The River Doesn't Need a Faster Channel
NFF: Medicine is extraordinarily good at responding to illness. That's the problem.
The most valuable thing I can do for a patient is make the next visit unnecessary. There’s no billing code for that.
I think about this a lot in my practice. I see patients virtually, across all 50 states, running near full-scope primary care. The breadth of that panel gives you a kind of pattern recognition you don’t get in a geographically bounded practice — different populations, different access histories, different points of entry into the system. And one of the patterns that gets clearer at scale is this: the patients consuming the most care are almost never the sickest patients. They’re the patients the system failed to catch before the problem compounded.
The fifteenth visit of the year is downstream of something. Usually a lot of somethings.
The Channel Is Not the Problem
Every major healthcare innovation I’m watching right now is a supply-side optimization. AI scribes cut documentation time. Virtual care eliminates the commute. Async messaging fills the gaps between visits. Prior auth automation reduces the administrative drag on an already-delayed decision. These are real problems, and the solutions are genuinely useful. I use some of them. I’m building with the logic of some of them.
But watch the assumption underneath all of it.
The implicit premise is that demand for sick care is fixed — a river that needs a better channel. So we engineer the channel. We make it faster, cheaper, lower friction. More access. Better throughput. Reduced time-to-care. The pitch is always some version of the same sentence: we can get care to more people, more efficiently, than before.
Nobody is asking whether the volume of care we’re routing is the right target.
Why No One Touches Demand
Medicine has never seriously tried to shrink consumer demand for sick care. Not because clinicians don’t want to. Because the entire economic architecture of healthcare is built on volume.
Fee-for-service doesn’t reward the visit that never happened. Health systems don’t post earnings calls about hospitalizations prevented. The quality metrics that actually drive institutional behavior are mostly process measures — did we do the thing, not did the thing need doing. Demand reduction, inside that model, registers as revenue destruction. So the incentive to build toward it is close to zero.
The result is a system that’s extraordinarily sophisticated at responding to illness, and has almost no infrastructure for making illness less likely.
So we build faster ERs. We don’t ask why the emergencies compound. We automate the prior auth. We don’t ask why the chronic disease went unmanaged for a decade before it showed up in an acute visit. We add the virtual slot. We don’t ask why the patient needed visit number fifteen.
The system treats each of those visits as a unit of successful care delivery. They are also, every one of them, a signal that something upstream failed.
Prevention Theater Is Not the Answer
I want to be careful here, because the obvious response to this argument is wellness programs and prevention initiatives, and most of what gets labeled prevention in healthcare is not what I’m talking about.
A wellness portal is not demand reduction. A step-count badge is not demand reduction. A chronic disease app that tracks symptoms without changing the conditions that produce them is not demand reduction. That’s demand management with softer marketing.
Real demand reduction is behavior change infrastructure embedded in the daily life of a patient before the problem becomes a problem. Not at the point of the visit. Not at the point of the test. Before any of that.
This has been tried at scale. Kaiser Permanente built the closest thing American medicine has ever produced to a demand-reduction model — integrated care, salaried physicians, a financial incentive to keep patients well rather than keep them coming back. It works, in meaningful ways, and the outcomes data shows it. But it also carries a reputation most people in healthcare know without having to say out loud: the patient who just wants to see the ENT can’t always get to the ENT. The system optimizes for population health and sometimes does it by slowing individual access. That’s not a conspiracy. It’s what happens when you tune a system for efficiency without fully solving for what the patient actually wants.
The tension isn’t between prevention and access. It’s between system logic and consumer need. Kaiser figured out the system side. The consumer side remained an acceptable casualty.
That’s the gap. And it’s the gap I’ve been thinking about for a few years.
There’s a path that doesn’t require choosing between them — that keeps the demand-reduction logic intact without turning the primary care physician into a gatekeeper the patient has to outmaneuver. It doesn’t require fixing fee-for-service. It requires solving a different problem, one that’s been sitting inside the existing architecture the whole time. Most builders aren’t looking there because the supply-side keeps pulling their attention back.
I’m not ready to put the full model in writing yet. The framing still matters. But the direction is clear enough that I can say this: the solution isn’t more access and it isn’t less access. It’s building a system that makes patients genuinely want to need less of it — and that meets them where they are when they do.
The Most Defensible Position
The builders who figure out how to make their patients need less care will own the most defensible position in healthcare over the next decade. Not because it’s noble, though it is. Because it’s a moat. A patient who stays healthier, longer, with fewer acute episodes, is not an abstraction. That’s a retention story. A cost story. A differentiation story that no supply-side optimization can replicate.
The visit is not the product. The visit is the failure mode.
Whatever comes after that assumption changes everything about how you build. More on this soon.
If you’re working on this problem — or watching someone else try to — leave it in the comments.
Hey — I came across your writing and really liked how you think.
I’m exploring something similar from a different angle — writing about human behavior through a system design lens (like debugging internal patterns).
Just started publishing on Substack. If you ever get a moment to read, I’d genuinely value your perspective.
Also happy to support your work — feels like there’s an interesting overlap here.
There is kind of a billing code for making the next visit unnecessary, it’s CCM… we use it heavily for that