Iris Holm: You know what kind of week it's been when the most clarifying thing you read is a budget line.
Cyrus Reed: Ha — okay, what budget line?
Iris Holm: CDC: ninety percent of U.S. healthcare expenditures — five-point-three trillion dollars a year — goes to chronic and mental health conditions. Not acute. Chronic.
Cyrus Reed: And yet — wait, and yet who does the system actually reward? The surgeon. The emergency physician. The person who pulls someone back from the edge.
Iris Holm: That's the structural tension. Acute care has an endpoint. A measurable crisis, a visible intervention, a patient who goes home. Chronic care has thirty years of management and an outcome that might just be the absence of something bad.
Cyrus Reed: The absence of something bad — how do you even build a career around that? Like, not philosophically, but actually — how does a clinician build a professional identity around something no one can see happened?
Iris Holm: Most don't. And the system's never required them to. The epidemiological transition — the whole twentieth-century shift where chronic disease overtook infectious disease as the primary cause of death — none of that reshaped the training or the reimbursement. The RAND Corporation finds patients with multiple chronic conditions accounting for over two-thirds of costs, and the structure still runs like it's 1950.
Cyrus Reed: No way — over two-thirds. So it's not even a marginal problem, it's the bulk of the load, and the system is — it's just oriented the wrong direction.
Iris Holm: And it compounds. The American Heart Association projects hypertension prevalence going from 48.6% in 2020 to 59.1% by 2050. That number is not a surprise finding. It's a forecast of what you get when prevention doesn't scale.
Cyrus Reed: So we have the data, we have the forecast — wait, is that the whole problem? We know and we're just not acting on it? Or is it something deeper, like the knowing doesn't even connect to the decision-making in any useful way?
Iris Holm: The knowing is there. The architecture isn't set up to act on it. That's what I want to pull apart — why rational actors inside this system keep making choices that look irrational from the outside.
Cyrus Reed: Because the system is rewarding them for it. They're not making mistakes — they're just responding to what the structure actually values.
Iris Holm: Right — but responding to what the structure values only gets you so far as an explanation, because it still doesn't name the mechanism. Why does the structure value acute care so specifically?
Cyrus Reed: Wait, I think it's actually pretty simple when you strip it down. Two doctors. One pulls someone back from a heart attack in the ER tonight. The other quietly kept someone's blood pressure controlled for a decade and that patient never had a heart attack. The system knows exactly, precisely, what the first doctor did. It has — there's a discharge note, a billing code, a moment. The second doctor? Nothing shows up anywhere, because nothing happened.
Iris Holm: That's the whole problem in one image.
Cyrus Reed: And the thing that gets me — it's not even that the system is being cruel about it, it's that non-events are just genuinely invisible. Like, how would you even write the metric? 'Stroke did not occur here'? That's not a line item.
Iris Holm: There is no discharge note for a stroke that didn't happen. No moment of triumph. The primary care physician managing blood pressure for ten years gets — nothing attributable. That's the attribution problem. Not a design flaw. The architecture.
Cyrus Reed: Wait — so it's not that the system is indifferent to chronic care, it's that chronic care outcomes are structurally defined by absence, and absence doesn't register.
Iris Holm: And you can't reward an outcome that has no presence in any record. Fee-for-service pays per procedure, per intervention — per thing that happened. Prevention doesn't generate a thing that happened.
Cyrus Reed: So the system defaults to rewarding — presence. The dramatic visible thing. And that's not irrational from inside the structure, that's just — that's what's measurable.
Iris Holm: Now here's where it gets uncomfortable. Andreas Conte and W. Jerjes — the Time Capsule Medicine pilot — they're arguing that if you simulate a twenty-year patient trajectory for medical students, you reshape how they think about chronic care. And maybe you do. But the attribution problem doesn't live in what students feel during training.
Cyrus Reed: No, it lives in the payment structure they graduate into.
Iris Holm: Exactly where the simulation can't follow them.
Cyrus Reed: So you can change what a doctor experiences in year one of medical school, but the moment they're billing under fee-for-service, the acute-care heroics narrative — the thing that made emergency medicine feel like the real work — that reasserts itself, because that's still what actually gets credited. And the prevented stroke is still invisible. It was always going to be invisible.
Iris Holm: And that's not even the bottom of it. Fee-for-service doesn't just ignore prevention — it actively has no billing code category for 'this did not deteriorate.' Every reimbursable event in the current architecture is something that occurred. A procedure. An encounter. A test ordered. The chronic care visit where nothing changes is, financially, almost indistinguishable from nothing.
Cyrus Reed: Wait — so behavioral counseling, longitudinal management, the actual mode of care that works for chronic disease — those are the least reimbursed activities in the whole system?
Iris Holm: By design. Not by accident.
Cyrus Reed: But then — okay, people keep proposing value-based care, capitation models, alternative payment structures. Are any of those actually — I mean, do they solve this, or are they just redistributing the same invisibility problem into a different accounting format?
Iris Holm: That's the question. And the honest answer is: proposed for decades, not demonstrated at scale. Value-based care sounds like it realigns the incentive. But the attribution problem doesn't disappear inside a capitation model — it just moves. Now a system gets rewarded for population outcomes it still can't cleanly trace to individual clinical decisions.
Cyrus Reed: So you've redistributed the credit problem without actually solving it. Huh.
Iris Holm: And the NIHCM analysis flags something that the payment reform conversation keeps stepping around — it's not only financial. Clinicians aren't culturally rewarded for chronic care either. The acute-care heroics narrative — the whole professional identity built around the antibiotic era, emergency surgery, the dramatic visible save — that's baked into training hierarchies, hospital prestige, what gets written up in grand rounds.
Cyrus Reed: So even if you fixed the reimbursement tomorrow, the identity signal is still wrong. The system is telling every resident, through every structure around them, that the ER is where the real work happens.
Iris Holm: Right. Two separate failure modes running in parallel — financial and cultural — and they reinforce each other. Fix one and the other holds the bias in place.
Cyrus Reed: Which is — wait, that's actually a harder problem than I thought it was ten minutes ago. Because you can legislate a payment model. You can't legislate what feels important to someone at two in the morning during their residency.
Iris Holm: And that's before we get to where the training identity problem and the drug development economics collide — that part makes this whole reimbursement argument look almost solvable by comparison.
Cyrus Reed: Oh — how does drug development even fit into this? Because yeah, now I need to know where that goes.
Iris Holm: Drug development has the exact same endpoint problem. The pipeline skew — cancer, rare diseases — that's not pharma being callous. That's pharma responding to trial economics.
Cyrus Reed: Wait — what do you mean trial economics specifically?
Iris Holm: A cancer trial has a clean endpoint. Tumor response. Survival at eighteen months. You can finance that, adjudicate it, get FDA approval on a legible timeline. A hypertension trial? You need a decade of outcome curves. Did the patient have a stroke? Did they not? When? Why?
Cyrus Reed: So the pharmaceutical pipeline skew — toward cancer and rare diseases and away from hypertension, diabetes, arthritis — that's not irrational either. It's the attribution problem wearing a different coat.
Iris Holm: Same mechanism. Different industry.
Cyrus Reed: But wait — those chronic conditions, hypertension, diabetes, they're the actual disease burden. Like, they account for the largest share of what's actually killing people, and the pipeline just... concentrates elsewhere because the trial math works out better.
Iris Holm: Orphan drug economics are genuinely favorable. Smaller trial populations, accelerated pathways, premium pricing. Chronic disease — massive market, but the development cost and timeline make the return harder to model. So capital flows toward clarity.
Cyrus Reed: Okay and now — now I'm thinking about the training side again, because this is — actually, no, these two things are colliding in a way I didn't expect. Medical students are trained around visible acute interventions, acute-care heroics, and then the drug development world they graduate into is also skewed away from the conditions that don't produce clean endpoints. So the whole professional environment, from day one of training to what drugs are even available, is just... pointing the same direction.
Iris Holm: And that's before the Andreas Conte and W. Jerjes study enters the picture. The Time Capsule Medicine pilot — students simulating twenty years of progressive chronic illness — it's explicitly trying to build experiential identity around long-term management. Which is exactly the right diagnosis of the problem.
Cyrus Reed: But it's a pilot. At the margins.
Iris Holm: Every structural signal outside that simulation still rewards acute. The training hierarchy, the hospital prestige, what gets written up, what gets billed — none of that moved.
Cyrus Reed: So you change what a student feels for a few hours in a sim lab, and then they walk back out into a system where — I mean, medical education historically organized everything around protocols, visible interventions, outcomes you could see inside a single shift. That identity was built over a century. One pilot study is not going to — it can't hold against that.
Iris Holm: Now drop the number that makes this concrete. Thirty-seven million Americans have chronic kidney disease. Most don't know it. The CDC data makes this explicit.
Cyrus Reed: Wait — most don't know it? How does that — thirty-seven million people, and the system hasn't flagged it?
Iris Holm: Because catching it early produces no visible endpoint. No discharge. No procedure. No moment. There's no reimbursable event for 'we found this before it became a crisis.' The system doesn't reward the catch — so the catch doesn't happen at scale.
Cyrus Reed: So the deepest irony is — the problems that are actually dominating disease burden, the ones the American Heart Association is projecting will get dramatically worse by 2050, those are precisely the problems the pipeline won't prioritize, the training doesn't build identity around, and the payment system can't see. Not by accident. That's just — that's what the logic produces.
Iris Holm: And I can't stop thinking about this. The American Heart Association's 59.1% figure — that's not a warning. It's not 'this could happen.' It's a description of what's already baked in. The forecast assumes no structural change, and nothing in what we've mapped today suggests structural change is coming.
Cyrus Reed: No, it's not. Which means — wait, does the system eventually just split? Like, does it bifurcate into two actual separate tracks — acute-care hospitals doing the crisis work they're built for, and some entirely separate chronic-disease management infrastructure that has to get invented almost from scratch? Because those 200 million people managing hypertension, diabetes, arthritis right now — they're not waiting for the redesign.
Iris Holm: They're absorbing the friction. Every day. That's the other possibility — no bifurcation. The system just never splits, and the cost of the mismatch keeps landing on the patient. On the person with Type 2 diabetes trying to navigate an institution that is genuinely, architecturally, built for someone else's emergency.
Cyrus Reed: Huh. And I don't — I mean, I genuinely don't know which of those is worse. The bifurcation might actually be fine, if the chronic track gets properly resourced. But I'm not sure it would. It would probably just be — the underfunded track. The one without the prestige.
Iris Holm: Frankly, that's where I land. Unresolved. The mismatch we've been mapping isn't moving toward a fix — not on its own timeline. And the people inside it aren't holding still waiting for NIHCM or anyone else to publish the right framework.
Cyrus Reed: That's — yeah. Good place to stop, I think. Thank you for working through this one.
