

Howard Forman: And I’m Howie Forman. We’re physicians and professors at Yale University. We’re trying to get closer to the truth about health and healthcare. Ordinarily we have a guest, but at least once a month we’re trying to do a deeper dive into a whole host of topics. ‌

Harlan Krumholz: I just get you by myself. I’m so happy.!‌

Howard Forman: I know. And when we’re actually physically in the studio, it makes a huge difference.‌

Harlan Krumholz: Physically In the studio, I get to see your beautiful face. ‌

Howard Forman: Exactly. ‌

Harlan Krumholz: That’s wonderful.‌

Howard Forman: And we’re both coordinated in our attire today in our polo shirts because it’s 95 degrees outside. ‌

Harlan Krumholz: It is hot. ‌

Howard Forman: And let me just say this for our listeners. You and I are both practically zealots about walking. Like we like to walk everywhere. And so there’s really not an option to come in dressed professionally if you’re going to walk in between buildings on Yale’s campus on a 95-degree day. So I do think this is professional attire if you’re going to be a walker.‌

Harlan Krumholz: Isn’t there something, if you’ve been faculty for a certain number of years then you no longer have to dress professionally?‌

Howard Forman: I hope so. ‌

Harlan Krumholz: Yeah, I hope so too.‌

Howard Forman: But I want to start off the podcast with a follow-up to last week we talked about the Duchenne muscular dystrophy medication. ‌

Harlan Krumholz: Yeah.‌

Howard Forman: I mean, it’s crazy to me.‌

Harlan Krumholz: Like usual, Howie, you’re a week ahead of the—‌

Howard Forman: No, but it’s crazy to me that we have huge follow-up. It’s like such a fast news cycle that we even have a lot more. So a reminder, there were three deaths in two adolescents and one much older adult from an FDA-approved treatment. ‌

Harlan Krumholz: Do we know how many people total have ever been treated with this? I mean, are those three out of how many?‌

Howard Forman: Oh, I think I looked this up last time. I think it was something like 900 people or something.‌

Harlan Krumholz: A large number of people have been, to me that’s a fairly robust number‌

Howard Forman: Or maybe much more than that. But it’s in that range at least. And the drug had received approval for... total approval for the 4-to-8-year-old ambulatory class. And then had received sort of an expedited approval in the more adult cases, but never full approval for those cases. And so you had had these losses in the older cases, you had not had consequences in younger cases. And all of a sudden they take the drug off the market. The FDA recommends moving it off the market and Sarepta goes ahead and then there’s even leaked anonymous words from the FDA saying, “It’s going to be really hard to get it back on the market.” And as you can imagine, the Duchenne muscular dystrophy community—parents, mostly—were very, very upset with this. These people sign up in advance to be able to get it. It’s scheduled months and months and months in advance. You’re waiting for your child to get treated for this disease that you know is going to devastate them. And then it’s pulled out from under them at that point. And so the thing that happens is after our podcast, just like a day after our podcast—‌

Harlan Krumholz: As soon as everyone heard it from us.‌

Howard Forman: Exactly. No, no. But after our podcast, the FDA does announce that there was an 8-year-old who had received this therapy and died. This is on Friday after our podcast. And that’s out of the blue. Nobody even talked about it. And maybe that was starting to explain why this whole thing happened and then three days later Monday they announced the kid did not die related to the treatment at all. The FDA announces this. So that would’ve already been enough big news to come out that you’re—‌

Harlan Krumholz: By the way, I just think that’s weird. Usually it’s like no matter what you die of, if it’s in proximity to the time you’ve received one of these kinds of therapies, people will attribute it to that.‌

Howard Forman: Absolutely.‌

Harlan Krumholz: So I’m really curious, did they, the person get hit by a car? What happened?‌

Howard Forman: I’ve read the release, it says nothing. It doesn’t say whether it was non-live because that’s the big complication. It doesn’t say anything. It doesn’t say whether he got treated three years earlier. It doesn’t say anything, but at least they say not related to the treatment. What made this even more curious is right after they have that announcement, they also announced that the FDA is reversing course, and now they’re going to allow deliveries of the drug for these children. So that was already a big thing. It was already really odd, the sequence of events. And then to put a crazy little bow on the whole circus, the FDA fired Vinay Prasad, and you and I briefly talked about this offline, he was functioning not only as the head of this agency that regulates this drug and many other drugs. He was also the FDA’s Chief Scientific Officer, Chief Medical Officer. To me, one of the things, the take-home messages, the whole thing is that when you start losing faith in the institutions and the people at those institutions, it makes us think all of this is one big conspiracy. So why don’t you take—‌

Harlan Krumholz: Well, I don’t want to preempt it. I think there are two major, at least two major stories in here. One is about how the FDA is operating, the degree to which they’re being transparent, how they’re making decisions, how those decisions are based on science and what do we know about them. And here’s a situation where they turned left and they turned right and there was a lot of external pressure on them and it was very confusing. I’m sure for those parents it was anxiety-provoking. ‌

Howard Forman: Devastating. ‌

Harlan Krumholz: I can understand the case actually on some sites, but I don’t know all the data like that, we still haven’t been, it hasn’t been shared exactly what’s transpired. So it’s hard. I think as you’ve suggested and you talked about last week, that’s an enormous story. And the end result is the person who’s kind of in charge of that decision gets let go.‌

Howard Forman: Right.‌

Harlan Krumholz: There’s another one, which is... really, it’s an Icarus story. Someone who flies too close to the sun. It’s a story of this remarkable chapter of Vinay Prasad at the FDA coming to a close. This is an outspoken critic. He became a smashmouth kind of persona in medicine going after people. I told you: nasty, smart, brilliant. ‌

Howard Forman: Same words I used. Yep.‌

Harlan Krumholz: Incredibly right. Then in recent months, aligns himself very closely to RFK Jr. Was starting to write really, I think almost obsequious pieces about, I was half-expecting him to say, “...and he’s a very handsome man.” ‌

Howard Forman: Right, exactly. ‌

Harlan Krumholz: Because he was clearly positioning himself for an important position. He ends up getting to head the Center for Biologics Evaluation Research. Peter Marks, who a very distinguished regulator for many years was forced out. Vinay Prasad, never having really any sort of record in regulation, except to be a gadfly…‌

Howard Forman: And never managed any organization before.‌

Harlan Krumholz: ...started to come in. And his departure comes three months after his appointment. So then we say, how many Scaramuccis has Prasad made? He’s made maybe 10.‌

Howard Forman: Not even, not even.‌

Harlan Krumholz: You know, Scaramucci! He’s the famous person who was appointed by Trump as the press secretary for what a matter of weeks before it turned sour. Howie, tell me what this means. This is what was told. He has a desire “not to be a distraction” and he wants to “spend more time with his family.” ‌

Howard Forman: And he also did not make his own statement at all at this point. ‌

Harlan Krumholz: And he hasn’t said anything official. So this is one of those stories where science, politics, and personality all collide. And I think it’s about him taking a hard line on certain evidence. You know, he had redefined the COVID vaccine. So I think talking about whether kids can get this, I mean, I just think it’s—‌

Howard Forman: And he had criticized Elevidys or however it’s pronounced. ‌

Harlan Krumholz: Before...‌

Howard Forman: Before he even joined the FDA, he had criticized the decision by Peter Marks earlier, which by the way, you could criticize it. We talked about that. It’s not to say you couldn’t criticize it, but it’s not like they didn’t already have him on record being against this.‌

Harlan Krumholz: So he had come out full scale, MAHA MAGA, and then Laura Loomer comes after him.‌

Howard Forman: Exactly.‌

Harlan Krumholz: And Rick Santorum comes after him.‌

Howard Forman: And let’s just make sure. So Rick Santorum is at least a former senator who, I don’t know why this is an important topic for him, but this became an important topic for him. And you don’t have to like Rick Santorum, but he at least has some standing in the world. Laura Loomer is a conspiracy theorist. I mean that’s the best way to describe her. ‌

Harlan Krumholz: Influential. Very influential. ‌

Howard Forman: Very influential, particularly with the president.‌

Harlan Krumholz: So this campaign starts, the Wall Street Journal has an editorial in which they called him “a one-man death panel” and “a Bernie Sanders acolyte in MAHA drag.”‌

Howard Forman: I love that line.‌

Harlan Krumholz: And I just want to say it’s just hard to describe or overstate how surreal all this is. Here’s someone who aligned himself with the administration’s skepticism of the pandemic era of public health, criticized child masking booster shots for young people, and then he got painted as—‌

Howard Forman: It wasn’t enough.‌

Harlan Krumholz: Too authoritarian, too evidence-focused. Even though, and then now left-leaning and then next thing you know, gone. Oof.‌

Howard Forman: No, and I think one important point for people to understand, though, is that in the midst of all of this, are human beings being treated with drugs that almost require approval in certain situations if there’s even the most basic evidence that can support that. And here you have courageous families with young children, mostly boys, essentially all boys who have been desperately waiting for this.‌

Harlan Krumholz: So there’s that. And I totally agree with you, but I’m also believe this brings in bright relief. Now the FDA is more political than ever. Absolutely. Whatever you think of Prasad, here was someone who in a week, because of the backlash from the very strong MAGA people, the campaign against him is out. So what’s the message to Makary, this commissioner of the FDA? I mean, you better—‌

Howard Forman: Bend the knee.‌

Harlan Krumholz: ...toe the line, bend the knee. But the question is, knee to what? ‌

Howard Forman: Right. In this case, this was not about MAGA or MAHA. This was about the fact that the more typical Republican line on drug therapies is “evaluate safety and efficacy.” If it crosses that bar, you approve it and you let doctors and their patients make decisions. And that is different than other people’s views on it. There is more of a heavy regulatory hand view. And there’s a view that says, and we can go back, you’re an expert on this, on Vioxx. The view on Vioxx is a nuanced view. And I think you agree with that. The problem with Vioxx was that Merck withheld data that could have informed people’s decision making. The decision about Vioxx was not a simple decision of “dangerous drug never should have been on the market.” You could have made the argument, I don’t know whether you do or not, you could have made the argument that “if we tell people that there’s a significant risk of increased cardiovascular bad outcomes with Vioxx, but it works well in this situation. And if you’re a doctor and you decide to prescribe it, you can go ahead and do that.” You could have made that argument. There would be other people who said, “We should never approve a drug like that.” And there is this range, and there are a lot of drugs like that.‌

Harlan Krumholz: No, and that reflected my opinion. My opinion wasn’t whether it should be taken off the market. I was involved in the litigation on behalf of plaintiffs, but my position always was about transparency.‌

Howard Forman: That’s right.‌

Harlan Krumholz: People should be able to make informed choices. No data should be hidden. No data should be manipulated. It should be out there in the way it should.‌

Howard Forman: No, and I a hundred percent agree with you on that. And for our listeners, you are a very influential, important arbiter of this because of your expertise in that. I mean, I fully support that, but I think a lot of people forget that Vioxx was not really a story about efficacy and safety per se. ‌

Harlan Krumholz: And if anyone’s forgotten what Vioxx was, it was a nonsteroidal COX-2 inhibitor that was meant to provide pain relief that was argued to be above and beyond what you get from a Motrin and ibuprofen and things like that. There was some controversy about that, too.‌

Howard Forman: That’s right.‌

Harlan Krumholz: But anyway, lots in that FDA story, right? The Sarepta stuff, what’s happened to Prasad, the politicization? What does it mean actually if they can mount a campaign within a week, and he is head of CBER, so all the biologics….‌

Howard Forman: It was his.‌

Harlan Krumholz: …and the chief medical officer—‌

Howard Forman: And the chief scientific officer!‌

Harlan Krumholz: And the favored individual in the entire FDA of the commissioner—and he’s gone.‌

Howard Forman: He’ll have lots of stories to tell about this; I’m sure he’ll weave it in his own way, but I think he’s learned at least one lesson that as they say, healthcare is not easy.‌

Harlan Krumholz: When I say what happens when the dogs catch the car, it’s sometimes a little rockier than you expected. ‌

Howard Forman: That was a good line. Yes. ‌

Harlan Krumholz: All right, what else? What else you got, man?‌

Howard Forman: So I want to go to something a little lighter. The CDC had two good stories this week. I’m going hopefully cover both of them, but one of them was a light one, and that is the story on marijuana in canola oil. I’ll explain more.‌

Harlan Krumholz: Marijuana in canola oil.‌

Howard Forman: Yes, yes. There was a case of mass THC—that’s the active ingredient in marijuana— intoxication was reported from Famous Yeti’s in a suburb of Madison, Wisconsin. Eighty-five people aged all the way up to 91 years of age became intoxicated after eating at famous Yetis. Famous Yetis is famous for their garlic bread, for their pizza, for their cheese bread and their grinders. But now they’re also famous for an unexpected but thankfully reversible adverse medical event. Quick-thinking physicians and the CDC solve this puzzle when numerous individuals showed up at emergency rooms after eating a meal at Famous Yeti’s with symptoms resembling marijuana or THC use. So how did this happen? Famous Yeti’s used a communal kitchen and borrowed some canola oil—and the CDC actually literally in their thing had a picture of the canola oil—without realizing that it had been infused with THC on the way to making baked marijuana products. ‌

Harlan Krumholz: Oh boy. ‌

Howard Forman: In a legal dispensary, 28 people were seen in an emergency room; three were admitted overnight, a few others sought outpatient—‌

Harlan Krumholz: What were their symptoms?‌

Howard Forman: “Dizziness, sleepiness, and anxiety” is what was written into the report. ‌

Harlan Krumholz: Wow. ‌

Howard Forman: Yeah. And not even—‌

Harlan Krumholz: And munchies, maybe. Munchies.‌

Howard Forman: I guess so. Although they just had Famous Yeti’s.‌

Harlan Krumholz: Yeah, right.‌

Howard Forman: So the lesson here, though, is to know where you are sourcing your ingredients obviously, and I only make light of this because the outcomes were really mostly fine.‌

Harlan Krumholz: Those are some smart epidemiologists!‌

Howard Forman: I know! And the ER doctors, apparently, they’re the ones, they first went in the homes and checked carbon monoxide levels because it simulated that.‌

Harlan Krumholz: But there were so many people.‌

Howard Forman: And then they started to connect the dots. And I just want to point out, this could have been much more serious. Remember, these people literally ate a meal and basically got intoxicated and a lot of them got in their cars and drove home. I mean, it could have been a problem, but luckily there was none.‌

Harlan Krumholz: Wow. All right, well you’re on the light side. Let me go back to, well, yeah, kind of heavy. So I don’t if you’ve been reading about the NIH and what’s going on with the pay line.‌

Howard Forman: I need you to explain it more to me. I did read this morning about this.‌

Harlan Krumholz: So there’s a funding storm brewing at the NIH, and this has been going on for a long time, but it’s starting to really hit hard. The NIH is slashing the number of research grants at funds. At some of the institutes pay lines are dropping now to 4%, four of a hundred grants that come in are getting funded—one out of every 25. And these are historically low rates. There were low rates for a long time; we’ve kind of drifted up as funding to the NIH got better. But this is something, I think, and a new policy requiring NIH to fund multi-year grants upfront rather than spending the payments across the four or five years. So that’s depleting the number of actual new grants they can give out because they’re not spreading it along. They’re under some pressure to spend now because they’ve been spending slowly, but now instead of funding more grants and releasing them, they’re funding some of these grants, the whole thing. ‌

And so get it again, you put in a five-year grant, you used to get the money one year at a time. That meant you were funding a lot more grants in any given year with the expectation you could fund next year.‌

Howard Forman: Right.‌

Harlan Krumholz: If you start saying, I’m going to write the whole check this year—‌

Howard Forman: Right.‌

Harlan Krumholz: ...then you’re funding fewer grants this year with uncertainty about what the future is going to be. Lots of people are talking about this feeling fairly catastrophic for the research community and highly demoralizing. And look, I’m someone who applies for a nice grant, so I have some conflict here, but I do believe that the medical funding structure we have, as much as it can be improved, has produced the best science in the world. And it’s been an engine of innovation. And now with the way it’s working now, it’s getting a little crazy. And I’ll tell you one other thing happened this week. So in the last cycle, I guess at least one person put in 20 grants. And the thought was, this was AI-assisted. By the way, I think AI should help you do grants. I mean, AI can probably help you do better grants.‌

Howard Forman: Sure, sure.‌

Harlan Krumholz: But the NIH reacted to this by saying, “We’re going to start restricting the number of grants any one person can put in as a PI,” as the principal investigator. So that’s so wild. It’s sort of like saying, “Well, our system is going to get overwhelmed.” And instead of fixing the system, saying like, “Fine, bring it on, we’ll screen them. We’ll identify the good ones.” ‌

Howard Forman: Right.‌

Harlan Krumholz: “We don’t care,” it’s like, “Okay, we’re going to just start restricting the number of grants that people can put in.” Is that going to net be positive? They say it’s at six right now for a year, but how’s that going to change? What’s that going to affect? It’s really kind of crazy. You know, Congress hasn’t intervened yet and they’re going to be debating now the appropriations, but while we’re gone, well, I guess they’re on recess now. So when they get back in September....‌

Howard Forman: Apparently this is going to lead to billions in dollars of appropriated but unspent dollars that Congress could repurpose next year. And there is some real concern, I think, that this whole tactic is to just save money and to completely delay ever paying for things that we anticipated and Congress appropriated.‌

Harlan Krumholz: So when you put this together with the question of whether overhead’s going to go down—‌

Howard Forman: Right.‌

Harlan Krumholz: ...the potential 40% cut that’s being discussed by the administration for next year’s NIH budget, then you’ve got our scientific community in a high level of uncertainty about what’s going to happen.‌

Howard Forman: And any economist would tell you that if you’re a student right now, going to either college or even graduate school, you’re changing the calculus around whether you go into academic research or not. I mean, there are a lot of places where smart people can go if you make this a less productive place to be. We’re not talking about getting rich, we’re just talking about a productive place where you can guarantee getting a job. You’re going to go into something else.‌

Harlan Krumholz: There’s lots of job freezes, hiring freezes. Yeah.‌

Howard Forman: So we’re going to come up with the other CDC story. It’s a quick one, but it was a follow-up to something we’ve talked about a couple of years. Once again, we’re learning about the perils of raw milk. We’ve just witnessed one of the largest outbreaks of raw milk–associated disease when 171 individuals—and the median age was seven, so it’s majority-children—were shown to have developed salmonella typhimurium, which I had never heard of that.‌

Harlan Krumholz: I’m proud that you could pronounce it.‌

Howard Forman: I’m not sure I’m pronouncing it correctly, but it is “typhimurium.”‌

Harlan Krumholz: Well, I want to salute the effort.‌

Howard Forman: Thank you. Thank you. After presumably consuming raw milk or raw milk products such as cheese, this appeared to arise from one farm and led to 22 hospitalizations. No deaths were reported. So another sort of good outcome in the end. It all comes down to one cow apparently, as I’ll point out. ‌

Harlan Krumholz: And how was it transmitted? ‌

Howard Forman: It was, so the cow to the raw milk. The raw milk to either raw milk itself or to—‌

Harlan Krumholz: You’ve been warning us about raw milk for a long time.‌

Howard Forman: For a long time. But just to put this in context, there were 143 cases of raw milk enteric disease reported from 2009 to 2021 in California. This is 171 from one outbreak.‌

Harlan Krumholz: Wow.‌

Howard Forman: So this is just a whole different level right now. By the way, 171 cases we only know about because they reported them. Got to be many, many more people that had it and had sort of subclinical diagnosis. And as I said, they found the cow, they isolated the cow, the farm’s back to being online producing raw milk. But I just want to remind people like raw milk, which is unpasteurized milk, has risks. There are people who believe there are benefits. There is no significant evidence base to support the notion of real benefit from the raw milk. And here we have a strong example of risks. So I would just tell people, be very, very cautious about raw milk products.‌

Harlan Krumholz: All right. So I can’t leave our last session of this term—we’re going to be on break after this—so of course I got to get back to some AI stuff. ‌

Howard Forman: Yeah, no, no. Okay.‌

Harlan Krumholz: So ambient AI in healthcare, I’ve talked about this before, it’s a sort of auto scribe. They’re listening to the doctors’ and the patients’ conversations and then turning these into notes. Ultimately they’ll be doing decision support. There’s another company that I haven’t mentioned before, Ambience Healthcare, just raised $243 million in Series C funding, pushing its valuation to, wait for it, $1.25 billion. Howie, you know, these are companies that were put together just in a very short period of time. ‌

Howard Forman: I know.‌

Harlan Krumholz: There’s all about algorithms. I’m not sure what the mode is. There’s a bunch of them that are out there. Of course at Yale we’re using Abridge, and it’s a huge round. By the way, I always wonder, 243, how’d they settle on 243? But anyway, probably because of the way people are throwing money in, but this is one of the biggest of the years. Backers include Oak, which is from Connecticut,‌

Howard Forman: Right, Ned Lamont’s—‌

Harlan Krumholz: Annie Lamont’s company, but also a16z Open AI startup fund. And this one is best known for its AI documentation tools, and it’s spreading quite a lot. They’re in more than 40 health systems, including the Cleveland Clinic, UCSF, Houston Methodist, and… Cleveland Clinic just signed a five-year deal with them. And it’s just going to be very interesting to see what happens that you’ve now got something—someone—lurking in the room—‌

Howard Forman: Always.‌

Harlan Krumholz: ...with the doctor and patient.‌

Howard Forman: And by the way, we haven’t talked about it, and I have not tested it at all, but Doximity, which is the social media–based company for physicians that is publicly traded, is now offering a free version of their own AI.‌

Harlan Krumholz: That’s so interesting.‌

Howard Forman: As long as you’re not, as part of a health system, you want to use it in your private office, it’s free for you. I’d be very curious to know how that works.‌

Harlan Krumholz: I wonder about the business model on that. You remember that in the pandemic, they did that for telemedicine, they immediately stood up a—‌

Howard Forman: Yes.‌

Harlan Krumholz: ...telemedicine program. So that was kind of interesting, too. And it’s going to give them access to a lot of data. And the question would be like, what are they going to do with that data and how’s that all going to work? It’s going to be a big question.‌

Howard Forman: I think that’s exactly what they’re aiming for. It’s sort of a Tesla. If you listen to Elon Musk, he’ll always tell you it’s not about the cars, it’s about the data.‌

Harlan Krumholz: The data.‌

Howard Forman: And maybe that is really what the mode is, ultimately.‌

Harlan Krumholz: Well, I wanted to ask you about one more before we get onto another topic, which is, so this company Aidoc, have you seen that? ‌

Howard Forman: That’s what we use.‌

Harlan Krumholz: Well, that’s what I’m going to ask you about because it’s also gaining a lot of momentum, doing rays, getting popularities, moving really quickly. And I read Yale New Haven, and it’s primarily for radiology right now, but they’re going to spread out. So what’s your experience?‌

Howard Forman: I’ve consistently had a good experience with them. ‌

Harlan Krumholz: What happens? How does it work? ‌

Howard Forman: So we have, I don’t know if we have 10 widgets, but we have widgets that will alert us when there is a pulmonary embolus in a pulmonary embolus CT. It’ll also alert—‌

Harlan Krumholz: No, no, but does that mean that the AI—you’ve read it, now you’ve got all these scans in the repository at Yale, and this thing is crawling through the Yale database and then sending you a note to say, “Hey, we just discovered something over here.” Is that what’s happening?‌

Howard Forman: Real time. Real time.‌

Harlan Krumholz: So you’re in the ED. Somebody is coming in with a suspected blood clot in their lungs; they’re getting a scan. You’re reading the scan, and then what’s happening?‌

Howard Forman: It’s giving you an alert saying, “We think there’s an intracranial bleed. We think there’s a cervical spine fracture.”‌

Harlan Krumholz: Is it flash up on your screen—‌

Howard Forman: Right.‌

Harlan Krumholz: …or how does it work? ‌

Howard Forman: It effectively flashes on the screen. It’s in the lower left corner.‌

Harlan Krumholz: Does it give you a percentage? There’s a percentage chance or it’s just telling you you should look closer.‌

Howard Forman: Exactly. And it’ll even give you a hotspot to look at if it’s a big image.‌

Harlan Krumholz: Now you’re someone who wrote a really great paper several years ago, that was where you had double-read scans in the ED. So you’re really an expert on quality and whether or not the variability within the ED radiology readings and what people need to do. I thought that was a brilliant piece you wrote. It raised a really important question as to how much is being missed and what double reads for high consequence scans be important. Do you see this as the second read?‌

Howard Forman: Oh, so there’s no question. It’s an extra set of eyes. I’ve called it that many times when I’ve written piece, like review pieces about this. It is an extra set of eyes. And when you’re in a reading room of residents, and maybe you had this experience when you read cardiac echoes, an extra set of eyes never hurts. I mean—‌

Harlan Krumholz: But is it forcing you to overread? Because now it’s like getting you—once it taps you on the shoulder and says “What about this?” it’s very hard for you to overturn that. ‌

Howard Forman: So there are—‌

Harlan Krumholz: That’s called “automation bias.”‌

Howard Forman: There are clearly people that are irritated by that, and I think depending on your level of confidence, it can fluster people at times.‌

Harlan Krumholz: How often will you overrule it?‌

Howard Forman: Oh, more than half the time. Because for certain of the widgets, it’s terrible. For an intracranial bleed, it’s terrible because it’s going to alert me about an intracranial bleed on someone who’s having a follow-up scan from an intracranial bleed. I already knew it. It’s going to take calcifications inside the brain and say it’s a bleed. Not true. It’s going to show me a scar from previous surgery and say it’s a bleed. And I’ll say, “No.” So it is a lot of that. The pulmonary embolus one is probably the most impressive one. That’s the one where the false positives are relatively low and the true positives where I would’ve missed it‌

Harlan Krumholz: And it’s worth it to you. ‌

Howard Forman: to me, it’s definitely worth it. It’s definitely worth it. ‌

Harlan Krumholz: And the other question I always wonder is the moat. So this is just really about data and algorithms. So can’t anyone enter this market?‌

Howard Forman: I think they can. I think the big thing with Aidoc is, or a-doc as I think they call themselves now, is that they’re “non-intrusive,” that’s always been my phrase for it. Their way of doing it and the alerts and everything like that, it’s actually very effective. It does not make you feel like you’re being—‌

Harlan Krumholz: And is it a closed loop system? And what I mean by that is that when you overrule it, it knows it because you’re sharing data back with it, and it’s learning how to get better all the time. ‌

Howard Forman: If it is, it’s doing that very poorly. And I say that advisedly, I don’t want to insult the people at Aidoc, but it has not gotten—in my opinion, it has gotten not materially better. Every time we’ve installed something, it’s been good and then it does not seem to get better. I’m one of the few people that will notify them about false positives, but I still don’t see that going anywhere. I feel like it’s going into a black box and they may or may not be using it. They’re certainly not figuring out some very common false positives and just saying, “Gone. We don’t have to look at these anymore.” It doesn’t happen. So I presume they will make some substantial improvements over time. If I have one criticism of them, it’s that they have not materially improved after they install each widget.‌

Harlan Krumholz: So interesting the way it’s all getting automated. There was one thing you told me a couple weeks ago that I thought was really, really interesting, which is that, just the way radiology is going there is that a lot of your colleagues are reading from home now people aren’t coming into the institution and they’re just home. They’re using these tools, but they’re not actually onsite anymore.‌

Howard Forman: No, no. I don’t know what percent it is, but if I had to guess I’d bet a majority of our cases are being read remotely by our own people, but remotely in their homes and certainly nationally, even more than that. If you look at job boards right now, people will not even take a pure onsite job anymore. The best they’ll do is a hybrid onsite and a virtual job. And a lot of people are doing pure virtual jobs. I know of people who are taking a job in New Jersey and living in Seattle.‌

Harlan Krumholz: For everyone listening, we’re talking about radiology. We’re talking about radiologists.‌

Howard Forman: Yeah, radiologists. But I mean, I know of people personally who literally with a straight face say, “Oh, I got a great job in New Jersey, and I just bought a home off an island in Washington state.” And I’m like, that—‌

Harlan Krumholz: And I told you, this was just so strange to me because when I grew up, we would huddle with the radiologists. I mean, we would have rich conversations, they would show us things, we would talk to them, we talked to ’em about the patients. And I just think some of this loss, let alone you had mentioned to me what the residents experience.‌

Howard Forman: I agree. I think I work on site personally a hundred percent. I don’t have a workstation at home and I work with the residents and I really enjoy the fact that the clinicians come back. I can tell you in the last week, at least 10 visits from colleagues both in trauma surgery, emergency medicine, and from pediatric medicine, pediatric emergency medicine, come back with good questions, want to go over a case, just want to talk through a patient. And I think it makes a material difference. I don’t think it happens as easily virtually.‌

Harlan Krumholz: But for the workforce, you couldn’t force people to come back now, just the market’s not that. ‌

Howard Forman: Certainly not a hundred percent. No way. ‌

Harlan Krumholz: Yeah, it’s interesting.‌

Howard Forman: Yep. ‌

Harlan Krumholz: All right, what else you got?‌

Howard Forman: So it’s the 60th anniversary of the signing of Medicare and Medicaid.‌

Harlan Krumholz: I was thinking that we should talk about that!‌

Howard Forman: We should just at least mention it. I think it’s good to remind people that first of all, one thing I like to say in class, everybody is either already on Medicare or aspires to be at some point. So I think that’s true. And Medicaid of course is the core of our healthcare safety net. And combined the two programs directly cover roughly 40% of the population and indirectly touch every single one of us. So let’s just hope that we can continue to improve these programs that’s been a core part of your career rather than dismantle or undermine them.‌

Harlan Krumholz: Isn’t this really about the power of policy? I mean, look, it was a moment in time. Johnson’s able to push the legislation through. It changed lives. I mean, it changed our healthcare system. It also caused some perversions in the healthcare system too, because it really doubled down all the fee-for-service. ‌

Howard Forman: It’s far from perfect, far from perfect. ‌

Harlan Krumholz: But it did mean that if you were going to become 65, you in this country for the first time could know that you were going to be covered.‌

Howard Forman: That’s right. And we cover them reasonably well. I mean particularly that we also have Medicaid, which means that those low-income Medicare beneficiaries actually get help with their deductibles and their copays and their premiums. We do a reasonably good job around the edges for a large number of people in the country. And we still leave too many behind.‌

Harlan Krumholz: Just so speaking of Medicare, I wanted to actually just finish this string of things I’ve been talking about over the last couple of weeks. I’ve talked about what’s going on in the insurance industry and how are things going. ‌

Howard Forman: I know.‌

Harlan Krumholz: I talked about Elevance and they had to readjust their expectations. Their stock dropped. Oscar, by the way, also had that. UnitedHealthcare came out this week. And this is the largest Medicare Advantage insurer in the country. And it’s saying we’re pulling out—with their Quarter 2 report they’re saying, we’re going to pull out of certain markets. That’s going to impact maybe 600,000 beneficiaries. And on this call that they had, the CEO now, Tim Noll, said that their pricing assumptions were way off and now they’re predicting more than $6.5 billion additional Medicare costs than what they thought they were going to get for 2025, $6.5 billion more than they had calculated. They’ve got actuaries, they’ve got people, all these people. I mean they’ve got highly sophisticated bean crunch bean counters.‌

Howard Forman: It’s hard to predict this stuff. It just is.‌

Harlan Krumholz: Well, and then more than half of this excess cost is coming from Medicare Advantage.‌

So where on one hand we’re seeing in D.C. people are enamored of Medicare Advantage. They think they’re going to be able to control the costs, but this year at least Medicare Advantage is becoming more difficult to handle. And I’ve also heard from some employers who are getting sticker shock. It’s cutting into their margins; it’s interrupting their businesses. I have not seen people write about this much. And I’m talking about you’ve got the self-insured, which are the large businesses, but these medium-sized businesses are being hit really hard with these increases. And then you’ve got towns who are trying to manage their budgets and having to raise taxes just to cover the increasing medical care costs. So I’m going to predict this year, this is all going to come to a head. This is going to be really, we’re finally getting to a point where people are—‌

Howard Forman: It may. I just—‌

Harlan Krumholz: Well, especially if you make all these people uninsured too, it’s going to be a lot of pressure on the system.‌

Howard Forman: That’ll happen a little bit later than this next year. But I’m with you. Look, we have not been able to fully explain why costs have been so high, both for hospitals as well as for the insurance companies this past year. We know some of it, the GLP-1s are a part of it. I mean, we have a much larger population now coming in for treatment. And by the way, even the GLP-1s are not simple anymore. Novo Nordisk stock went down by like 20% this week. ‌

Harlan Krumholz: Shuffling with a new CEO and stuff like that. ‌

Howard Forman: And they’re not doing as well as people... for a long time they looked like a meme stock.‌

Harlan Krumholz: A juggernaut.‌

Howard Forman: Just going straight up in line. And even they have now come down considerably. So healthcare’s hard. And I am very cautious and humbled every year by what hits us on all different sides. Our hospital has still not gotten back on its feet in the same way it was before the pandemic. And I don’t think anybody can precisely explain how that happened.‌

Harlan Krumholz: We knew fresh new ideas. I really like our friend Zeke Emanuel was writing about this and saying—of course he’s advocating for Democrats—don’t just call for people to be insured or roll back the Big Beautiful Bill. Think of what do we need for the redesign? How is this going to look better in the future? And how can we address some of the shortfalls because it’s far from perfect.‌

Howard Forman: We can link in our show notes. It’s a piece in The Bulwark where he really lays out specific ideas. And one of them is like, stop having people enroll one year at a time. Commit them to five years at a time so that insurance companies actually can make some types of strategic investments that have payoffs over a multi-year window. We don’t do that particularly well. And quite frankly, I sit on the board of University Health at Yale, and so I get to see how our Yale University Health Services functions and they make very hard decisions all the time. And I think they do an extremely good job both balancing holding down costs but also being able to provide access to GLP-1s and other treatments as they emerge no matter how expensive they are, including CRISPR technologies and other genetic treatments. Gene treatments.‌

Harlan Krumholz: Yeah. Yale’s kind of unique in this way. I mean, there’s just a few like Yale and it’s... yeah, that’s cool that you’re on the thing. So Howie, we’re at summer break. I just want to thank you again. What a pleasure it is to do this with you. Been fun. It’s a real privilege. And for everyone out there, we thank you for listening. So let’s just end. We will do our final outro here. ‌

Howard Forman: Yeah, no, and let me just say also, we have a great lineup for next year. I mean, I get a little—‌

Harlan Krumholz: Well, we’re going to hit escape velocity next year. Is that what you’re saying?‌

Howard Forman: I think we’ve already hit it. And let me just also say to the people out there, the number of people that have come up to me in the last few weeks by email or in person and have told me that they were excited about one or more episodes spontaneously, without any reason to do it. ‌

Harlan Krumholz: No. I’m sending them money, and I’m asking them to.‌

Howard Forman: Well, I appreciate that, Harlan. Someone has to do the marketing, but I just want to say it takes a village, by the way. I mean we’re going to thank Gloria and Tobias and Miranda again, and I thank you. But it is—‌

Harlan Krumholz: Who knew it would grow to this?‌

Howard Forman: Four years. We’re going to have our 200th episode coming up in this next academic year.‌

Harlan Krumholz: And of course for me, it’s just selfish. I get to talk to you every week. So this is great.‌

Howard Forman: Listen, you’ve always been like my big brother and my idol and—‌

Harlan Krumholz: Oh my God, ‌

Howard Forman: I’m so glad that I get to work—‌

Harlan Krumholz: No, come on. You’re the role model for all of us, anyway. So you’ve been listening to Health & Veritas with Harlan Krumholz and Howie Forman.‌

Howard Forman: So how did we do? To give us your feedback or to keep the conversation going, email us at health.veritas@yale.edu or follow us on any of social media.‌

Harlan Krumholz: And we love your feedback, especially your positive feedback.‌

Howard Forman: But we’ll take negative feedback!‌

Harlan Krumholz: Please keep giving Howie positive feedback. We love it.‌

Howard Forman: I need it. I need it.‌

Harlan Krumholz: But rank us, rate us, helps people find us. We appreciate it.‌

Howard Forman: And if you have questions about the MBA for Executives program, the Yale School of Management, reach out via email for more information or check out our website at som.yale.edu/emba.‌

Harlan Krumholz: Health & Veritas is sponsored by the Yale School of Management, the Yale School of Public Health. We are so grateful for them. And we’re really lucky to be working with superstar undergraduates. Thanks you, Howie. Thanks for your recruitment from your class, Tobias Liu and Gloria Baek. Tremendous, amazing. And we have a remarkable producer who somehow every week makes us sound better than we actually are, which we always appreciate.‌

Howard Forman: And they’re all joining us virtually or in person in the studio today. So we get to thank them in person.‌

Harlan Krumholz: And I get to work with the very best co-host of all time. Thank you so much, Howie. Talk to you soon. ‌

Howard Forman: Right back at you, Harlan. Thank you, Harlan. Talk to you soon.‌