Aspirin for Primary Prevention: A Risk Score Just Under the Guideline's Cutoff
A single patient, his calculated ten-year cardiovascular risk sitting just under the threshold most guidelines use before they'll even discuss aspirin. The disagreement isn't about his risk calculation — it's about how much weight a family history the calculator doesn't ask about should carry.
S.A., a 58-year-old man, has spent the last eleven years as a dental hygienist at a practice he describes as barely changed since he started — same chairs, same patients now bringing in their own children. He runs three mornings a week, not competitively, mostly to keep his knees from complaining, and until this visit his only real medical history was hypertension, diagnosed six years ago and controlled since on lisinopril 10 mg daily. He has never smoked, drinks occasionally, and has no history of peptic ulcer disease, gastrointestinal bleeding, or falls.
He's here for what was billed as a routine physical, but his physician used the visit to run a formal ten-year cardiovascular risk calculation — prompted less by anything acute than by a passing comment S.A. made about his father, who had a heart attack at fifty-eight, the same age he is now. His labs came back with a total cholesterol of 211 mg/dL, an LDL of 138 on his current moderate-intensity statin, an HDL of 52, and a systolic blood pressure of 128 on treatment. Entered into the Pooled Cohort Equations, those numbers put his ten-year ASCVD risk at 8.7% — squarely in the range guidelines call intermediate, but under the 10% figure the USPSTF specifically uses to decide whether aspirin is even worth discussing for a man his age. The calculator has no field for his father's heart attack.
At the follow-up, deciding what the calculator left out
His family history is about as specific as this kind of information gets — a first-degree relative with an MI at the exact age he is now. I'll concede the technical point before it's made for me: the 2018 ACC/AHA cholesterol guideline defines premature ASCVD in a male relative as an event before fifty-five, so his father at fifty-eight sits three years outside the risk-enhancer as the guideline actually writes it. What it doesn't sit outside is the reason the guideline named that enhancer in the first place — the Pooled Cohort Equations were never claimed to be a complete picture, just a starting number, and a cutoff at fifty-five is a convention, not a biological edge. His bleeding-risk side of the ledger is clean — no ulcer history, nothing that raises his hemorrhagic risk above baseline — which is exactly the population the 2019 ACC/AHA primary-prevention guideline had in mind when it said aspirin might be considered, not routinely given, in select higher-risk adults 40 to 70.
8.7% isn't wrong. It was never supposed to be the whole answer for a patient carrying a family history this specific.
The three trials that actually changed this recommendation in 2022 — ASPREE, ARRIVE, and ASCEND — enrolled populations far closer to modern background therapy than the older studies aspirin's primary-prevention reputation was built on, and all three found the same thing: whatever ischemic benefit remains at intermediate risk is reliably offset by major bleeding. That isn't a qualitative judgment call, it's a population result, and it's the evidence the 10% threshold was actually calibrated against.
The risk-enhancer isn't invented, but treating it as license to cross a hard evidence-based threshold isn't the same as what the guideline intended by naming it. And there's a lower-risk move sitting right in front of us that nobody's arguing about: his LDL is still 138 on a moderate-intensity statin. Fixing that does more for his actual risk than starting a drug the guideline no longer routinely offers him.
Both of you are arguing from different kinds of evidence — a qualitative enhancer against a population trial base — and I don't think more discussion resolves that gap, because it isn't actually a disagreement about facts. It's a disagreement about what kind of evidence should be allowed to move a number that's already close. A coronary artery calcium score gets us out of that: it's his own arteries, not his father's history and not a cohort average, and the 2018 guideline was written with exactly this scenario in mind.
Agreed: continue lisinopril unchanged, increase atorvastatin to 40 mg, and obtain a coronary artery calcium score before revisiting aspirin. Nobody starts aspirin today.
Not agreed, and the reason the CAC order carries a real branch point rather than settling the question outright:
The preventive cardiologist agreed this would meaningfully soften his case for aspirin, though not, in her view, close it — a family history this specific isn't fully answered by one imaging result.
The aspirin conversation returns with a patient-specific number behind it, not just a family history and a population equation arguing past each other.
The preventive cardiologist and the clinical pharmacologist left holding genuinely different priors about what a zero CAC score would actually settle — and about how much weight a family history the equation doesn't ask for should carry on its own, independent of whatever the score eventually shows.