The approach
Prevention, practiced as a discipline.
Standard care waits for the event. The Ethos approach works the decades before it: in three acts. Measure the actual causes of heart attacks and strokes, not the proxies that are easy to bill. Interpret the results honestly, in absolute terms. Then accompany members across the years that prevention actually takes.
I · Measure
Measure the causes,
not the conveniences.
The standard panel measures what is easy to bill. The proposed Ethos assessment is built around the mechanisms that actually produce events: particle number, inherited lipid risk, arterial inflammation, plaque itself, and the metabolic substrate beneath them.
LDL-C measures the cargo. ApoB counts the particles.
LDL-C estimates the cholesterol mass carried inside lipoprotein particles; ApoB counts the particles themselves, and it is particles that enter the artery wall. When cargo and count disagree, the particle number is the better guide. A routine panel never measures it. The proposed assessment does.
Lp(a): inherited, common, and invisible to routine panels.
Lipoprotein(a) is set almost entirely by genetics and appears on no standard cholesterol test. In the BioHEART-CT cohort (1,718 patients undergoing CT angiography), adding Lp(a) reclassified cardiovascular risk, with a net reclassification index of 16%. One measurement, once in a lifetime, closes the gap.
After cholesterol, the inflammation that remains.
Even with lipids controlled, residual arterial inflammation continues to drive events. In CANTOS, anti-inflammatory therapy significantly reduced major adverse cardiovascular events, with no change in lipids at all. The proposed assessment measures that axis with hs-CRP rather than assuming it away.
Risk calculators estimate. A calcium scan shows.
Risk equations guess from averages. In MESA, coronary artery calcium delivered the largest risk reclassification of any marker tested, and adding a calcium score to standard estimates reclassified 40–41% of people, including 24% of those advised to take a statin who proved to have no detectable calcium. When your profile calls for it, the scan tells both directions of the truth: who should treat, and who can safely wait.
Waist and visceral fat, measured the same way every time.
The scale cannot distinguish fat under the skin from fat around the organs, and it is the second kind this practice watches. Waist circumference and visceral adiposity are measured at every visit, by the same trained measurer, with the same instrument and the same protocol. A number taken the same way for years becomes a trend you can act on, rather than noise you argue about.
“No test is ordered unless the result can change what we do.”
Every order answers to that rule first. Final selection of tests is always an individual clinical decision, made with you.
II · Interpret
Both numbers,
always.
Relative risk is the language of headlines; absolute risk is the language of decisions. A “36% relative reduction” can be 1.1% in absolute terms. Both statements are true, and only one of them tells you what a treatment is likely to do for you.
The relative reduction you will be quoted
The absolute reduction it can actually mean
So both numbers are always shown: the relative and the absolute, the benefit and the number-needed-to-treat. The statin example is instructive: in secondary prevention the number-needed-to-treat is 63 per year; in primary prevention, 217. Same drug, same relative effect: a different absolute answer, because the starting risk differs. That is why we measure first.
The evidence is reported to you the same way it appears in the literature. In the CTT meta-analyses, each 1 mmol/L of LDL-C lowering carries an ≈20–25% relative reduction in major vascular events. What that means in absolute terms depends on your measured baseline, which is the whole point of the first act.
III · Accompany
Prevention is a relationship
measured in decades.
None of this works as a one-time transaction. The proposed model is built around the long arc (the same physician, a defined schedule, and guaranteed windows), so the person interpreting your trend is the person who has watched it form.
Your own physician, nearly every time.
The proposed model holds a continuity index of at least 0.75 (you see your own physician for at least 75% of visits) because a baseline only means something to the person who knows it.
Every changed number earns a second look.
When a treatment changes, the proposed model rechecks the number it was meant to move within 6–8 weeks. A prescription without a recheck is a hypothesis left untested.
When something changes, the door is open.
The proposed model holds routine appointments within 48 hours and urgent concerns the same day, with guaranteed response windows, because a question about your heart should not wait three weeks for a slot.
Thirty minutes or longer, and a panel small enough to know you.
The proposed model schedules preventive visits at 30 minutes or longer and caps panels at 600–800 members per physician, against panels exceeding 2,500 in conventional primary care. Prevention does not survive fifteen-minute increments.
Membership
The discipline becomes a membership.
Panels, visit length, access windows, and what membership includes: the proposed model, stated plainly.
References · this page
Every figure, sourced.
- BioHEART-CT: 1,718 patients undergoing coronary CT angiography; adding Lp(a) yielded a net reclassification index of 16%.
- MESA: coronary artery calcium delivered the largest risk reclassification among tested markers; adding CAC to standard risk scores reclassified 40–41% of individuals (24% of those recommended for statins had CAC = 0).
- CANTOS (Ridker, 2017): anti-inflammatory therapy significantly reduced major adverse cardiovascular events, with no change in lipids.
- CTT meta-analyses: ≈20–25% relative reduction in major vascular events per 1 mmol/L of LDL-C lowering.
- The Ethos model (panel caps, visit length, continuity targets, recheck intervals, and access windows) is a proposed practice design, pending launch. Figures describe the design, not an operating history.