Jonathan Ingles: Ben, quick one before we get into it — have you ever been told you should be on a medication that you didn't take?
Ben Okonkwo: Oh, hm — yeah, actually. Vitamin D, I think, three years ago. Just never filled it.
Jonathan Ingles: You are the problem. Statistically, you are the problem. Because under the 2018 ACC and AHA guidelines, roughly half of statin-eligible adults weren't receiving the drugs — that's in a JAMA editorial from July 20th, 2026.
Ben Okonkwo: Wait — that's the same publication that just announced the new eligibility expansion?
Jonathan Ingles: Same issue. University of Pittsburgh, University of Michigan — they ran the numbers on the 2026 guidelines the ACC and AHA issued in March. 87.5 million Americans now eligible. 56.6% of adults 30 to 79. An increase of 21.5 million over what 2018 authorized. And on the same day JAMA prints that, they print an editorial saying the prior generation of eligible patients aren't even on the drug.
Ben Okonkwo: So the framing you're building toward is — this expansion is solving the wrong problem.
Jonathan Ingles: Not building toward it — it's obvious. The newly eligible 21.5 million are younger, lower short-term risk. People who feel fine. The ACC and AHA didn't see a medical gap. They rewrote the criteria because the prescription rates weren't where they wanted them, and this was an easier lever than fixing physician-patient communication at scale.
Ben Okonkwo: Now, I want to separate the motive claim from the evidence claim there, because those aren't the same argument — and one of them is much harder to prove than the other.
Jonathan Ingles: Fair. But before you walk the evidence chain — explain what actually changed mechanically, because I think most people don't have that.
Ben Okonkwo: Right — so here's the plain version. The old calculator was basically a smoke detector that only went off when flames were already close. The PREVENT calculator, which the ACC and AHA dropped into the March 2026 guidelines, asks: is there a faint smell of smoke thirty years from now? That's not a tweak — that's a structural reorientation of what the tool is even measuring.
Jonathan Ingles: Sixteen months.
Ben Okonkwo: November 2024 to March 2026, yes. Now, interesting thing is — PREVENT wasn't just bolted on top of the old thresholds. The thresholds got rebuilt too. High risk is now ten percent ten-year probability. Intermediate is five percent. Borderline — and this is the part that actually — wait, structurally significant — borderline is now three percent. And if your ten-year risk is below three percent, you can still qualify if your thirty-year projected risk clears ten percent.
Jonathan Ingles: So someone who shows zero near-term signal gets medicated on a projection that won't resolve for three decades.
Ben Okonkwo: Or — and I'm not defending it yet, I'm just mapping it — a thirty-eight-year-old with an LDL above 160 or a strong family history of premature ASCVD now gets a statin conversation that previously started at forty. That's a named threshold in the guideline. Statins have been in use for over forty years, so the drug class itself isn't the novelty. The question is whether the model driving who gets it is mature enough to carry that weight.
Jonathan Ingles: And that's where I want to press — PREVENT is sixteen months old at the time of adoption. Prior risk models had decades of real-world outcome data behind them. This one doesn't.
Ben Okonkwo: That's the exact question I have. The validation studies exist — but those came from the developers. Independent real-world calibration at scale, for a thirty-year projection horizon? That data can't exist yet. Thirty-year predictions don't have a thirty-year track record. So we're treating modeled risk as if it were measured risk, and — yeah. That gap matters.
Jonathan Ingles: That's the whole argument. You can't falsify a thirty-year claim for twenty-nine more years.
Ben Okonkwo: And that unfalsifiability cuts sharpest exactly where the new pathway is doing the most work. Picture a 36-year-old — Saturday afternoon, she's picking up a prescription at a pharmacy. Her 10-year cardiovascular risk is 2.4%. Under the old frame, she doesn't even get the conversation. But PREVENT's 30-year model tips her over 10%, so she qualifies. No symptoms. No disease. Just a number from a calculator that's been in the field sixteen months.
Jonathan Ingles: That is the kernel. Right there.
Ben Okonkwo: And the gap I can't close — actually, I want to be precise here — it's not that 30-year modeling is inherently wrong. It's that independent external validation over a long timeframe is limited compared to older calculators. The developers validated PREVENT. That's not the same as replicated real-world calibration across diverse populations over decades.
Jonathan Ingles: Joshua Sussman — chair of the AHA's Primary Care Science Committee — said physician-patient dialogue is central to how this gets applied. Fine. But you've just labeled 87.5 million people eligible. That's a population-level frame. It doesn't support individualized dialogue, it pressures it.
Ben Okonkwo: That's — yeah. That tension is real.
Jonathan Ingles: The irony is almost too clean. Sussman defends the guideline on individualization grounds while the number being broadcast guarantees it gets processed as a mandate.
Ben Okonkwo: So your partial win — and I think it is a win — is specifically this: the sub-3% 10-year pathway into pharmacotherapy rests on a projection that cannot be validated in the population it most affects for another three decades. That's not a political claim. That's a study-design limitation.
Jonathan Ingles: Exactly what I've been saying.
Ben Okonkwo: Now — and I'm flagging this because it gets worse — there's a sequencing question underneath all of this about what the benefit-risk ratio actually looks like in low short-term risk populations versus where the expansion is largest. That's the part I want to get into.
Jonathan Ingles: And that's the payoff — the benefit-risk ratio is least symmetric exactly where the expansion is largest. Low short-term risk means the absolute risk reduction from statins is small. The drug is certain. The benefit is probabilistic and far out. That's not a wash, that's an asymmetry.
Ben Okonkwo: Right — but let me be precise about what that asymmetry means clinically, because this is where I want to be careful. Statins have forty years of replicated evidence in moderate-to-high-risk populations. That's robust. But that evidence base doesn't automatically transfer to healthy adults qualifying on a thirty-year model output. Those are different populations, different absolute risk baselines.
Jonathan Ingles: The JAMA editorial named this directly.
Ben Okonkwo: It did. And here's the structural problem — the framing we've been given comes almost entirely from the ACC and AHA. The institutions that wrote the guidelines are also the dominant voices characterizing what the guidelines mean. Independent critical clinical voices are underrepresented in this conversation. That's not a conspiracy claim, it's just — hm, it's a sourcing gap that shapes what looks like consensus.
Jonathan Ingles: So the calibrated version isn't 'the guidelines are corrupt.' It's something narrower and actually harder to dismiss.
Ben Okonkwo: Right — the defensible claim is this: the implementation gap is larger than the eligibility gap. Half of the people already eligible under 2018 rules aren't on statins. That number, 21.5 million newly eligible people under a sixteen-month-old calculator — that doesn't close the adherence problem. It just makes it wider and harder to track.
Jonathan Ingles: Expand only after you've solved the fifty percent treatment gap. That's the sequencing argument. And it's not ideology — it's basic implementation logic.
Ben Okonkwo: And I'll commit to that. The evidence licenses cautious expansion — not because the guidelines are wrong on the science, but because pharmacotherapy for a population defined by a model prediction rather than observed disease needs more than sixteen months of real-world calibration before it's the load-bearing infrastructure for 87.5 million people.
Jonathan Ingles: Fine. It's not a marketing scheme. It's a sequencing problem. Which, honestly, is worse — because a sequencing problem means the institutions knew what the implementation gap looked like and expanded anyway.
Ben Okonkwo: Yeah. That's — I think that's where I actually land too. And the uncomfortable part is that adding 21.5 million people to the eligible list doesn't treat a single one of them. Not one. Until the ACC and AHA solve what Joshua Sussman called the physician-patient dialogue problem, that number is a projection, not a promise.
Jonathan Ingles: Mm. Fair place to stop.
Ben Okonkwo: Good talk. Genuinely.
