Chest pain: timing, not typicality

Clinical Rounds, the GPAtlas podcast. Cardiovascular. 20 minutes. Published 31 August 2026. Free to listen.

Clinical source: our own Ocean article "Chest Pain", last reviewed 26 May 2026.

Transcript

Sarah: Hello, and welcome to GPAtlas Clinical Rounds. I'm Sarah, and this is the podcast where we take one thing you'll genuinely see this week and work out what to do with it.

Ben: And I'm Ben. Today, chest pain, and the three questions that decide where the patient goes.

Sarah: Mr Osei is fifty-eight, he smokes, and he's come in because he had about forty minutes of tightness across his chest yesterday evening while he was carrying shopping in. It's gone now. He feels completely fine. He thinks it was indigestion, and he took some Gaviscon and it settled, which he offers as proof. And he's mildly embarrassed to be taking up an appointment.

Ben: Three things. Why what settled it tells you nothing. The three questions that sort every chest pain into an ambulance, a same-day, or a routine referral. And the presentations that don't look like this at all.

Sarah: So the first thing our article does, and I think it's the right frame, is tell you what your job actually is. The first task is not to reach a precise diagnosis but to decide, quickly, who needs emergency admission and who can be assessed in slower time. You are triaging. You are not diagnosing.

Ben: And what does the triage is timing, not typicality. The three questions are: is the pain present now, did it start within twelve hours, and what does the ECG show. And our article is explicit that the answers, not the typicality of the pain alone, decide between nine nine nine, same-day assessment and routine referral.

Sarah: So let me give the ladder in full, because it's the spine of the episode. Pain now, or within twelve hours, with an abnormal ECG: nine nine nine. Pain within twelve hours that's resolved with a normal ECG, or pain twelve to seventy-two hours ago: same-day assessment. And pain more than seventy-two hours ago, in a stable patient: rapid access chest pain clinic.

Ben: Mr Osei had his pain yesterday evening. So depending on the hour, he's in the twelve to seventy-two hour band, which is same-day assessment. Not a clinic letter. Not come back if it happens again. Today. And he's sitting there apologising for wasting your time.

Sarah: Timing decides it. Not how typical the story sounds.

Sarah: Now the two pieces of false reassurance, and Mr Osei has handed me one of them already. He says the antacid settled it. And our article's instruction is unambiguous: do not use the response to GTN to decide whether pain is cardiac, because it relieves oesophageal spasm too.

Ben: And the same logic runs the other way with antacids. Chest pain that settles with Gaviscon has not been proven to be reflux. Something settled, and pain that's going to stop anyway often stops around the time you take something. The test result you think you have is a coincidence with a story attached.

Sarah: The second piece of false reassurance is the ECG, and this is the one that trips people because we do it in the surgery and it feels definitive. NICE is explicit, and our article quotes it: a normal resting twelve-lead ECG does not exclude an acute coronary syndrome.

Ben: And it says what does exclude it, which is the useful half. Resolved pain and a clean ECG in a patient with a suspicious history still warrant urgent same-day assessment with troponin, because it is the trend in troponin, not the appearance of the ECG, that excludes myocardial injury. A trend needs two samples over time. You cannot produce that in a surgery, which is precisely why he has to go somewhere that can.

Sarah: And there's one more sign that feels like it should help and doesn't. Reproducible chest wall tenderness does not reliably exclude an acute coronary syndrome, because a minority of confirmed infarcts have tenderness on palpation. So pressing on it and finding it sore is not a discharge criterion, even though it feels like one.

Ben: A normal ECG excludes nothing. It is the troponin trend that does.

Sarah: Then the group that doesn't present like Mr Osei at all, which is where the misses cluster. Our article says older people, women and people with diabetes may have a myocardial infarction with little or no chest pain, and that breathlessness, fatigue, nausea, sweating or simple collapse may be the only clue.

Ben: And it draws the practical conclusion for you, which I'd underline: keep the threshold for an ECG low in these groups, because the textbook history that triggers admission may never appear. If you wait for the crushing central chest pain radiating to the jaw before you get an ECG, in those three groups you will wait through the infarct.

Sarah: Which changes what the consultation looks like. An eighty-year-old woman with two days of feeling wrung out and a bit sick, and no chest pain at all, is not obviously a cardiac presentation. And that's the point. The absence of the symptom the pathway is named after doesn't take her off the pathway.

Ben: In those three groups, the textbook history may never arrive.

Sarah: The numbers that mean nine nine nine regardless of the story. Altered consciousness. Central cyanosis. A respiratory rate over thirty. A heart rate over a hundred and thirty. Oxygen saturations below ninety-two per cent. Temperature above thirty-eight point five. A systolic below ninety or a diastolic below sixty, unless that's normal for them.

Ben: And that last caveat matters, because a systolic of eighty-eight in a small elderly woman who always runs at ninety is a different finding to the same number in Mr Osei. Physiological instability is relative to the person, and it's the only one of those thresholds with a get-out.

Sarah: Those numbers are an ambulance whatever the story is.

Sarah: So what do you actually do while you wait for the ambulance. Aspirin three hundred milligrams, chewed or dispersible, not enteric-coated, as soon as possible, unless he's clearly allergic. Then seventy-five daily. And the formulation detail is not fussiness, it's absorption speed, which is the entire point of giving it.

Ben: GTN, with a rule attached that gets forgotten in an emergency: avoid it if he's hypotensive, and never within twenty-four hours of a PDE-5 inhibitor, so sildenafil or tadalafil, because of profound hypotension. And that is a question you have to actually ask, in front of whoever is in the room with him, which is awkward and necessary.

Sarah: And oxygen, which is not routine. Only if the saturations are below ninety-four, targeting ninety-four to ninety-eight. And in COPD or anyone at risk of carbon dioxide retention, target eighty-eight to ninety-two through a twenty-eight per cent Venturi mask. Oxygen for everyone with chest pain is an old habit and it isn't harmless.

Ben: Three hundred, chewed. And ask about the PDE-5 inhibitor.

Sarah: If it turns out to be stable angina rather than an acute event, our article says something practical that's easy to defer: start secondary prevention, aspirin and a statin, and an anti-anginal, while you're arranging the clinic review. You don't wait for the clinic to start treating.

Ben: A beta-blocker like bisoprolol, or a calcium channel blocker like amlodipine, or diltiazem or verapamil, first line. And one combination never to make: never verapamil with a beta- blocker, because of the risk of severe bradycardia or asystole. That's the one to have hard-wired, because both drugs are reasonable alone and the combination is dangerous.

Sarah: And two specific situations worth knowing. Cocaine-associated chest pain: avoid beta- blockers acutely, because of unopposed alpha vasoconstriction. And if it is genuinely non- cardiac, musculoskeletal or reflux, then treat it as such with simple analgesia or an antacid or PPI trial, rather than leaving him with nothing and a vague reassurance.

Ben: Never verapamil with a beta-blocker.

Sarah: And the safety-net, which our article words for you and which applies to every patient you send home with a non-cardiac label. Anyone discharged with a non-cardiac diagnosis must know to call nine nine nine if the pain becomes severe, crushing or prolonged, or is accompanied by breathlessness, sweating or collapse.

Ben: And note what that sentence assumes: that you might be wrong. Which you might. The safety- net isn't an add-on to a confident diagnosis, it's an acknowledgement that a primary care assessment of chest pain has limits, and that the patient is the person who'll notice first if you've called it wrong.

Sarah: And on driving, since he'll ask and it's a question we're bad at answering. After an acute coronary syndrome, an ordinary licence holder stops for one week after successful angioplasty if the ejection fraction is forty per cent or above, and four weeks otherwise.

Ben: Assume you might be wrong. That is what the safety-net is for.

Sarah: Let me go back to Mr Osei, because the hardest part of that consultation isn't clinical, it's that he doesn't want to go. He feels fine. He has a plausible explanation. He came in half expecting to be told it was nothing, and now I'm saying he needs assessing today, which to him sounds like a massive overreaction to some indigestion that's already gone.

Sarah: And what I've found doesn't work is escalating the fear. Telling him this could be your heart, you could be having a heart attack, is both possibly untrue and quite likely to make him dig in. What works better is being honest about the limits of what I can do here. I can't tell from this room whether that was your heart. There's a blood test that answers it, and it has to be done twice a few hours apart, and I can't do that here.

Sarah: That's true, it isn't frightening, and it explains the ambulance or the same-day referral as a practical necessity rather than a judgement about how ill he is. It also sets up the thing I want him to understand for next time, which is that a normal examination and a normal ECG in a surgery genuinely don't settle this question.

Ben: And it's worth being clear with yourself about why that is, because it's the reason the pathway is built on timing. Troponin rises and falls over hours. A single value early is not interpretable, and the ECG can be normal throughout an evolving event. So the thing that answers the question is a trend, and a trend is a property of time, not of clinical skill.

Sarah: Which reframes what you're doing when you refer. You aren't admitting defeat, and you aren't being defensive. You're recognising that this specific question has an answer, and that the answer requires a resource you don't have in the building. That's a much more comfortable position to consult from than trying to be clever about typicality.

Ben: The answer is a trend over hours. You cannot produce one in a surgery.

Sarah: And one last thing about the group we mentioned earlier, because it's the one I most want people to change. Our article says older people, women and people with diabetes may have a myocardial infarction with little or no chest pain. And the practical consequence is that the trigger for an ECG in those groups cannot be chest pain, because chest pain may never arrive.

Sarah: So the trigger has to be something vaguer, and that's uncomfortable. Breathlessness, fatigue, nausea, sweating, a collapse. Which are also the symptoms of about forty other things, most of them benign. Nobody can ECG every tired person. But in those three groups specifically, our article says to keep the threshold low, and that's the instruction to act on.

Ben: In those three, the trigger cannot be chest pain. It may not come.

Ben: Let me lay the ladder out once more, because if you take one thing from this it should be reproducible from memory at four in the afternoon. Three questions. Is the pain present now. Did it start within twelve hours. What does the ECG show.

Ben: Pain now, or within twelve hours, with an abnormal ECG: nine nine nine. Pain within twelve hours, resolved, with a normal ECG: same-day. Pain twelve to seventy-two hours ago: same- day. Pain more than seventy-two hours ago, stable: rapid access chest pain clinic. Four outcomes, decided by two facts and a tracing.

Ben: And notice what isn't in that ladder. Whether it was crushing. Whether it radiated. Whether it came on with exertion. Whether the antacid helped. All of that is the history you'll take anyway, and none of it is what sorts him. Our article's phrase is that the answers, not the typicality of the pain alone, decide between the three destinations.

Ben: Which is a genuinely useful thing to internalise, because typicality is where clinical judgement feels like it should live, and it's also where the misses come from. The atypical presentations we talked about are precisely the ones where a typicality-based system fails, and the timing-based one still works.

Sarah: Two facts and a tracing. Not how typical it sounded.

Ben: Exam corner. A fifty-eight year old smoker describes forty minutes of central chest tightness eighteen hours ago, which has fully resolved. He is pain-free and haemodynamically normal, and his ECG in the surgery is normal. He took an antacid at the time and the pain settled. What is the most appropriate action? A, reassure and treat as reflux. B, refer routinely to a rapid access chest pain clinic. C, arrange same-day assessment. Or D, arrange an exercise ECG.

Ben: It's C. His pain was twelve to seventy-two hours ago, which is the same-day assessment band, and a normal ECG does not exclude an acute coronary syndrome because it's the troponin trend that does that. A is the trap the article warns about twice over: the response to an antacid or to GTN must not be used to decide whether pain is cardiac. B is the right destination only for pain more than seventy-two hours ago in a stable patient. And D isn't the pathway here at all.

Sarah: So. Three things for Monday. One. Ask when it happened before you ask what it felt like, because timing sorts the patient and typicality doesn't. Two. Nothing that settled the pain is evidence about its cause, whether that's an antacid or GTN. And three. In older people, women and people with diabetes, get the ECG early, because the history that would have triggered you may never appear.

Sarah: One thing to reflect on, if you're logging this. Think of the last chest pain you sent home. Did you give them the specific words, severe, crushing or prolonged, breathlessness, sweating, collapse, or did you say come back if you're worried? The full transcript and the references are on the episode page. Ben and I are synthetic voices. The medicine isn't. See you next week.

More from Clinical Rounds

Every episode of Clinical Rounds, free to listen with a full transcript.