ESC 2021: acute and chronic heart failure
International ESC Current
2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure
McDonagh TA, et al. Eur Heart J. 2021;42:3599–3726
DOI 10.1093/eurheartj/ehab368 · Summary checked against the source: 2026-11
Read with the 2023 focused update
The ESC 2023 focused update upgraded SGLT2 inhibitors in HFmrEF and HFpEF to Class I, upgraded IV iron, and added a post-discharge up-titration recommendation. Where this page and the update differ, follow the update.
At a glance
- Diagnosis: NT-proBNP ≥125 or BNP ≥35, then echocardiography, then classify by LVEF. [p. 3615]
- HFrEF: four Class I drugs for everyone: ACE-I (or ARNI as a replacement), beta-blocker, MRA, and dapagliflozin or empagliflozin. Add loop diuretics for congestion. [p. 3620]
- ICD: Class I if ischaemic, IIa if non-ischaemic. LVEF ≤35% after ≥3 months of optimal therapy. [p. 3625]
- CRT: Class I for LBBB with QRS ≥150 ms. IIa for LBBB 130–149 ms or non-LBBB ≥150 ms. Not for QRS <130. [p. 3627]
- HFmrEF: everything except diuretics was IIb in 2021 (SGLT2i is now Class I under the 2023 update). [p. 3629]
- Acute HF: check for cardiogenic shock or respiratory failure first, then a CHAMPIT cause. Use IV loop diuretics and judge the response by urine sodium and urine output. [p. 3653–3654]
Definitions
Adapted from ESC 2021 Table 3 [p. 3613]
| HFrEF | HFmrEF | HFpEF | |
|---|---|---|---|
| 1 | Symptoms ± signs | Symptoms ± signs | Symptoms ± signs |
| 2 | LVEF ≤40% | LVEF 41–49% | LVEF ≥50% |
| 3 | – | – | Objective evidence of structural or functional abnormality consistent with LV diastolic dysfunction or raised filling pressure, including raised natriuretic peptides |
For HFmrEF, other evidence of structural heart disease (larger left atrium, LV hypertrophy, abnormal LV filling) makes the diagnosis more likely.
Diagnosis
flowchart TD
A["Suspected HF<br/>risk factors · symptoms and/or signs · abnormal ECG"] --> B{"NT-proBNP ≥125 pg/mL<br/>or BNP ≥35 pg/mL?"}
B -- No --> U["HF unlikely<br/>consider other diagnoses"]
B -- "Yes (or HF strongly suspected,<br/>or NP not available)" --> E["Echocardiography"]
E --> F{"Abnormal findings?"}
F -- No --> U
F -- Yes --> C["HF confirmed<br/>define phenotype by LVEF"]
C --> R["≤40%<br/>HFrEF"]
C --> M["41–49%<br/>HFmrEF"]
C --> P["≥50%<br/>HFpEF"]
R --> T["Find the cause and start treatment"]
M --> T
P --> T
Adapted from ESC 2021 Figure 1 [p. 3615]
HFrEF
flowchart TD
A["Class I for all with HFrEF:<br/>ACE-I (or ARNI as replacement) · beta-blocker · MRA<br/>dapagliflozin / empagliflozin · loop diuretic for fluid retention"] --> B{"Device criteria"}
B -- "LVEF ≤35%, QRS <130 ms,<br/>where appropriate" --> I["ICD<br/>ischaemic: Class I<br/>non-ischaemic: Class IIa"]
B -- "LVEF >35%, or device<br/>not indicated or appropriate" --> S
B -- "Sinus rhythm, LVEF ≤35%,<br/>QRS ≥130 ms" --> C["CRT-D / CRT-P<br/>QRS 130–149 ms: IIa<br/>QRS ≥150 ms: I"]
I --> S["If symptoms persist,<br/>consider Class II therapies"]
C --> S
Adapted from ESC 2021 Figure 2 [p. 3620]
Class I: drugs for all patients with HFrEF (NYHA II–IV) [p. 3620]
Class I: in selected patients [p. 3622]
Class IIa [p. 3622]
- Ivabradine: LVEF ≤35%, sinus rhythm, resting HR ≥70 despite the maximum tolerated beta-blocker plus ACE-I/ARNI and MRA [IIa, B]. Also when a beta-blocker can't be used [IIa, C].
- Hydralazine + isosorbide dinitrate: self-identified Black patients with LVEF ≤35% (or <45% with a dilated LV), NYHA III–IV, despite ACE-I/ARNI, beta-blocker and MRA. [IIa, B]
Class IIb [p. 3622–3623]
- Vericiguat: NYHA II–IV with worsening HF despite standard therapy. [IIb, B]
- Hydralazine + isosorbide dinitrate if an ACE-I, ARB and ARNI can't be used. [IIb, B]
- Digoxin in sinus rhythm despite standard therapy, to reduce admissions. [IIb, B]
- Sacubitril/valsartan started without a previous ACE-I (de novo). [IIb, B · text p. 3622]
Starting ARNI
The patient needs adequate BP and eGFR ≥30. Leave a 36-hour washout after the last ACE-I dose. [p. 3622]
Devices
ICD: Class I [p. 3625]
- Secondary prevention: after recovering from a ventricular arrhythmia that caused haemodynamic instability, with expected survival >1 year in good functional status. Exclusions: a reversible cause, or the arrhythmia occurred <48 h after an MI. [I, A]
- Primary prevention, ischaemic: NYHA II–III, LVEF ≤35% despite ≥3 months of optimal therapy, and not within 40 days of an MI. [I, A]
ICD: Class IIa
CRT: Class I [p. 3627]
CRT: Class IIa
Class III
HFmrEF and HFpEF (2021)
Adapted from ESC 2021 p. 3629 [p. 3629]
| Drug | Class |
|---|---|
| Diuretics for congestion | I, C |
| ACE-I | IIb, C |
| ARB | IIb, C |
| Beta-blocker | IIb, C |
| MRA | IIb, C |
| Sacubitril/valsartan | IIb, C |
Updated in 2023: SGLT2i is now Class I, A.
- Screen for and treat the cause, plus cardiovascular and non-cardiovascular comorbidities (Class I).
- Diuretics for congestion (Class I).
- Updated in 2023: SGLT2i is now Class I, A. See the 2023 update.
Acute heart failure
flowchart TD
A["Suspected acute HF<br/>(urgent phase after first contact)"] --> B{"Cardiogenic shock and/or<br/>respiratory failure?"}
B -- Yes --> S["Pharmacological support<br/>ventilatory support<br/>mechanical circulatory support"]
B -- No --> C["Immediate phase (first 60–120 min):<br/>identify the acute cause — CHAMPIT"]
C --> D{"C acute coronary syndrome<br/>H hypertensive emergency<br/>A arrhythmia<br/>M mechanical cause<br/>P pulmonary embolism<br/>I infection<br/>T tamponade"}
D -- Yes --> T["Start specific treatment straight away"]
D -- No --> F["Further treatment by clinical presentation"]
Adapted from ESC 2021 Figure 12 [p. 3653]
Four clinical presentations
Adapted from ESC 2021 Table 21 [p. 3647]
| Acute decompensated HF | Acute pulmonary oedema | Isolated RV failure | Cardiogenic shock | |
|---|---|---|---|---|
| Mechanism | LV dysfunction, sodium and water retention | Raised afterload and/or LV diastolic dysfunction, valve disease | RV dysfunction and/or pre-capillary pulmonary hypertension | Severe cardiac dysfunction |
| Onset | Gradual (days) | Rapid (hours) | Gradual or rapid | Gradual or rapid |
| Blood pressure | Normal to low | Normal to high | Low | Low |
| Profile | Wet–warm or wet–cold | Wet–warm | Wet–cold | Wet–cold |
| Main treatment | Diuretics; inotropes or vasopressors if hypoperfused; short-term MCS or RRT if needed | Diuretics, vasodilators | Diuretics for congestion; inotropes or vasopressors if hypoperfused | Inotropes or vasopressors, short-term MCS, RRT |
Diuretics in acute HF
flowchart TD
A{"Already on an oral loop diuretic?"} -- No --> B["IV furosemide 20–40 mg"]
A -- Yes --> C["IV dose = 1–2 × the daily oral dose"]
B --> D{"Spot urine sodium ≥50–70 mEq/L at 2 h<br/>or urine output ≥100–150 mL/h at 6 h?"}
C --> D
D -- Yes --> E["Repeat the same dose IV every 12 h"]
D -- No --> F["Double the IV dose<br/>up to the maximum IV dose"]
E --> G{"Urine sodium ≥50–70 mEq/L<br/>or urine output ≥100–150 mL/h?"}
F --> G
G -- Yes --> H["Continue until fully decongested"]
G -- No --> I["Combination diuretics<br/>(add thiazide, metolazone or acetazolamide)"]
H -.- K["Check creatinine and electrolytes at least every 24 h"]
I -.- K
Adapted from ESC 2021 Figure 13 [p. 3654]. The maximum daily IV loop dose is usually 400–600 mg of furosemide (up to 1000 mg in severe kidney impairment).
Class I: initial treatment of acute HF [p. 3655–3656]
- Oxygen if SpO₂ <90% or PaO₂ <60 mmHg. [I, C]
- Intubation for progressive respiratory failure despite oxygen or non-invasive ventilation. [I, C]
- IV loop diuretics for every patient admitted with signs or symptoms of fluid overload. [I, C]
- Thromboembolism prophylaxis (for example LMWH) if not already anticoagulated. [I, A]
Class IIa
Class IIb
Class III
Iron deficiency (2021 version)
Class I [p. 3669]
Class IIa (upgraded in 2023)
Iron deficiency = serum ferritin <100 ng/mL, or ferritin 100–299 ng/mL with TSAT <20%. Erythropoiesis-stimulating agents aren't indicated for anaemia in HF.
Clinical pearls
- ESC uses CHAMPIT for the acute causes of HF.
- In the non-acute setting, NT-proBNP <125 makes HF unlikely.
- ESC 2021 dropped "mid-range" and renamed HFmrEF "mildly reduced".
- Diuretic response after IV furosemide: urine sodium >50–70 mEq/L at 2 h, or urine output >100–150 mL/h over the first 6 h.
- Small, transient rises in creatinine during decongestion aren't linked to worse outcomes once the patient is free of congestion. When starting a RAAS inhibitor, ARNI or SGLT2i, a creatinine rise of <50% (up to 266 µmol/L) or an eGFR fall of <10% is acceptable. [p. 3669]
Classes and levels of evidence
How ESC 2021 grades its recommendations. Adapted from ESC 2021 Tables 1 and 2 [p. 3608]. Hover over any class or level cited on this page to see what it means.
| Class | Meaning | Wording used |
|---|---|---|
| Class I | Evidence and/or general agreement that a treatment or procedure is beneficial, useful and effective | is recommended or is indicated |
| Class IIa | Weight of evidence or opinion is in favour of usefulness or efficacy | should be considered |
| Class IIb | Usefulness or efficacy is less well established by evidence or opinion | may be considered |
| Class III | Evidence or general agreement that the treatment or procedure isn't useful or effective, and in some cases may be harmful | is not recommended |
| Level | Meaning |
|---|---|
| Level A | Data from multiple randomised clinical trials or meta-analyses |
| Level B | Data from a single randomised clinical trial or large non-randomised studies |
| Level C | Consensus of expert opinion and/or small studies, retrospective studies, registries |
Class II covers conflicting evidence or a divergence of opinion about usefulness or efficacy. It's split into IIa and IIb.