The most-cited prescription in primary care, for most of the past forty years, has been a $5 bottle of low-dose aspirin. Not the most-prescribed; the most-cited. Generations of patients arrived at the practice with a daily 81 mg aspirin in the medication list and no clear answer to the question who put you on this. The answer, more often than not, was the cumulative practice of medicine in the United States — built up out of the 1989 Physicians' Health Study, the meta-analyses of the 2000s, and the kind of caution-as-default that takes hold when the cost of a treatment is small and the perceived downside is small.
Then in 2022 the U.S. Preventive Services Task Force, having reviewed the evidence again, did the careful unflashy thing of saying that the perceived downside was bigger than we had assumed. USPSTF moved aspirin for primary prevention of cardiovascular disease in adults aged 60 and older to a D recommendation; for adults 40 to 59 with a 10-year ≥ 10% atherosclerotic CVD risk, the recommendation became C. The bleeding risk, at the population level, was non-trivial; the cardiovascular benefit, at the population level, was smaller than once believed; the calculus had tipped.
I want to spend the rest of this bulletin on what the calculus actually looks like in the room, and on the practical question of how a small primary-care practice rewrites an intake script when the standing recommendation has changed but a non-trivial proportion of the existing panel is still on aspirin.
Why USPSTF and BNF can disagree without either being wrong
The British National Formulary publishes a different set of recommendations for the United Kingdom, and the BNF text on aspirin is more permissive in some primary-prevention contexts than the USPSTF position. This is not, in 2026, a surprise. The two bodies are weighing different things. USPSTF is a technology-assessment body whose remit is whether a preventive intervention has net benefit at the population level, with a specific U.S. context (a higher proportion of the population with risk factors that worsen bleeding risk; a different cardiovascular event-rate baseline; a different frequency of comorbid GI pathology). BNF is a pharmacological compendium that incorporates NICE guidance and a UK-specific context. They are reading the same primary literature and assigning different weights to the harms and the benefits.
This is the friction the bulletin is named for. The friction is not a problem. The friction is the right tool for thinking about a low-cost, low-perceived-risk intervention whose benefit and harm are both small relative to clinical noise.
What the room actually looks like
The patient in the room is a 64-year-old who has been taking aspirin 81 mg for a decade because the previous PCP said to. They have hypertension, well-controlled. They do not have known CVD. Their 10-year ASCVD risk by the ACC/AHA risk estimator is 9%. They have no GI history, no anemia, no known hereditary bleeding disorder.
The conversation we have:
- "The recommendation about aspirin for primary prevention has changed since you started taking it. The benefit at your level of risk is now thought to be very small, and the bleeding risk is non-trivial. I want to walk you through what the change means."
- I show them the USPSTF page on the computer screen. I let them read it. I do not summarize it for them.
- I say: "If we stop the aspirin today, the change in your projected 10-year heart-attack-or-stroke risk is on the order of half a percent. The change in your bleeding risk is also on the order of half a percent, in the other direction. The numbers are small. The decision is mostly a question of which kind of small risk you would rather hold."
- If they want to keep taking it because they have done so for a decade and have had no bleeding event, that is reasonable. If they want to stop because the benefit is now thought to be smaller than they were originally told, that is also reasonable. I document the conversation.
The script above is the one we have used, in some version, for two years. What changed in February 2026 was less the script and more the place at which it occurs.
The intake script change
Until February, the aspirin conversation happened opportunistically — at the annual visit, or when the medication list was reviewed at a quarterly hypertension visit. The conversation happened to the patients who came in often. It did not happen to the patients who came in once a year, on the same day, regardless of clinician.
In February we made two operational changes:
- The pre-visit medication-reconciliation worksheet, completed by the medical assistant ten minutes before the clinician enters the room, now flags any aspirin 81 mg in the medication list as a "discussion item" with a printed one-paragraph summary of the USPSTF recommendation. The clinician sees the flag and the patient sees the summary before the conversation begins.
- The annual visit template now contains a structured field for "shared decision-making documentation" specifically for aspirin in primary prevention. The field has three checkboxes: continued at patient request, discontinued at patient request, continued because patient has known CVD or another secondary-prevention indication. None of these is the default. The clinician enters one.
Both changes are small. The first is operational; the second is documentary. Together they have, in the eight weeks since they went live, produced a measurable difference: 22 of our adult patients aged 60+ have had a documented aspirin shared-decision conversation since February that they would not have had until their next opportunistic visit otherwise. Of those, 11 elected to continue, 9 elected to discontinue, and 2 had a secondary-prevention indication that we had not previously surfaced.
What the patient should expect
If you are an adult on aspirin for primary prevention and you are seen here this year: expect us to bring it up. Expect us to bring it up even if everything else is going well. Expect us to ask you to read the USPSTF page yourself, in the room, before you decide. Expect us to write the decision down in plain language. Expect us, if you decline to discuss it today, to bring it up again at the next annual visit. The friction is the point.
What we got wrong
For the first eighteen months after the USPSTF 2022 recommendation, we relied entirely on the opportunistic conversation. The aspirin discussion was, in effect, available only to patients who came in often. That is not equitable; the patients who come in less often are systematically the patients with less time, less continuity, and (for entirely different reasons) more downside risk from a bleeding event. The intake-script change should have happened in 2023.
References
- U.S. Preventive Services Task Force, Aspirin Use to Prevent Cardiovascular Disease: Preventive Medication, 2022 — uspreventiveservicestaskforce.org.
- British National Formulary online — pharmaceuticalpress.com/bnf-and-bnfc/online.
- ACC/AHA ASCVD Risk Estimator Plus — tools.acc.org.
- Whelton PK et al., 2017 ACC/AHA Hypertension Guideline — ahajournals.org.
- Bibbins-Domingo K et al., USPSTF 2016 statement on aspirin for primary prevention (the prior recommendation that was updated in 2022) — acpjournals.org.
- Mahmoud AN et al., Aspirin and risk of major bleeding meta-analysis — jamanetwork.com.
- NICE — Cardiovascular disease: risk assessment and reduction (CG181) — nice.org.uk.
- Choosing Wisely — clinician lists — choosingwisely.org.