Decompensated (acute) heart failure: decongestion, cardiogenic shock and early follow-up

What changed in the 2026 ESC guideline

If you revised from the 2021 ESC guideline or its 2023 focused update, several examinable points have moved. The table lists the section 7 changes as the 2026 document states them.^[1]

^[1]

The 2026 guideline also introduces new names for treatments. Foundational medical therapy (FMT) is beta-blocker, ACE-I/ARNI/ARB, MRA and SGLT2 inhibitor for HFrEF, and MRA and SGLT2 inhibitor for HFpEF. Guideline-directed medical therapy now covers FMT plus additional medical therapy (AMT), and guideline-directed interventional therapy is abbreviated as GDIT.^[1]

Overview and definition

DHF can bring a patient to an unplanned admission, an emergency department visit, or an unplanned ambulatory visit needing HF treatment. It may be a worsening of known HF or the first presentation (de novo HF). In patients with pre-existing HF, worsening HF is a deterioration in symptoms or signs; DHF is specifically characterised by the need for treatment modification.^[1]

The risk is high. Death occurs in 3%–10% during the admission, 20%–30% at 1 year and more than 50% after 5 years, and 1-year rehospitalisation ranges between 35% and 44%. De novo DHF may carry higher in-hospital death but lower post-discharge death and rehospitalisation than DHF in pre-existing HF.^[1]

The 2022 AHA/ACC/HFSA guideline notes that most HF hospitalisations for decompensation are not truly "acute". They follow a gradual rise in filling pressures on pre-existing structural heart disease, often with precipitating factors that can be identified.^[2]

^[2]

The ESC Heart Failure Long-Term Registry (6629 patients) predates the 2026 terminology and used its own six phenotypes.^[3] In it, 'decompensated heart failure' was one phenotype alongside pulmonary oedema and cardiogenic shock, whereas in ESC 2026 pulmonary oedema and cardiogenic shock are themselves categories of DHF.^[3][1] In the registry, 61.1% had the decompensated HF phenotype, 13.2% pulmonary oedema, 14.4% HF with an acute coronary syndrome, 4.8% hypertensive HF, 3.5% right HF and 2.9% cardiogenic shock.^[3]

^[3]

Systolic pressure at initial presentation also stratified 1-year mortality: 34.8% below 85 mmHg, 29.0% at 85–110 mmHg, 21.2% at 110–140 mmHg and 17.4% above 140 mmHg. These differences tended to diminish after discharge; in patients who survived at least 6 months after discharge, 1-year mortality did not vary significantly by phenotype or SBP.^[3]

Classification: four clinical categories

The 2026 guideline recognises four categories of DHF based on clinical presentation. Pathophysiology, timing of onset and speed of deterioration differ between the four categories, but they can overlap.^[1]

^{[1] [2] [4]}

ESC 2026 names congestion and hypoperfusion as the two major clinical manifestations of DHF.^[1] In the AHA/ACC/HFSA guideline, initial triage includes clinical assessment of the haemodynamic profile for severity of congestion and adequacy of perfusion.^[2] The ESC-EORP-HFA registry classified 7865 patients at admission into four profiles defined by congestion and perfusion: the split was dry-warm 9.9%, wet-warm 69.9%, wet-cold 19.8% and dry-cold 0.4%. In-hospital mortality was 2.0%, 3.8%, 12.1% and 9.1% respectively.^[4]

The 2026 ESC guideline uses the SCAI classification, introduced by the Society for Cardiovascular Angiography and Interventions, to grade the severity of cardiogenic shock.^[1]

^[1]

Pathophysiology

^[1]

Congestion can come from four sources: fluid overload, fluid redistribution, increased vascular permeability and raised intracardiac pressures.^[1]

In APO, the dominant process is fluid redistribution: pulmonary congestion without total volume overload, mainly from blood shifting into the pulmonary circulation. Such patients usually do not need high doses of diuretics, but they may need vasodilators.^[1]

Decompensated right-sided HF is often associated with tricuspid regurgitation and can impair LV filling and reduce systemic cardiac output through ventricular interdependence. RV function and its effect on the kidney and liver usually set the clinical course, so kidney and hepatic function need close monitoring.^[1]

Hypoperfusion is not a synonym for hypotension. Cardiogenic shock can present as hypoperfusion with normal blood pressure, and this form has a better prognosis than hypotensive shock.^[1]

Treatments act on this physiology too. Non-invasive positive pressure ventilation raises intrathoracic pressure, reducing venous return and right and left ventricular preload; it may also lower cardiac output and blood pressure. It needs caution in patients with reduced preload reserve and hypotension, such as decompensated right-sided HF.^[1]

Clinical presentation

The 2026 guideline's Table 12 sorts the findings into three groups. Use it as the checklist at the bedside.^[1]

^[1]

Clinical criteria for the diagnosis of APO include signs and symptoms of respiratory distress (dyspnoea with orthopnoea, respiratory failure with hypoxaemia with or without hypercapnia, tachypnoea above 25 breaths/min, and increased work of breathing) caused by massive pulmonary congestion. It typically develops over a short period.^[1] The AHA/ACC/HFSA guideline notes that some patients present with pulmonary oedema and severe hypertension needing urgent blood-pressure reduction, more commonly when LVEF is preserved.^[2]

In the AHA/ACC/HFSA guideline, most patients admitted with HF have congestion without apparent hypoperfusion. In decompensated chronic HF with low EF, right- and left-sided filling pressures usually rise in proportion, but up to 1 in 4 patients have a mismatch. Resting hypoperfusion is often underappreciated in chronic HF but can be suspected from a narrow pulse pressure and cool extremities.^[2]

Differential diagnosis

^{[1] [2]}

Clinical and bedside assessment

If DHF is suspected, assess congestion and hypoperfusion at first medical contact. Start non-invasive blood pressure, respiratory rate, pulse oximetry, temperature and an ECG early. In a haemodynamically unstable patient with suspected hypoperfusion, measure serum lactate promptly.^[1]

Triage on haemodynamic and clinical status decides the level of care and monitoring. Risk scores have been proposed, but they still need validation. Pre-hospital care should not delay transfer to the right setting.^[1]

Investigations

Natriuretic peptides. The non-age-adjusted rule-out thresholds include NT-proBNP below 300 pg/mL and MR-proANP below 120 pg/mL. Age-related NT-proBNP rule-in thresholds make DHF likely above 450 pg/mL under 50 years, above 900 pg/mL at 50–74 years and above 1800 pg/mL over 74 years.^[1]

• If suspicion persists despite a value below rule-out, seek specialist assessment and an echocardiogram.^[1]
• NT-proBNP above 5000 pg/mL indicates very high risk, and admission is generally recommended.^[1]
• The AHA/ACC/HFSA guideline notes that natriuretic peptides are less sensitive in obesity and HFpEF, and less specific in sepsis.^[2]
• NT-proBNP results are unaltered by ARNI treatment, whereas ARNIs affect BNP degradation.^[1]

Imaging. Perform transthoracic echocardiography when NT-proBNP is high or suspicion of DHF is high. Where trained staff are available, lung ultrasound identifies pulmonary congestion and is useful for serial assessment.^[1]

Perfusion. In the ESC 2026 cardiogenic shock section, the biochemical manifestation of inadequate tissue perfusion is best measured by arterial lactate, and a value above 2 mmol/L is required. Acute kidney injury and/or acute hepatic injury are additional markers of hypoperfusion.^[1] The AHA/ACC/HFSA clinical shock criteria need both a pressure criterion and a hypoperfusion criterion.^[2]

^[2]

Triggers. Initial investigations aim both to diagnose DHF and to detect triggers rapidly. Thyroid function is recommended in de novo HF or DHF.^[1]

^[1]

Monitoring during treatment. The AHA/ACC/HFSA guideline lists careful measurement of fluid intake and output, vital signs, standing weight at the same time each day, and congestion and perfusion signs. Electrolytes, urea and creatinine are checked daily during active medication adjustment.^[2]

Management: three phases

In-hospital care runs through three phases. Patients with cardiogenic shock usually stay in phase 1 for several days and may never reach phase 2. Uncomplicated left-sided, right-sided or APO presentations can pass through all three quickly.^[1]

^{[1] [1]}

Phase 1: resuscitation and first decisions

When CS is suspected and temporary MCS may be needed, consult the Shock Team. In these patients, and in any other life-threatening presentation, diagnostic work-up and rescue treatment start promptly and simultaneously. Identification and possible treatment of acute causes is crucial.^[1]

Do not stop foundational therapy reflexively. In patients already on FMT, continuation or rapid reintroduction is recommended; discontinuation is not recommended unless there are clear signs of hypoperfusion or specific clinical indications.^[1] For patients hospitalised with HFrEF, the AHA/ACC/HFSA guideline says withholding or reducing the beta-blocker should be considered with marked volume overload or marginal low cardiac output, and that oral therapy should not be withheld for mild or transient falls in blood pressure.^[2] B-CONVINCED randomised patients with LVEF below 40% on stable beta-blockade, admitted with acutely decompensated HF, to continuation or discontinuation of the beta-blocker. Continuing the beta-blocker met the non-inferiority criterion for the main end point, improvement in both dyspnoea and general well-being at 3 days as judged by a physician blinded to therapy (92.8% versus 92.3%), and more patients were on beta-blockade at 3 months (90% versus 76%).^[25]

Oxygen and ventilation

^[1]

Studies, including RCTs, show that indiscriminate oxygen in non-hypoxaemic patients may be harmful.^[1] In adults with acute cardiogenic pulmonary oedema, a Cochrane review found that NPPV, compared with standard medical care, probably reduces intubation (RR 0.49, NNTB 13). It may reduce hospital mortality (RR 0.65, NNTB 17, low-quality evidence).^[11]

Intubation criteria in DHF are cardiac or respiratory arrest, progressive worsening of altered mental status, or progressive respiratory failure despite NIV or intolerance of it. That failure means hypoxaemia with PaO₂ below 60 mmHg, PaCO₂ above 50 mmHg and pH below 7.35. Persistent haemodynamic instability and a need for airway protection complete the list.^[1]

Decongestion

Intravenous loop diuretics are the cornerstone. They are recommended for all patients with DHF admitted with fluid overload (Class I, Level A). The Task Force assigns Level A because of clear clinical benefit; no adequately powered RCT has ever been done or is likely to be done, for ethical reasons.^[1]

^[1]

The AHA/ACC/HFSA 2022 guideline notes that diuretic titration in recent trials of hospitalised HF was often started with at least 2 times the daily home dose, mg for mg, intravenously.^[2]

DOSE. Patients were randomised to bolus every 12 hours or continuous infusion, at either their previous oral dose or 2.5 times it. There were no significant differences in global symptom assessment or renal function between bolus and infusion, or between high and low dose.^[5] High dose gave more diuresis but more transient worsening renal function (23% versus 14%), defined as a serum creatinine rise of more than 0.3 mg/dL at any time during the 72 hours after randomisation. There was no evidence of worse clinical outcomes at 60 days, but the trial was not powered to detect differences in clinical events.^[5] The trial enrolled chronic HF patients on moderate-to-high outpatient doses, so its findings may not apply to newly diagnosed HF or to patients with more modest diuretic requirements.^[5]

Adding a second diuretic (Class IIa, Level B1). Short-term i.v. acetazolamide or oral hydrochlorothiazide should be considered in patients with fluid overload previously treated with loop diuretics.^[1]

• ADVOR. Patients had clinical volume overload and NT-proBNP above 1000 pg/mL (or BNP above 250 pg/mL). Acetazolamide 500 mg i.v. once daily, added to loop diuretic at twice the oral maintenance dose, achieved successful decongestion (no signs of volume overload within 3 days, without escalation of decongestive therapy) in 42.2% versus 30.5% (RR 1.46).^[6] Worsening kidney function, hypokalaemia, hypotension and adverse events were similar; death or HF rehospitalisation during 3 months of follow-up was 29.7% versus 27.8% (HR 1.07).^[6]
• CLOROTIC. Hydrochlorothiazide or placebo was added to an i.v. furosemide regimen; hydrochlorothiazide increased 72-hour weight loss (2.3 versus 1.5 kg) without improving patient-reported dyspnoea. Impaired renal function was more frequent, and mortality and rehospitalisation did not differ.^[7]
• Neither trial showed benefit on hard end points. The ESC text says hydrochlorothiazide was more likely to worsen kidney function and cause hypokalaemia, warranting caution.^[1]

Urinary sodium guidance. Urinary sodium-guided diuretic therapy may be considered during the first days of treatment to improve natriuresis and diuresis (Class IIb, Level B1).^[1] In PUSH-AHF, treatment was intensified when spot urinary sodium was below 70 mmol/L. 24-hour urinary sodium excretion, a primary end point, was higher (409 versus 345 mmol), but 180-day death or HF rehospitalisation was 31% in both arms.^[8][1] In ENACT-HF, a standardised natriuresis-guided diuretic protocol in patients with signs of volume overload raised day-1 natriuresis (282 versus 174 mmol) and shortened length of stay (5.8 versus 7.0 days).^[9] No trial has yet shown a benefit of urinary sodium-guided therapy on death or recurrent HF hospitalisation.^[1]

Creatinine during diuresis. Small, transient creatinine rises are not linked to poor outcome if decongestion is achieved. Adverse outcomes are seen when worsening kidney function comes with a poor diuretic response; they are absent when the response is good.^[1] The AHA/ACC/HFSA guideline says diuresis should not be stopped prematurely for small creatinine changes. Rises around 0.3 mg/dL do not predict worse outcomes except when patients are discharged with persistent congestion.^[2]

Fluid and sodium. Keep fluid intake below urine output and sodium intake below natriuresis; fluid restriction may be appropriate in hyponatraemia, based on individual assessment.^[1]

Ultrafiltration. Its role is uncertain; it may be considered in selected patients responding poorly to diuretic escalation.^[1] In an RCT by Bart et al. (2012; worsened renal function and persistent congestion), ultrafiltration was inferior to stepped pharmacological therapy on the 96-hour bivariate change in creatinine and body weight, owing mainly to a creatinine rise. At 96 hours creatinine rose with ultrafiltration and weight loss was similar. Over follow-up, serious adverse events were more frequent with ultrafiltration (72% versus 57%).^[12]

SGLT2 inhibitors in hospital

In-hospital initiation of an SGLT2 inhibitor is recommended in DHF after initial stabilisation, to improve quality of life and congestion symptoms and reduce HF hospitalisation (Class I, Level B1).^[1] The ESC recommends early initiation both as part of FMT and as a tool for decongestion.^[1]

EMPULSE randomised 530 patients with de novo or decompensated chronic HF, regardless of LVEF, to empagliflozin 10 mg once daily or placebo. Randomisation took place in hospital once stable, a median of 3 days after admission.^[10] More patients had clinical benefit, a hierarchical composite assessed at 90 days by win ratio (stratified win ratio 1.36). The benefit held across de novo and chronic HF, ejection fraction and diabetes status.^[10] ESC Table 11 lists dapagliflozin and empagliflozin at 10 mg once daily as both starting and target dose.^[1]

Vasodilators, inotropes, vasopressors and opiates

^[1]

Vasodilators. I.v. nitroglycerine or nitroprusside may be effective, particularly with predominant vasoconstriction and volume redistribution. In APO, with raised filling pressures and blood pressure but little systemic fluid accumulation, vasodilators can reduce pulmonary congestion. Monitor blood pressure closely, because vasodilator-related hypotension may be harmful.^[1] The ESC supplement advises extreme caution in left ventricular hypertrophy and/or severe aortic stenosis. It adds that favourable effects were described in LV systolic dysfunction with aortic stenosis when vasodilators were given with careful haemodynamic monitoring.^[1] In a meta-analysis of 46 RCTs (28,374 patients), vasodilators did not reduce mortality (RR 0.95) but were associated with less tracheal intubation (RR 0.54).^[13]

The AHA/ACC/HFSA guideline adds practical cautions. Patients with hypertension, coronary ischaemia or significant MR may be suitable for nitroglycerin, but tachyphylaxis may develop within 24 hours. Up to 20% of patients with HF may develop resistance even to high doses.^[2] Nitroprusside can cause marked hypotension, so it typically needs invasive blood-pressure monitoring and is usually used in intensive care. Longer infusions have rarely been associated with thiocyanate and cyanide toxicity, particularly with renal insufficiency and significant hepatic disease.^[2]

Inotropes. Because of the risks of arrhythmias and myocardial ischaemia, inotropes are not routinely recommended.^[1] A Cochrane review found no convincing data that any inotrope or vasodilator reduces mortality in cardiogenic shock or low-output states.^[14] A double-blind RCT (Mathew 2021) randomised patients with cardiogenic shock to milrinone or dobutamine: there was no significant difference in the primary composite outcome (49% versus 54%, RR 0.90) or in in-hospital death (37% versus 43%).^[15] No relevant levosimendan RCTs in CS exist; by mode of action, levosimendan or a PDE-III inhibitor may be preferred over dobutamine in patients on beta-blockers.^[1]

Vasopressors. Some studies, despite limitations, support norepinephrine as first choice over dopamine or epinephrine. The target mean arterial pressure in CS is not well defined, and vasopressin has not been studied in CS.^[1] An RCT by De Backer et al. (2010) randomised patients with shock to dopamine or norepinephrine as first-line vasopressor; arrhythmic events were more frequent with dopamine than with norepinephrine (24.1% versus 12.4%). In a subgroup analysis it was associated with higher 28-day mortality in the 280 patients with cardiogenic shock.^[16] The ESC notes this subgroup benefit of norepinephrine lacked a significant interaction P value.^[1] In OptimaCC (cardiogenic shock after acute MI), cardiac index was similar, but refractory shock was more frequent with epinephrine (37% versus 7%) and the trial stopped early.^[17]

^[1]

Opiates. Limit them to carefully selected patients with severe anxiety or distress refractory to other treatment, because of the risk of respiratory failure.^[1]

Cardiogenic shock and temporary mechanical circulatory support

The SHARC consensus quoted by ESC defines cardiogenic shock as a cardiac disorder with clinical and biochemical evidence of sustained tissue hypoperfusion. ESC adds that CS is generally characterised by hypoperfusion and hypotension, but no specific blood-pressure cut-off is defined.^[1] In the AHA/ACC/HFSA framing, hypotension is the main manifestation but is not enough for the diagnosis.^[2] The AHA/ACC/HFSA guideline groups causes into acute decompensation of chronic HF, acute myocardial dysfunction without prior HF, and survivors of cardiac arrest; in acute MI, urgent revascularisation is paramount.^[2]

Temporary MCS should be considered as an option in highly selected patients, to stabilise haemodynamics and organ perfusion. A multidisciplinary Shock Team guides the modality and type: an HF cardiologist, a cardiothoracic surgeon, an interventional cardiologist and an intensivist, at a high-volume advanced HF centre.^[1]

^[1]

"No risk of hypoxic brain injury" means no out-of-hospital arrest with persistent Glasgow coma scale below 8 after ROSC, and no resuscitation or resuscitation lasting under 10 min. For MI-related mechanical complications, the table footnotes add three points. The choice of passive IABP or active percutaneous MCS should be based on shock severity as assessed by the Shock Team. With a large ventricular septal defect, temporary MCS should be used with caution because of possible shunt changes. With ventricular free-wall rupture, temporary MCS should be avoided, although VA-ECLS can be considered to allow emergent surgery in profound shock or arrest without signs of irreversible brain injury.^[1]

^{[1] [18] [19] [20] [21]}

VA-ECLS raises LV afterload through retrograde flow and may harm the heart; active unloading with microaxial pumps or IABP aims to mitigate this. RCTs of ECLS in non-ischaemic CS, and of microaxial pumps in non-infarct CS, have not yet been done.^[1]

Specific scenarios

Acute pulmonary oedema. Treatment aims to relieve pulmonary congestion, treat respiratory failure and keep peripheral perfusion. Investigate the cause of decompensation in every patient.^[1]

Decompensated right-sided HF. Pulmonary embolism and acute MI involving the right ventricle need to be excluded during initial evaluation and treated. Care combines diuretics, judicious fluid management, inotropes and vasopressors, and MCS when indicated. Use non-invasive ventilation with caution in patients with reduced preload reserve and hypotension, because it reduces preload and may lower blood pressure.^[1]

AF with rapid ventricular rates. Urgent cardioversion is recommended in DHF with rapid rates and haemodynamic instability (Class I, Level C). I.v. amiodarone or digoxin may be considered in unstable HFrEF with AF for acute rate control (Class IIb, Level C).^[1] Amiodarone is recommended for pharmacological cardioversion in HFrEF. In HFrEF, propafenone, flecainide and dronedarone are linked to poorer outcomes and should be avoided.^[1] In stable patients with HFrEF and AF, beta-blockers are recommended as first-line short- and long-term rate control (Class I, Level C). In the same stable patients, digoxin should be considered if the rate stays high despite beta-blockers, or if they are contraindicated or not tolerated (Class IIa, Level C).^[1] Oral anticoagulation is recommended in clinical AF at elevated thromboembolic risk by CHA₂DS₂-VA score (Class I, Level A). DOACs are recommended in preference to VKAs in HF (Class I, Level B1), except in moderate or severe mitral stenosis or mechanical prosthetic valves, where VKAs are recommended.^[1]

Ambulatory decompensation. In ambulatory patients with chronic HF who are worsening, i.v. diuretics or intensified oral diuretics can be considered. Outpatient i.v. diuretic care covers day-hospital and hospital-at-home models; sessions usually consist of a 3–6 h i.v. infusion dosed from the existing oral regimen.^[1] Thiazide-like diuretics, particularly oral metolazone 2.5–5 mg, can be used in advanced HF with diuretic resistance, as part of sequential nephron blockade, or when eGFR is below 30 mL/min/1.73 m², with close monitoring of eGFR and electrolytes.^[1]

Thromboprophylaxis

ESC 2026 makes no definitive recommendation because no RCT has compared anticoagulation with placebo in DHF. It states prophylaxis is generally indicated with immobilisation or other risk factors.^[1]

The AHA/ACC/HFSA guideline is more specific. In patients admitted specifically for decompensated HF with creatinine clearance above 30 mL/min, trials suggest enoxaparin 40 mg subcutaneously once daily, unfractionated heparin 5000 units subcutaneously every 8 or 12 hours, or rivaroxaban 10 mg once daily reduce radiographic venous thrombosis. Effects on mortality and clinically significant pulmonary embolism are unclear.^[2] Extending prophylaxis after discharge reduced VTE but increased bleeding, and overall did not appear to provide additional benefit.^[2]

Phase 2: stabilisation and foundational therapy

Once the patient is stable, start or optimise long-term FMT while decongestion continues.^[1] RCTs show that starting MRAs, SGLT2 inhibitors or ARNIs in this phase is feasible and safe. The guideline calls the traditional sequential approach suboptimal and time-consuming, and prefers simultaneous implementation. It gives no specific recommendation on how many classes to start together or in what order.^[1]

A more conservative approach may be kept for specific high-risk groups, such as patients with low eGFR, hypotension or a history of hyperkalaemia.^[1] An SGLT2 inhibitor (dapagliflozin or empagliflozin) and an MRA (steroidal MRA for HFrEF; steroidal or non-steroidal MRA for HFpEF) are each recommended in symptomatic HF independent of LVEF (both Class I, Level A).^[1]

PIONEER-HF. Patients with HFrEF hospitalised for acute decompensated HF were randomised after haemodynamic stabilisation to sacubitril–valsartan (target 97/103 mg twice daily) or enalapril (target 10 mg twice daily).^[23] The time-averaged NT-proBNP reduction (weeks 4 and 8 versus baseline) was greater: 46.7% versus 25.3%, without significant differences in worsening renal function, hyperkalaemia, symptomatic hypotension or angioedema.^[23] Chronic dosing, titration and ACE inhibitor switching rules are covered in the HFrEF topic.

Phase 3: pre-discharge and early follow-up

This phase covers the last days in hospital and the first weeks after discharge. The aims are fewer recurrent episodes, further FMT optimisation, and decisions on AMT and GDIT.^[1]

^[1]

Residual congestion. Residual congestion is associated with poor outcomes and a high risk of rehospitalisation.^[1] In the registry, 30.9% of survivors left hospital congested, with higher 1-year mortality (28.0% versus 18.5%).^[4] The AHA/ACC/HFSA guideline reports persistent congestion at discharge in 25% to 50% of patients.^[2] The ESC tools for assessing decongestion in the pre-discharge phase (Figure 14) are optional because no RCT evidence exists for them. A congestion score of 0 means no dyspnoea, orthopnoea, fatigue, rales, oedema or jugular distension.^[1]

STRONG-HF. Patients aged 18–85 years who were not on full doses of guideline-directed treatment were randomised before discharge. High-intensity care up-titrated treatment to 100% of recommended doses within 2 weeks of discharge, with four scheduled outpatient visits over the 2 months after discharge monitoring clinical status, laboratory values and NT-proBNP.^[22] HF readmission or death by day 180 fell from 23.3% to 15.2% (risk ratio 0.66), and the trial stopped early.^[22] By day 90, adverse events were more frequent (41% versus 29%), but serious (16% versus 17%) and fatal (5% versus 6%) events were similar.^[22]

The ESC notes STRONG-HF's limits: a highly selected population, suboptimal treatment in the control arm and no SGLT2 inhibitor use. It still recommends fast uptitration after HF hospitalisation as safe and feasible.^[1] Safety indicators, including blood pressure, heart rate, creatinine, potassium and NT-proBNP, help guide uptitration. A fall in eGFR within a week of discharge was linked to more congestion and readmission in STRONG-HF.^[1]

Iron. If not done earlier, the pre-discharge phase may also be used to give iron to patients with iron deficiency, based on AFFIRM-AHF.^[1] In AFFIRM-AHF (LVEF below 50%), ferric carboxymaltose missed its primary end point, total HF hospitalisations and cardiovascular death up to 52 weeks (RR 0.79, p=0.059). It reduced total HF hospitalisations (RR 0.74), with no difference in cardiovascular death.^[24]

Discharge medicines and programmes. The AHA/ACC/HFSA guideline notes that most patients who needed i.v. diuretics will need a loop diuretic at discharge.^[2] Enrol patients in disease-management and rehabilitation programmes, with education and lifestyle guidance.^[1]

Prognosis and disposition

In the ESC-EORP-HFA registry, profile at admission tracked outcome. Wet-cold patients had the highest in-hospital mortality (12.1%), and, classified at admission, wet-warm patients had higher 1-year mortality than dry-warm (adjusted HR 1.78).^[4] Patients whose worsening HF progresses to advanced HF often have frequent hospitalisations and need prompt evaluation for advanced HF treatments.^[1]

Early consultation with an advanced HF centre is recommended for patients with, or at risk of, advanced HF who are motivated and have no absolute contraindication to transplantation or durable MCS (Class I, Level A).^[1] Access to an integrated palliative care team is recommended in advanced HF (Class I, Level B2).^[1]

Special populations

• Older patients. Age-related NT-proBNP thresholds may aid triage and reduce unnecessary tests.^[1]
• Chronic kidney disease. Start with a higher i.v. loop dose. Hydrochlorothiazide dose rises as eGFR falls; acetazolamide may be preferable in more advanced CKD.^[1]
• Patients on beta-blockers needing inotropic support. Levosimendan or a PDE-III inhibitor may be preferred over dobutamine.^[1] The AHA/ACC/HFSA guideline advises caution when starting beta-blockers in patients who needed inotropes during hospitalisation.^[2]
• Advanced HFrEF with hypoperfusion. Down-titrating or stopping beta-blockers or ivabradine should be considered in selected patients with organ hypoperfusion despite initial treatment (Class IIa, Level C).^[1]

Evidence and regional differences

^{[5] [6] [7] [8] [10] [12] [15] [22] [23] [1]}

ESC 2026 versus AHA/ACC/HFSA 2022.

• Starting diuretic: in ESC 2026, diuretic-naive patients usually respond well to 40 mg i.v. furosemide, and twice the usual oral dose i.v. should be considered for patients on loop diuretics.^[1] AHA/ACC/HFSA 2022 notes that recent trials often started with at least twice the daily home dose intravenously.^[2]
• Thromboprophylaxis: ESC 2026 makes no definitive recommendation (no RCT in DHF) but states that, in case of immobilisation or other risk factors, thromboembolic prophylaxis is generally indicated.^[1] AHA/ACC/HFSA 2022 cites enoxaparin, unfractionated heparin and rivaroxaban regimens.^[2]
• Terminology: ESC 2026 uses decompensated HF and four clinical categories.^[1] AHA/ACC/HFSA 2022 uses acute decompensated HF triaged by congestion and perfusion.^[2]

Australia and New Zealand (NHFA/CSANZ 2018)

The National Heart Foundation of Australia and the Cardiac Society of Australia and New Zealand (NHFA/CSANZ) published their Guidelines for the Prevention, Detection, and Management of Heart Failure in Australia in 2018; they were intended to replace the 2011 update.^[26] This section draws on the full guideline text published in Heart, Lung and Circulation, which gives guidance on the clinical care of adult patients with heart failure in Australia.^[26] Its recommendations were formulated with GRADE methodology, which gives the strength of a recommendation for or against an intervention; each row of the NHFA/CSANZ 2018 recommendation tables below condenses one selected recommendation and keeps its printed strength and quality of evidence.^[26] For topics with a limited evidence base, or where the impact of interventions on clinical outcomes was considered modest, the guideline includes comments in its practice advice sections.^[26] The guideline received final approval in June 2018, so it predates the AHA/ACC/HFSA 2022 and ESC 2026 guidelines used in the rest of this topic.^[26] Comparison rows set selected 2018 rows beside the matching ESC 2026 rows; any 2018 row that differs from a newer recommendation elsewhere in this topic should be read as history rather than current practice.^[26] It uses the term acute heart failure; the 2026 ESC guideline replaces acute HF with decompensated HF.^[26][1]

Assessment and acute treatment

^[26]

Practice advice in the acute phase

• Oxygen therapy is not recommended in acute heart failure with oxygen saturation of 94% or above.^[26]
• In hypoxic patients given oxygen therapy, a target saturation of 94–98% should be achieved; in patients at risk of hypercapnoea (type II respiratory failure), 88–92% should be targeted.^[26]
• Non-invasive ventilatory support is recommended in acute pulmonary congestion with persisting hypoxaemia (SaO₂ below 94%) and tachypnoea (respiratory rate above 25/min) despite oxygen therapy; patients with acute pulmonary congestion and type II respiratory failure with hypercapnoea and acidosis who need non-invasive ventilatory support are suitable candidates for BiPAP.^[26]
• Opiates should generally be avoided in acute heart failure, particularly with hypotension or a risk of aspiration or hypoventilation.^[26]
• Intravenous loop diuretics such as furosemide are first-line therapy in acute heart failure with congestion, and the intravenous dose should be at least equal to the oral dose taken at home.^[26]
• In patients not previously on diuretics, initial therapy is 20–40 mg intravenous furosemide; in renal impairment the dose may need to be increased.^[26]
• With no response to an intravenous loop diuretic, intravenous vasodilators, oral thiazides or an MRA such as spironolactone may be added, provided the patient is not hypotensive.^[26]
• Intravenous nitrate therapy is generally preferred among the vasodilators, but beyond 24–48 hours it may result in tolerance.^[26]
• IABP does not improve outcomes in cardiogenic shock associated with acute MI.^[26]
• While ultrafiltration may be considered in acute heart failure with congestion not responding to diuretics and other maximal therapy, it remains unclear how to best select these patients; it does not improve survival, length of hospital admission or rehospitalisation rates compared with diuretics.^[26]

Natriuretic peptide cut-offs

As a pragmatic rule-out guide, NHFA/CSANZ 2018 proposes BNP below 100 ng/L and NT-proBNP below 300 ng/L, noting that precise cut-offs vary between trials and with individual patient characteristics such as age, weight and renal function.^[26] It adds that BNP and NT-proBNP levels are generally lower in HFpEF than in HFrEF, so their rule-out reliability is significantly weaker in suspected HFpEF.^[26] Its Table 4 gives rule-in values of BNP above 400 ng/L, and NT-proBNP above 450 ng/L under 50 years, above 900 ng/L at 50–75 years and above 1800 ng/L over 75 years.^[26] A footnote to that table says defining cut-off values, particularly for rule-in, is complicated and somewhat limited in accuracy because multiple factors influence natriuretic peptide levels.^[26] ESC 2026 gives a non-age-adjusted rule-out threshold for DHF of NT-proBNP below 300 pg/mL, and age-related "likely" rule-in thresholds, which may reduce unnecessary exams, of NT-proBNP above 450 pg/mL under 50 years, above 900 pg/mL at 50–74 years and above 1800 pg/mL over 74 years.^[1]

Discharge and follow-up

The 2018 guideline notes that starting treatment that improves long-term prognosis, such as a beta blocker, before hospital discharge increases the likelihood of long-term maintenance treatment.^[26] As practice advice, patients should ideally be reviewed within the first 7–14 days of discharge from hospital, regardless of the type of appointment.^[26]

^[26]

Where ESC 2026 differs

Where the 2026 ESC guideline gives a different recommendation, the table sets the two side by side; read the 2018 column as history.^[26][1]

^{[26] [1]}

Later Australian consensus statement (2026)

A consensus statement published in 2026 says there is a real need for unified recommendations to optimise guideline-directed medical therapy (GDMT) for patients hospitalised with acute heart failure in Australia.^[27] Using a modified Delphi method, an expert panel of nine Australian clinicians with expertise in heart failure management developed consensus statements to guide healthcare professionals in Australia on optimising GDMT.^[27] The statement outlines a strategy for starting GDMT while patients are in hospital and optimising it to maximum tolerated doses rapidly after discharge, and is designed to enhance the current acute heart failure guidelines.^[27]

Exam pearls

• NT-proBNP rule-out below 300 pg/mL (not age-adjusted); DHF is likely above 450 pg/mL (under 50 years), 900 pg/mL (50–74 years) or 1800 pg/mL (over 74 years).^[1]
• Oxygen when SpO₂ is below 90% or PaO₂ below 60 mmHg (I, C); indiscriminate oxygen in non-hypoxaemic patients may be harmful.^[1]
• Non-invasive positive pressure ventilation should be considered for respiratory distress (respiratory rate above 25 breaths/min, SpO₂ below 90%) and started as soon as possible (IIa, B2).^[1]
• I.v. vasodilators may be considered as initial therapy above SBP 110 mmHg (IIb, C); inotropes may be considered below SBP 90 mmHg with hypoperfusion not responding to standard treatment, including fluid challenge (IIb, C).^[1]
• A satisfactory diuretic response can be defined as urinary sodium ≥70 mEq/L at 2 h, or urine output ≥100 mL/h over the first 6 h. Below 3 L at 24 h with persisting congestion: consider further escalation on day 2.^[1]
• In cardiogenic shock, arterial lactate above 2 mmol/L is required as the biochemical manifestation of inadequate tissue perfusion.^[1]
• Class III in DHF: routine opiates (unless severe or intractable pain or anxiety), temporary MCS in unselected cardiogenic shock caused by acute MI, and IABP in unselected cardiogenic shock.^[1]

References27ShowHide

1. [1]Køber L et al. 2026 ESC Guidelines for the management of heart failure. Eur Heart J, 2026.PMID 42661420
2. [2]Heidenreich PA et al. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation, 2022.PMID 35363499
3. [3]Chioncel O et al. Clinical phenotypes and outcome of patients hospitalized for acute heart failure: the ESC Heart Failure Long-Term Registry. Eur J Heart Fail, 2017.PMID 28463462
4. [4]Chioncel O et al. Acute heart failure congestion and perfusion status - impact of the clinical classification on in-hospital and long-term outcomes; insights from the ESC-EORP-HFA Heart Failure Long-Term Registry. Eur J Heart Fail, 2019.PMID 31127678
5. [5]Felker GM et al. Diuretic strategies in patients with acute decompensated heart failure. N Engl J Med, 2011.PMID 21366472
6. [6]Mullens W et al. Acetazolamide in Acute Decompensated Heart Failure with Volume Overload. N Engl J Med, 2022.PMID 36027559
7. [7]Trullàs JC et al. Combining loop with thiazide diuretics for decompensated heart failure: the CLOROTIC trial. Eur Heart J, 2023.PMID 36423214
8. [8]Ter Maaten JM et al. Natriuresis-guided diuretic therapy in acute heart failure: a pragmatic randomized trial. Nat Med, 2023.PMID 37640861
9. [9]Dauw J et al. Protocolized Natriuresis-Guided Decongestion Improves Diuretic Response: The Multicenter ENACT-HF Study. Circ Heart Fail, 2024.PMID 38179728
10. [10]Voors AA et al. The SGLT2 inhibitor empagliflozin in patients hospitalized for acute heart failure: a multinational randomized trial. Nat Med, 2022.PMID 35228754
11. [11]Berbenetz N et al. Non-invasive positive pressure ventilation (CPAP or bilevel NPPV) for cardiogenic pulmonary oedema. Cochrane Database Syst Rev, 2019.PMID 30950507
12. [12]Bart BA et al. Ultrafiltration in decompensated heart failure with cardiorenal syndrome. N Engl J Med, 2012.PMID 23131078
13. [13]Lukoschewitz JD et al. Vasodilators for Acute Heart Failure - A Systematic Review with Meta-Analysis. NEJM Evid, 2024.PMID 38804781
14. [14]Uhlig K et al. Inotropic agents and vasodilator strategies for the treatment of cardiogenic shock or low cardiac output syndrome. Cochrane Database Syst Rev, 2020.PMID 33152122
15. [15]Mathew R et al. Milrinone as Compared with Dobutamine in the Treatment of Cardiogenic Shock. N Engl J Med, 2021.PMID 34347952
16. [16]De Backer D et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med, 2010.PMID 20200382
17. [17]Levy B et al. Epinephrine Versus Norepinephrine for Cardiogenic Shock After Acute Myocardial Infarction. J Am Coll Cardiol, 2018.PMID 29976291
18. [18]Møller JE et al. Microaxial Flow Pump or Standard Care in Infarct-Related Cardiogenic Shock. N Engl J Med, 2024.PMID 38587239
19. [19]Thiele H et al. Extracorporeal Life Support in Infarct-Related Cardiogenic Shock. N Engl J Med, 2023.PMID 37634145
20. [20]Thiele H et al. Intraaortic balloon support for myocardial infarction with cardiogenic shock. N Engl J Med, 2012.PMID 22920912
21. [21]Morici N et al. Early Intra-Aortic Balloon Support for Heart Failure-Related Cardiogenic Shock: A Randomized Clinical Trial. J Am Coll Cardiol, 2025.PMID 40162941
22. [22]Mebazaa A et al. Safety, tolerability and efficacy of up-titration of guideline-directed medical therapies for acute heart failure (STRONG-HF): a multinational, open-label, randomised, trial. Lancet, 2022.PMID 36356631
23. [23]Velazquez EJ et al. Angiotensin-Neprilysin Inhibition in Acute Decompensated Heart Failure. N Engl J Med, 2019.PMID 30415601
24. [24]Ponikowski P et al. Ferric carboxymaltose for iron deficiency at discharge after acute heart failure: a multicentre, double-blind, randomised, controlled trial. Lancet, 2020.PMID 33197395
25. [25]Jondeau G et al. B-CONVINCED: Beta-blocker CONtinuation Vs. INterruption in patients with Congestive heart failure hospitalizED for a decompensation episode. Eur Heart J, 2009.PMID 19717851
26. [26]Atherton JJ, Sindone A, De Pasquale CG, et al. National Heart Foundation of Australia and Cardiac Society of Australia and New Zealand: Guidelines for the Prevention, Detection, and Management of Heart Failure in Australia 2018. Heart Lung Circ, 2018.PMID 30077227
27. [27]Sindone AP, Amerena J, De Pasquale CG, et al. Consensus Statement on Optimisation of Patient Care After Hospitalisation for Acute Heart Failure. Heart Lung Circ, 2026.PMID 41539881