Beyond the Calcium Score: What a Modern Heart Scan Can Actually See

One of the patients I think about most often did everything right.

He was in his mid-forties, lean, at his ideal weight, the kind of person who exercised six days a week and did not miss. No family history of heart disease. At every annual physical for years he had been told he was in perfect health, and on paper he was.

We ordered a coronary calcium score, a quick, inexpensive scan that looks for calcified plaque in the arteries of the heart. His was not zero. A CT angiogram then confirmed what the number hinted at: significant coronary artery disease, in a man who by every lifestyle measure should have had none. His full story is written up separately, as the clean bill of health that wasn't.

I open with him because he is the reason this article exists. A heart scan can see things a physical exam cannot. But a number is where the story starts, not where it ends. What follows is an honest account of what a calcium score gets right, where it misleads, and what the newer scans can and cannot add.

What the calcium score gets right

Let me start with the case in favor, because it is strong.

A coronary artery calcium score, often called a CAC score or an Agatston score, is one of the genuinely useful tests in preventive cardiology. For a middle-aged adult whose risk sits in the murky intermediate zone, the place where a risk calculator says "maybe" and a physician is honestly unsure whether to start a statin, the calcium score is the best tie-breaker we have. In the MESA cohort it reclassified a large share of intermediate-risk adults, moving many of them up or down into a category where the right decision became clear.

It is also honest about time. A true zero score buys you what cardiologists call a warranty period, a window during which your risk of a heart event is genuinely low. Depending on your other risk factors that window runs roughly 3 to 7 years, and it is shorter if you have diabetes. That is real, actionable reassurance, with an expiration date attached.

And it is cheap and concrete. A calcium score typically costs somewhere between $100 and $400 out of pocket, takes a few minutes, and delivers a radiation dose of about 1 millisievert, roughly what you get from a few months of ordinary background radiation. For the right person, a 45 to 70 year old weighing a statin, that is a great deal of decision-making value for very little.

So when the test fits, I use it, and I trust it. The trouble starts when people ask it to do more than it can.

Where it misleads

The calcium score has three blind spots that matter.

The first is the one that caught my patient. Calcium is a late-stage feature of coronary disease. Plaque begins soft, as a lipid-rich deposit in the artery wall, and only later, sometimes years later, does the body wall it off with calcium. That means a younger person can have dangerous, unstable, non-calcified plaque and still score exactly zero. The score does not measure disease. It measures one late chapter of it, and it lags the story.

The second is a genuine paradox that trips up smart patients. Statins raise your calcium score. This sounds alarming and is almost always the opposite of alarming. Part of it is arithmetic: the Agatston formula weights denser calcium more heavily, and statins make plaque denser. Part of it is biology: statins help convert soft, rupture-prone plaque into stable, calcified plaque. So a rising score in someone already on treatment may reflect plaque stabilizing rather than disease progressing. I say "may," not "does," because a rising number can also be real progression, and the score alone cannot cleanly tell the two apart. That ambiguity is exactly why chasing the number over time is a trap.

Which is the third blind spot: repeat scanning is mostly noise. For most people a second and third calcium score rarely change what we do, the densification effect muddies the comparison, and guidelines do not endorse routine re-scanning. One good score, interpreted well, is usually worth more than a series of them watched anxiously.

What comes after the score

When the score or the overall picture warrants a closer look, the imaging has moved well beyond a calcium number.

A coronary CT angiogram, or CTCA, uses contrast dye to image the artery wall itself, which means it can see the soft, non-calcified plaque a calcium score misses entirely. This is not a fringe test. In the SCOT-HEART trial, patients with stable chest pain who were managed with the help of a CTCA had about 41 percent fewer heart attacks and coronary deaths at 5 years than those managed the standard way, and the benefit held at 10 years.

I want to be careful here, because this is where honest and promotional part ways. SCOT-HEART is the most favorable large trial of CT angiography, and it is not the whole story. Other major trials are more sober. PROMISE found no outcome difference between CT angiography and standard functional testing. DISCHARGE found CT angiography merely as good as invasive angiography, not better. Even in SCOT-HEART, much of the benefit is credited not to the pictures themselves but to the fact that seeing plaque prompted doctors to start more patients on preventive therapy. The scan helped mostly by changing what happened next. That is the honest way to hold it: a CT angiogram sees what is truly in your arteries better than a calcium score, even where the outcome trials are mixed, and the value lies in what you do with the picture.

On top of the CTCA sits a newer layer that gets the most breathless coverage: artificial intelligence that analyzes the same images, with no additional scan and no additional radiation.

The first kind is AI plaque quantification, sometimes called AI-QCT (one FDA-cleared example is Cleerly). It measures the amount, type, and location of plaque across the whole coronary tree, far more precisely than a human reader eyeballing the scan. In the CONFIRM2 registry, adding these measurements improved the prediction of future cardiac events, though how much depends on the comparison. Against a basic clinical risk model the improvement looked large, a standard measure of discrimination rising from about 0.63 to 0.76; against the trained reading physicians already use, the gain was modest, from roughly 0.79 to 0.81. The tool adds real information, but it does not transform the picture, and you should be wary of anyone who quotes only the flattering number.

The second kind is CT-derived fractional flow reserve, or FFR-CT (one example is HeartFlow). Instead of measuring plaque, it estimates whether a given narrowing is actually choking off blood flow, the kind of question that used to require threading a wire into the artery. When it is reassuring, it can safely spare a patient an invasive catheterization.

The honest reframe is this. A plain calcium score is now table stakes. The frontier is imaging the whole vessel wall and its function. That frontier is real. It is also where the evidence starts to thin out.

The honest catch

Everything I just described is more informative than a calcium score. More informative is not the same as proven to help you live longer, and the gap between those two things is the whole ballgame.

In late 2025 the American College of Cardiology published a scientific statement on exactly these AI plaque tools. Its conclusion, from the field's own experts, was cautious: there is not yet sufficient evidence that acting on AI plaque measurements improves how patients actually do, and there is no consensus yet on when or how to use them. This is the same lesson as the biomarker hype I have written about before. A richer description of your arteries is not automatically a better outcome. It can just as easily mean more incidental findings, more anxiety, and more cost, without a proven change in whether you have a heart attack.

There is a cost and access reality too. These advanced scans and analyses are expensive, inconsistently covered by insurance, and unevenly available.

One more thing, in the interest of straight dealing. My practice has no financial relationship with any of the companies named here. We do sometimes order these tests when they are the right tool for a specific patient, but we hold no equity, no referral arrangement, and no promotional tie to any of them. I name the brands only because you will encounter them, and you deserve to know who is talking.

When the scan earns its place

So when is a closer look genuinely worth it, evidence gap and all? For the right person, and for reasons that have little to do with promising a longer life.

The first reason is the blind spot we already met. A calcium score of zero is not the same as no plaque. A younger patient with a strong family history or a genetic marker like elevated Lp(a) can carry dangerous soft plaque while scoring a perfect zero, and a CT angiogram can find it. In studies of patients with chest pain, a real minority with a zero score still had plaque the calcium scan missed; the yield is lower in someone healthy being screened, so this is a reason to look closer in the right patient, not a reason for everyone with a zero to worry.

The second reason is quieter. A scan puts your cholesterol, your genetics, and your family history into the context of your own anatomy rather than a population average, and seeing your own arteries turns out to be one of the few things that reliably changes behavior. In randomized trials, people shown images of their own silent atherosclerosis went on to improve their cholesterol, their blood pressure, and how faithfully they took their medication. For someone stuck in the ambiguous middle of the risk spectrum, honestly unsure whether to start a statin, a picture that is both truer and more motivating can make that decision clearer and easier to keep.

That is a good reason to look closer. It is simply a different reason than the promise that the scan itself will make you live longer, and the two should never be confused.

Common questions

Should I get a coronary calcium score?

For many people in their forties to seventies whose heart risk is genuinely uncertain, and who are trying to decide whether to start a statin, yes, it is one of the most useful tests available. It is less useful for the very young, whose plaque has not calcified yet, and for people already on a statin for known disease, where the number is harder to interpret. The right answer depends on your specific risk picture, which is a conversation worth having with a physician.

What does a score of zero mean, and does a high score mean I will have a heart attack?

A zero score is good news with a time limit. It suggests your near-term risk is low for a window of a few years, but it is not a permanent guarantee, and it does not rule out soft plaque that has not calcified. A high score raises your risk but is not a verdict. It is a reason to treat aggressively and to look closer, not a prediction that an event is coming. A number is a measure of risk, not fate.

Calcium score versus CT angiogram, what is the difference?

A calcium score is a number: it counts calcified plaque and nothing else. A CT angiogram is a picture: it uses contrast dye to show the whole artery wall, including the soft, non-calcified plaque a calcium score cannot see, and how much any narrowing restricts blood flow. The calcium score is the cheap, fast screen. The angiogram is the detailed look you order when the screen or the clinical picture calls for it.

How much does a calcium score cost?

A calcium score is typically around $100 to $400 out of pocket, and it is often not covered by insurance. A CT angiogram costs more. The AI analyses layered on top add further cost and are the least consistently covered of all.

How much radiation is involved?

A calcium score delivers about 1 millisievert, roughly a few months of ordinary background radiation. A modern coronary CT angiogram is higher but still modest, on the order of a few millisieverts on current low-dose scanners, with one large real-world registry finding a median near 3 to 4. The AI analysis adds no radiation at all, because it reruns on images you already have.

What are Cleerly and HeartFlow?

They are examples of two different kinds of AI analysis applied to a coronary CT angiogram. One kind quantifies plaque in detail across the arteries. The other estimates whether a narrowing is actually limiting blood flow, which can help avoid an invasive test. Both are genuinely useful in the right hands. Neither has yet been proven to change long-term outcomes, which is why interpretation matters more than access.

Why this is a concierge conversation

None of this fits into a 15-minute visit.

Deciding whether you need a heart scan at all, choosing which one, and then reading the result without over- or under-reacting takes time, and takes a physician who follows the primary literature rather than the press release. The value my patients get is not a longer menu of tests. It is judgment: knowing that a calcium score is the right first move for one person and a waste for another, that a nonzero score in a fit 45 year old means look harder rather than panic, and that a glossy AI plaque report is informative without being the last word.

That is the work I find most rewarding, and it is the work that caught my patient's clean bill of health for the mistake it was. His number was not zero. Someone took that seriously, looked closer, and changed the course of his life before anything happened. That is what these tools are for, in the hands of someone paid to interpret them for you rather than to sell them to you.

If you want to think through what advanced cardiac screening would and would not do for you, our Executive Health evaluation is built around exactly this kind of question, and you can always contact us to start the conversation.

Paul Abramson, MD is the founder of My Doctor Medical Group, a concierge medical practice in San Francisco. He trained as an engineer at Stanford before completing medical school and residency at Stanford and UCSF, a combination that shapes his approach to medicine: rigorous about evidence, skeptical of complexity for its own sake, and focused on what actually works.

This content is for educational purposes only and does not constitute medical advice. Consult your physician before making changes to your health regimen.

Sources

• Polonsky TS, et al. Coronary artery calcium score and risk classification for coronary heart disease prediction. JAMA, 2010;303(16):1610-1616 (MESA). PMID 20424251.
• Nasir K, et al. Implications of coronary artery calcium testing among statin candidates. Journal of the American College of Cardiology, 2015.
• Dzaye O, et al. Warranty period of a coronary artery calcium score of zero. JACC: Cardiovascular Imaging, 2021. DOI: 10.1016/j.jcmg.2020.06.048.
• Puri R, et al. Impact of statins on serial coronary calcification during atheroma progression and regression. Journal of the American College of Cardiology, 2015;65(13):1273-1282. DOI: 10.1016/j.jacc.2015.01.036.
• Lee SE, et al. Effects of statins on coronary atherosclerotic plaque: the PARADIGM study. JACC: Cardiovascular Imaging, 2018. DOI: 10.1016/j.jcmg.2018.04.015.
• SCOT-HEART Investigators. Coronary CT angiography and 5-year risk of myocardial infarction. New England Journal of Medicine, 2018;379:924-933. DOI: 10.1056/NEJMoa1805971. Ten-year follow-up: The Lancet, 2024.
• Douglas PS, et al. Outcomes of anatomical versus functional testing for coronary artery disease (PROMISE). New England Journal of Medicine, 2015.
• DISCHARGE Trial Group. CT or invasive coronary angiography in stable chest pain. New England Journal of Medicine, 2022. DOI: 10.1056/NEJMoa2200963.
• Villines TC, et al. Prevalence and severity of coronary artery disease and adverse events among symptomatic patients with coronary artery calcium scores of zero (CONFIRM registry). Journal of the American College of Cardiology, 2011;58(24):2533-2540. PMID 22079127.
• Rozanski A, et al. Impact of coronary artery calcium scanning on coronary risk factors and downstream testing (EISNER randomized trial). Journal of the American College of Cardiology, 2011;57(15):1622-1632. PMID 21439754.
• Naslund U, et al. Visualization of asymptomatic atherosclerotic disease for optimum cardiovascular prevention (VIPVIZA): a pragmatic, open-label, randomised controlled trial. The Lancet, 2019;393:133-142. PMID 30522919.
• CONFIRM2 Registry. AI-driven quantitative coronary CT angiography in suspected coronary artery disease. JACC: Advances, 2026. DOI: 10.1016/j.jacadv.2026.102618.
• Quantitative coronary plaque analysis in clinical practice: 2025 ACC Scientific Statement. JACC: Cardiovascular Imaging, 2025. DOI: 10.1016/j.jcmg.2025.11.008.
• Stocker TJ, et al. Reduction in radiation exposure in cardiovascular computed tomography imaging (PROTECTION VI). European Heart Journal, 2018;39(41):3715-3728. PMID 30165629.
• NICE Medical Technologies Guidance MTG32: HeartFlow FFRCT for estimating fractional flow reserve from coronary CT angiography. 2017, updated 2021.