Many years ago, when I was working as a registrar in Emergency Medicine and starting my PhD, my supervisor (a social scientist) got me to think about the assumptions I was making in my research. “Is there such a thing as a myocardial infarction?”, he asked. I instinctively replied with something along the lines of, “Well, of course!” We can see that people develop coronary atheroma, that plaques rupture, that this makes blood clot, which blocks arteries, which causes injury to the myocardium. That can stop the heart from beating as well and it can cause sudden death by lethal arrhythmias or even rupture of the myocardium itself. It’s kind of obvious that myocardial infarction exists. We have lots of evidence.
But, as I often do, I went home and over-thought this question, and I still find myself asking it to this day. OK, so let’s say that there is a thing called a ‘myocardial infarction’. If that’s so, then what is it? And that’s a harder question than you might think. The more we think about it, the more we realise that ‘myocardial infarction’ is a social construct used to conveniently put similar patients into groups rather than being a single absolute reality.
So why do we need a definition? Well, a definition helps us to standardise what we do. This is important for health equity, for understanding patterns of disease, and for working out what treatments will work best for specific groups of patients. That’s where the idea of a ‘universal definition of myocardial infarction’ came in. Last week, the fifth universal definition was released, and its publication was announced at the European Society of Cardiology Congress, led by first author (and a great friend of mine) Nick Mills – Professor of Cardiology in Edinburgh.
Let’s go through what this means for our practice, and what’s new. You can also catch my summary of this by video…
THE BOTTOM LINE
In the previous (fourth) universal definition of myocardial infarction (MI), we grouped MIs into five types: type 1, type 2, type 3, type 4 (further divided into a, b and c) and type 5. That was quite complicated. On the face of it, the new (fifth) universal definition simplifies all of this (which I think is very welcome), as follows…
PRIMARY MYOCARDIAL INFARCTION
Primary MI is the equivalent of the old ‘type 1’ MI – essentially an MI that occurs spontaneously due to coronary disease – but there are some important differences. With the previous definition, only cases with atherothrombosis were called ‘type 1’. Now, any coronary pathology is to be labelled as a ‘primary MI’. This includes SCAD, coronary vasospasm, coronary artery dissection, coronary embolism, and stent thrombosis or coronary artery bypass graft failure – providing that the failure has occurred more than 30 days after the procedure.
SECONDARY MYOCARDIAL INFARCTION
This is the equivalent of the old ‘type 2’ MI, with the exception that the coronary causes I just mentioned (SCAD, etc.) have moved out of this group. ‘Secondary myocardial infarction’ now only includes cases where there is an imbalance between oxygen supply and demand to the myocardium because of another condition (e.g., sepsis; gastrointestinal haemorrhage; dysrhythmia). But there’s another caveat: we now also need evidence of obstructive coronary artery disease (but without acute coronary pathology like a ruptured plaque or thrombus), and/or there has to be evidence for a new (or presumed new) regional wall motion abnormality, e.g., on echo.
PROCEDURE-RELATED MI
This replaces the slightly complicated classification we previously had. It now means that anyone who has an MI identified within 30 days of a coronary procedure (PCI or CABG) will be classed as having ‘procedure related MI’, replacing the old type 4 and 5 MIs.
WHERE DID TYPE 3 MI GO?
Type 3 MI was that caused by sudden cardiac death, and it was felt to be redundant. MIs causing sudden cardiac death can now be placed into one of the above three categories.
So, for most clinicians this makes things more straight forward. We have three options: primary MI (caused by coronary pathology), secondary MI (caused by something else putting strain on the heart) and procedure-related MI, which is exactly what it says (providing the procedure was within the last 30 days). So let’s take a look at what else is new…
SEX-SPECIFIC TROPONIN CUTOFFS ARE OFFICIALLY IN
On average, women have lower troponin concentrations than men. All troponin test manufacturers therefore now derive separate reference ranges for men and women. The fifth universal definition now explicitly states that we should use them. For most of us, this will have no implications because we’ll already be using them. However, if your hospital is using the Roche high-sensitivity troponin T assay (gen 5), you may still be using a single cutoff (14 ng/L outside the USA, 19 ng/L in the USA). This may be because the sex-based differences are relatively small in absolute terms (17 ng/L in men; 9 ng/L in women) and the precision of the assay – at least until recently – meant that the assay would not be high-sensitivity at the upper limit of normal for women. This will inevitably change relatively soon when the Gen 6 assay is released.
This is likely to be a positive move for health equity. The under-diagnosis and under-treatment of women with MI has been widely documented. Moving to sex-specific troponin cutoffs may help to address some of the issues around health equity, though we should remember that this is only part of the issue – and always aware of the potential for our own cognitive biases (e.g., ‘what someone with a heart attack looks like’) to fully address this issue. The Guardian has given some great coverage to this very positive development in the 5th universal definition.
IMAGING IS NO LONGER OPTIONAL IF YOU WANT DIAGNOSTIC CERTAINTY
A really notable aspect of fifth universal definition is the much more stringent requirements for patients to have imaging and/or coronary angiography to confirm the diagnosis of MI. In low resource settings, there’s an acknowledgment that imaging may not be undertaken, and MI may still be diagnosed (on the balance of probability) in those cases. However, the document now repeatedly states that:
- A diagnosis of MI can be likely on clinical grounds
- But MI is confirmed through coronary and/or cardiac imaging.
For primary MI, confirmation requires either:
- Demonstration of an acute coronary pathology, or
- A new regional wall motion abnormality or new loss of viable myocardium consistent with ischaemia.
For secondary MI, imaging becomes even more important. Secondary MI is only confirmed when imaging demonstrates:
- Obstructive coronary artery disease without acute coronary pathology, and/or
- New regional wall motion abnormality
- New loss of viable myocardium.
This is important for emergency physicians, especially for secondary MI. We are used to interpreting troponin rises in the context of other conditions – sepsis, tachyarrhythmia, etc. We’re used to seeing evidence of ischaemia (e.g., chest pain, ST depression) in such patients. And we would have then assigned a diagnosis of ‘type 2 MI’. Now, we’re required to order imaging to confirm the diagnosis. This might have important implications for resource utilisation. But what should we do with the findings? Unfortunately, we still can’t be entirely clear about that because we don’t have evidence that any given treatment is effective in type 2 MI. At least, not yet – perhaps evidence will come soon!
HOW ABOUT CHRONIC MYOCARDIAL INJURY?
Another issue that will be very familiar to emergency physicians is chronic myocardial injury. Many patients will have an elevated troponin when we measure it. We’re used to repeating the measurement and labelling the myocardial injury as chronic if there is no significant change to demonstrate a rise and/or fall. This especially affects older adults and those with chronic kidney disease. So what does the 5th universal definition change?
First, let’s look at what they say about ‘delta’ values. How do we decide if there has been a significant change in troponin on serial sampling? Many studies have shown that ‘absolute’ changes (maximum value – minimum value) are better than relative (percentage) changes in troponin for deciding who has MI. But does the fifth definition tell us which to use? Actually, it tells us that this is a nuanced area and we need to use our clinical judgement. At lower troponin values, an absolute change may be better. At higher troponin values, a relative change may be better. Early after symptom onset, we might see a big change on serial sampling. Later (12-24h) the change might be minimal. The new definition therefore suggests no single value, and leaves the judgement up to clinicians. The timing of serial samples will also affect the delta. The new definition overtly recommends 1 to 2 hours between samples for this reason, perhaps because cutoffs are likely to have been validated with those timings, though by following that logic we might expect that validated 0/3h delta criteria for any given troponin assay would also still be acceptable.
WHAT ELSE IS IMPORTANT FOR CLINICIANS TO KNOW?
There are two more things you might be wondering.
First, what about risk scores and clinical decision aids like the HEART score, T-MACS, EDACS, and MI-3? They aren’t in there. I think they’re out of scope – so the definition seemingly doesn’t affect their use.
Second, does the new definition move us away from the traditional concepts of STEMI and NSTEMI towards OMI (occlusive MI) and non-OMI? No, not quite. The traditional terms are retained, but there is a nod to OMI patterns on the ECG. The definition specifically mentions patterns like Wellens, de Winter T waves, posterior MI, Aslanger pattern, and Sgarbossa criteria for left bundle branch block.
IN SUMMARY
I think the key messages for emergency physicians are as follows:
- Think primary, secondary and procedure-related MI, not Type 1 and Type 2.
- Use sex-specific troponin thresholds whenever available.
- Understand that both primary and secondary MI now require objective evidence, usually from imaging.
- Chronic myocardial injury is still suspected by seeing elevated troponin levels without a significant change on serial sampling, but establishing whether there is a change is left to clinical judgement.
- Chronic myocardial injury is confirmed by further investigations (likely imaging), though as it isn’t clear what we should do about it that may not be universally done.
- We’ve moved closer to an OMI-based understanding of coronary occlusion, but the STEMI/NSTEMI terminology remain
And remember that the guidance is still very clear that troponin elevation alone is myocardial injury, not myocardial infarction.
Rick
Comment by Nick Mills (first author of the 5th universal definition of myocardial infarction) / September 9, 2026
Excellent summary Rick. You have nicely described the key changes. One of the main reasons for the update was to make it easier to apply consistently in practice. Accurate and early diagnosis of myocardial infarction can improve outcomes through prompt and effective treatment, but the effectiveness of treatment depends on the cause and the setting in which myocardial infarction arises. The previous definition recognised this, but was not intuitive or aligned with the way that we evaluate patients with possible myocardial infarction in practice.
Where someone presenting with the spontaneous onset of symptoms or signs of myocardial ischemia is found to have acute myocardial injury the working diagnosis is acute myocardial infarction and the assumption is that this is due to atherothrombosis. Indeed, treatment in the Emergency Department is often initiated on this basis. However, without performing coronary angiography we cannot determine the cause or tailor the treatment effectively. It therefore made more sense to include all primary acute coronary pathologies within the same classification. Encouraging cardiologists to define the aetiology in the final diagnosis (e.g. primary myocardial infarction due to atherothrombosis, spontaneous coronary artery dissection or vasospasm), will improve understanding for patients and communication with other specialists that will help with their future assessment and care. It also recognises that all acute coronary causes of myocardial infarction are equally important, including those that are more likely to be experienced by women, such as coronary artery dissection or vasospasm.
Invasive coronary angiography may not be available or the risks of the procedure for some patients may be considered prohibitive. That does not prevent a clinical diagnosis of primary myocardial infarction, but we need to recognise that the diagnosis is less certain and we cannot further classify the diagnosis according to the underlying cause.
Secondary myocardial infarction is now only considered as a working diagnosis in patients presenting with an alternative acute illness resulting in myocardial ischemia due to supply demand imbalance. Where this is due to obstructive coronary artery disease or the extent of myocardial injury is substantial enough to cause new ventricular impairment then the diagnosis of secondary myocardial infarction is confirmed and will have direct treatment implications. The priority will always be to manage the acute presenting condition, and the timing and need for cardiac investigation always needs to be carefully considered, but hopefully the introduction of more objective diagnostic criteria here will reduce uncertainty.
Do you want to get involved?
Consensus documents cannot improve care or outcomes though. For that we are going to need help to implement this in practice, and further research to help us develop more effective care pathways for patients with possible myocardial infarction. If you are interested in getting involved then we are starting a new programme called ADOPT-UDMI, which aims to promote adoption of the new classification of myocardial infarction through education, evaluation and research. We are looking for clinicians from all specialties to join as hospital, country or regional leads. There are many ways to help from just promoting the new definition or ADOPT-UDMI programme through social media channels, giving an educational presentation, or participating in a global audit or research study. Please email us at [email protected] or complete this short form to register interest – ADOPT-UDMI get involved [forms.cloud.microsoft].
