Heart Failure's 2026 Definition Drops Fixed Ejection-Fraction Cutoffs

It is late on a weeknight and someone is sitting at a kitchen table with a printout from an appointment that felt routine while it was happening. Halfway down the page is a phrase that did not come up in the room, or came up so quickly it did not register: stage B, pre-heart failure. Nothing hurts. Nothing has stopped. The heart in question is beating perfectly well, which is the part that makes the word failure so difficult to sit with at eleven at night.

So a week of headlines announcing that four cardiology organizations have redefined heart failure lands on that person with a particular weight. The redefinition is real, and it is worth understanding precisely, because what changed is the vocabulary clinicians use to sort patients — not anything about the heart of the person reading the page.

The source is an expert consensus document published in Circulation and released online on June 29, 2026, under the title Second Universal Definition of Heart Failure. Its authorship is joint across the American Heart Association, the American College of Cardiology, the European Society of Cardiology and the World Heart Federation, and bibliographic databases index it as a review rather than as original research.

It succeeds the 2021 collaboration among the same kinds of bodies, which the abstract credits with establishing standardized terminology and introducing stage B — pre-heart failure — for individuals at risk. The stated reason for revisiting the framework five years on is that advances in prevention, diagnosis and management, together with patient-centred perspectives, had moved past the old wording. The stated reason any of this was necessary in the first place is blunter: the authors write that previous definitional ambiguity hindered research, surveillance and prevention.

What the document is not is equally important. No one was randomized. Nothing was tested against a control. There is no effect size here, because measuring an effect was never what this document set out to do. It is a naming instrument, and the appropriate question to ask of it is whether the names are better ones, not whether they work.

The headline change is that heart failure is now classified using ejection-fraction categories — reduced, preserved and improved — rather than fixed cutoffs. Ejection fraction is the share of blood a ventricle pushes out with each beat, and for years it has functioned as the number that decided which kind of heart failure a person had. A framework organized around categories rather than a hard line concedes something clinicians have long known and charts rarely reflected: the measurement moves between scans, and a person sitting a point either side of a boundary was being sorted partly by measurement noise.

The second change is a proposed universal classification of cause. Heart failure that follows a heart attack, heart failure that follows chemotherapy and heart failure that follows a failing valve have shared a single label while sharing very little else, and a common vocabulary for cause is what lets a registry in one country be compared with a registry in another.

The third is trajectory. The document examines how heart failure moves over time, including improvement and recovery — which is what the improved category is for. It encodes the observation that a person is not permanently defined by the worst number their chart ever held.

Running underneath all three is an emphasis the abstract makes explicit: social determinants and geographic disparities affect outcomes, and the framework is meant to be usable worldwide rather than in well-resourced health systems alone.

Here the coverage and the source part company. Nearly every account of this document reports that the definition changed; the operative detail — where the categories begin and end — is not in the public abstract. The full text is available by subscription, and the paper has not been deposited in PubMed Central, so a person who has just met one of these category names on their own chart can learn what the category is called and not what qualifies a measurement for it.

That gap is worth naming rather than papering over. A number quoted from memory or lifted from secondary coverage would read as more helpful than a stated absence, and it would be exactly the kind of number that ends up in a patient's head with no source underneath it. What can be said with confidence is the shape of the change: categories in place of cutoffs, cause classified in common terms, and trajectories that include getting better.

Definitions decide who gets counted. Before stage B had a name, the population it describes — people at risk, without symptoms — was largely invisible to the machinery that funds and studies medicine: hard to enrol in a trial, hard to count in a registry, hard to write a prevention programme for. Naming a group is the step that makes it addressable, which is why the authors frame standardized terminology as the thing that unblocks prevention, early detection and global management rather than as housekeeping.

For the person at the kitchen table, the honest reading is narrower and calmer than the headlines. A label appearing on a chart is a statement about how a health system now sorts risk, and the sorting has been redrawn to be less brittle than a single percentage. What clinicians weigh next is how a category translates into monitoring intervals and follow-up in their own setting, and whether record systems and coding will adopt the new terms quickly enough for the standardization to deliver what it promises.

The desk's position is that the honest headline is not that heart failure has been redefined but that cardiology has stopped treating a single percentage as a diagnosis. Moving from fixed cutoffs to categories, and adding a category for people who got better, corrects a framework that quietly told patients their worst echocardiogram was their identity — and the fact that this is the second universal definition in five years is a sign of a field revising in public rather than one that cannot make up its mind. The document deserves less credit on its own stated terms, though: an abstract that emphasizes geographic disparities and global usability sits in front of a full text that costs money to read and is absent from PubMed Central, which puts the standard furthest from the clinicians least able to pay for it. A definition meant to standardize the world is only as universal as its distribution, and that part has not been settled by publishing it.