Heart failure: under-treated, not optimised

Clinical Rounds, the GPAtlas podcast. Cardiovascular. 19 minutes. Published 31 August 2026. Free to listen.

Clinical source: our own Ocean article "Chronic Heart Failure (CHF)", last reviewed 23 August 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 together what to do about it.

Ben: And I'm Ben. I'm the one who's been through the guidance, so when Sarah tells you what she did, I'll tell you whether the evidence agrees with her. Today, the heart failure patient who looks stable, and is quietly under-treated.

Sarah: So. Seventy-one year old man, heart failure with reduced ejection fraction, diagnosed three years ago. He's on forty milligrams of furosemide, two point five of ramipril, and one point two five of bisoprolol. And he's fine. He says he's fine. His ankles are down, he's not breathless at rest, and every review for three years has said stable, continue. And when I actually looked at it properly, what I had was a man on a big dose of the drug that makes him feel better and a token dose of the drugs that keep him alive.

Ben: Three things today. Why that patient is under-treated rather than optimised, which is our article's phrase and it's the right one. The four pillars and what each actually does. And a genuinely new option for the other kind of heart failure, the preserved one, where until recently we had almost nothing.

Sarah: So, the four pillars. An ACE inhibitor, or an ARNI. A beta-blocker licensed for heart failure. An MRA. And an SGLT2 inhibitor. And the thing to hold onto is that each of those four independently reduces mortality. Independently. They are not alternatives and they are not a ladder you climb only when things get worse.

Ben: And set against that, the loop diuretic. Furosemide relieves symptoms but does not prolong life. That is the whole tension in Sarah's patient. The drug on the biggest dose is the one doing nothing for his survival, and the four that are doing the work are on starting doses. Our article puts it exactly: the mortality benefit comes from establishing and up- titrating all four pillars, and a patient parked on a large dose of diuretic with sub- target doses of disease-modifying therapy is under-treated, not optimised. Start them early and titrate to the maximum tolerated dose. Not simply started and left.

Sarah: So that's the first one. Four pillars, each one independently saving lives, and a diuretic that saves none of them. If the diuretic is the biggest dose on the list, look again.

Sarah: Now the practical bit, because titrating up is where nerve fails. The ACE inhibitor goes up roughly every two to four weeks towards target, so ramipril heading for ten milligrams a day, and you watch renal function and potassium. The beta-blocker is bisoprolol, carvedilol or nebivolol. And here's the line I want to read out properly, because it stops so many up-titrations. Do not withhold a beta-blocker on the grounds of age, COPD, peripheral vascular disease, erectile dysfunction or diabetes. None of those is a reason.

Ben: And the monitoring is specific, so it's worth having it exactly. Check renal function and electrolytes before you start an ACE inhibitor, ARNI, ARB or MRA. Then one to two weeks after starting. Then one to two weeks after each dose increase. Then every three to six months once you're at the maximum tolerated dose. And whenever renal function might be compromised. The numbers that make you stop and think are a creatinine rise of more than fifty per cent, or a potassium above five point five.

Sarah: And one that's purely ours, and free. Daily weights. A sudden rise of more than one and a half to two kilos over two days means fluid is coming back, and that's the moment to adjust the diuretic, before he's breathless and before anybody calls an ambulance.

Sarah: Titrate every two to four weeks. Bloods before, after, and after every increase. And a set of scales at home beats a hospital admission.

Ben: Now, the other heart failure, the preserved ejection fraction one, where the honest position for years has been that we had very little. That has changed. Finerenone, a non- steroidal MRA, is now recommended by NICE for symptomatic heart failure with preserved or mildly reduced ejection fraction, so an LVEF of forty per cent or above. Our article calls it only the second disease-modifying option in that group after SGLT2 inhibitors. It is specialist-initiated, with primary care monitoring under shared care, so it will arrive on your repeats.

Sarah: And because it lands on our repeats, the numbers are ours to know. Starting dose depends on the kidneys: twenty milligrams if the eGFR is sixty or above, ten if it's between twenty-five and fifty-nine. And the target is set by the eGFR at the start, not by everybody reaching the same dose. Recheck potassium and eGFR at four weeks and before each step up. And do not start it at all if the potassium is above five point oh or the eGFR is below twenty-five.

Sarah: Preserved ejection fraction finally has a second option. Specialist starts it, we monitor it, and the main risk is the potassium.

Ben: Let's do the diagnosis properly, because the NT-proBNP sets two things at once: whether this is heart failure, and how fast the clock is running. Below four hundred, untreated, makes heart failure unlikely. Four hundred to two thousand means specialist assessment and an echocardiogram within six weeks. Above two thousand means within two weeks, and our article notes that very high levels carry a poor prognosis. So the number isn't just positive or negative. It's a referral timescale.

Ben: And the thing it cannot do is tell you which heart failure you're dealing with. NT-proBNP does not differentiate between reduced, mildly reduced and preserved ejection fraction. Only the echo assigns the type. Which matters because everything we're about to say about treatment depends on that answer, and you can't shortcut to it from the blood test.

Sarah: And here's the one that catches people, and it caught me. A normal NT-proBNP does not always exclude heart failure. It's pushed down by obesity, by African or African-Caribbean family background, and by the very drugs we use to treat heart failure: diuretics, ACE inhibitors, ARBs, ARNIs, beta-blockers and MRAs. So the breathless, obese patient already on furosemide, with a level of three hundred and eighty, is exactly the patient in whom that number means the least. Our article says if clinical suspicion is high, discuss with the heart failure team rather than stopping there.

Ben: Two thousand means two weeks. And a normal level in an obese, treated patient proves nothing.

Sarah: Now, preserved ejection fraction, because this is where practice has genuinely changed and where a lot of us are still saying there's nothing much to offer. That is out of date. HFpEF gets an SGLT2 inhibitor and finerenone, a non-steroidal MRA, alongside a loop diuretic for congestion and proper treatment of the blood pressure, the atrial fibrillation, the obesity and the diabetes. So the honest framing now is that preserved ejection fraction has disease-modifying treatment, not just symptom control.

Ben: With two hard numbers on the finerenone. Do not start it if the potassium is above five point zero, or the eGFR is below twenty-five. And because it's a non-steroidal MRA, it's used instead of a steroidal one, so instead of spironolactone or eplerenone, not alongside. That last point is the one that will generate a serious hyperkalaemia if it goes wrong.

Sarah: Preserved ejection fraction is not nothing to offer any more.

Sarah: Monitoring, and I want to give the schedule in full because it's the thing most likely to be done badly in a busy practice. Renal function and electrolytes before you start an ACE inhibitor, ARNI, ARB or MRA. Again one to two weeks after starting. Again one to two weeks after every dose increase. Then every three to six months once you're at the maximum tolerated dose. And any time renal function might be compromised, so any acute illness.

Ben: And the two numbers that mean stop and think, which our article says local guidance should define action for: a creatinine rise of more than fifty per cent, or a potassium above five point five. Below those, a degree of movement is the price of treatment that works. The titration itself is roughly every two to four weeks, towards a named target dose, so ramipril ten milligrams daily rather than whatever they happen to be tolerating. Started and left at two point five is not treatment.

Sarah: And beta-blockers, where I think we are all more timid than the evidence allows. Our article says do not withhold a beta-blocker on the grounds of age, COPD, peripheral vascular disease, erectile dysfunction or diabetes. Those five are the reasons people usually don't start one. The genuine bars are narrow: second or third degree heart block without a pacemaker, or a heart rate below fifty.

Ben: Fifty per cent and five point five. Below those, keep going.

Sarah: What I actually say to the patient, because self-management is most of the year for them and ten minutes of it is us. Weigh yourself daily. A sudden rise of more than one and a half to two kilos over two days suggests fluid retention and may need the diuretic adjusting. That is a concrete, checkable instruction, and it's far better than telling somebody to watch for swollen ankles.

Sarah: And the advice to stop giving, because a lot of our patients are still carrying it from years ago. Avoid excessive salt, yes. But routine severe salt or fluid restriction is not recommended. You restrict fluid only if dilutional hyponatraemia occurs. So the patient rationing themselves to a litre a day because somebody told them to in two thousand and twelve can usually be released from that, and they will be delighted.

Ben: And there's a sick-day rule here too, the same one as in diabetes, because the SGLT2 inhibitor is now a heart failure drug taken by people who don't have diabetes at all. Rare but serious risk of ketoacidosis, including euglycaemic DKA with normal or near-normal glucose. Counsel to pause during acute serious illness, and test for ketones if the symptoms fit even when the glucose is normal.

Sarah: Two kilos in two days. And stop telling them to restrict fluids.

Sarah: Deprescribing is part of this, and it's the bit that never gets its own appointment. Three groups to actively hunt for and stop: anti-inflammatories, verapamil and diltiazem, and pioglitazone. All of them worsen heart failure. The anti-inflammatory is usually bought over the counter for a bad back and won't be on your screen, so you have to ask.

Ben: And the stable-patient package, which is easy to list and easy to forget. Annual influenza vaccination and a one-off pneumococcal. Refer to cardiac rehabilitation. Review in primary care at least six-monthly. And screen for iron deficiency, because intravenous iron is considered if the haemoglobin is under a hundred and fifty grams per litre with a ferritin under a hundred, or a transferrin saturation under twenty per cent. Note that threshold: you are looking for iron deficiency in people who are not anaemic by the usual definition.

Ben: Two more for completeness. If they're still symptomatic on optimal four-pillar therapy, that's an urgent specialist review, and the switch from an ACE inhibitor to sacubitril- valsartan needs the ACE inhibitor stopped at least thirty-six hours beforehand because of angioedema risk, and never co-prescribed with an ACE inhibitor or ARB. And on driving, for an ordinary licence, NYHA four means stop and notify. For a lorry or bus licence they must notify and need an ejection fraction of at least forty per cent.

Sarah: And the conversation that gets postponed until it's too late to have well. Discuss palliative care and advance care planning as the illness advances. Our article describes doing it sensitively and in small chunks, acknowledging uncertainty while being frank about the risk of sudden death. That's a hard sentence to say to somebody who feels reasonably well this month. It is much harder to say nothing and have the family find out the other way.

Ben: Hunt for the anti-inflammatories. They will not be on your screen.

Sarah: And one thing about how this fits into a ten-minute appointment, because reading that list back it sounds impossible. You are not doing all of it today. The four pillars get titrated over months, roughly every two to four weeks, and each of those is a short appointment with a blood test. What you are deciding today is only the next step up and when the bloods are due. Heart failure care in general practice is a sequence of small, boring, correct decisions, and the reason people end up under-treated is almost never that somebody made a wrong call. It's that nobody made the next one.

Ben: Which is why the six-monthly review matters so much. It's the appointment that exists purely to ask whether anybody has moved this forward since last time. And if the answer is no, and they're still on a starting dose two years later, that is the finding. Not the ejection fraction.

Ben: Right, exam corner. Sarah, and everybody driving, a few seconds. A sixty-six year old woman, breathless on exertion for two months, no known cardiac history. You send an NT- proBNP and it comes back at two thousand four hundred nanograms per litre. What does that mean for what you do next? A, heart failure is unlikely, look for another cause. B, refer for specialist assessment and echocardiography within six weeks. C, refer for specialist assessment and echocardiography within two weeks. Or D, start a loop diuretic and review in a month.

Ben: It's C. Above two thousand means echo and specialist assessment within two weeks, and very high levels carry a poor prognosis. B is the band below, four hundred to two thousand, which gets six weeks. A would be right only below four hundred, and even then with a caveat worth remembering: a normal NT-proBNP does not always exclude heart failure, because obesity pushes it down, as does African or African-Caribbean family background, and so do the very drugs we treat it with. And D treats her symptoms while leaving the diagnosis unmade.

Sarah: So. Three things for Monday. One. If the loop diuretic is the biggest dose on the list, that patient is under-treated, not optimised. Two. Do not withhold a beta-blocker for age, COPD, peripheral vascular disease, erectile dysfunction or diabetes. And three. NT-proBNP above two thousand is a two-week referral, four hundred to two thousand is six weeks, and a low result in an obese or already-treated patient still deserves a conversation.

Sarah: One thing to reflect on, if you're logging this. Pull up your heart failure register and look at the doses, not the diagnoses. How many of them are on a big diuretic and a small everything else? 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.