Heart Failure Comprehensive — Cardiology Guide

HFrEF · HFpEF · HFimpEF2026: no rigid EF cutoffs

4 pillarsHFrEF — start all before d/c

Stage C/Dsymptomatic HF

SGLT2iEvidence across the EF spectrum

▤

2026 definition · 2022 treatment

Part	Section	Pages
A · Foundations	Definitions / stages / NYHA / EF phenotypes · diagnosis & work-up · etiology & don't-miss causes · diagnostic testing · acute-HF triage	pp 2–4 + more
B · HFrEF	Four-pillar GDMT (overview → discharge implementation) · drug detail · add-ons (iron, ivabradine, H-ISDN, vericiguat, digoxin) · devices · when to refer	pp 5, 21–23, 28
C · Mildly reduced / HFpEF	LVEF 41–49% mgmt · HFpEF dx (H2FPEF · HFA-PEFF) · HFpEF mgmt (SGLT2i · finerenone · MRA) · HFpEF mimics	pp 6–9
D · Ischemic CM & revascularization	Ischemic CM + viability · revasc indications (CABG vs PCI in low EF) · secondary prevention	pp 10–12
E · Structural heart	Functional MR / TEER (COAPT vs MITRA-FR) · AS / TAVR (LF/LG) · TR · LAA closure · ASD/PFO in HF	pp 13–15
F · Electrophysiology	ICD (SCD-HeFT · DANISH) · CRT / conduction-system pacing · AF in HF (CASTLE-AF) · VT / PVC-induced CM	pp 16–18
G · Inpatient mgmt	ADHF (DOSE · ADVOR · CLOROTIC) · cardiogenic shock (SCAI · DanGer · tMCS) · inpatient GDMT + discharge (STRONG-HF) · sodium / fluid restriction · oral sodium & hypertonic saline	pp 19–21, 45–46
Reference	HFrEF drug detail & add-ons · RHC/hemodynamics · exam · acute-HF ED · discharge GDMT, adverse events, cost & access · iron deficiency, IV formulations & delivery	pp 22–29
H · Devices	Device atlas & DRI²P²S · ICD / CRT / conduction-system pacing / WCD · PA sensors · Barostim / CCM · CSA · research devices	pp 30–44

The HF hub integrates discharge GDMT and access (P.28), iron deficiency (P.29), and HF devices / DRI²P²S (PP.30–44). Focused companion guides cover the etiologies below.

↦

Cross-guide map — when to open a focused guide

public companion topics

If the HF is from…	Use the focused public guide
Thick walls / low-flow, ↑ troponin	Cardiac Amyloidosis (ATTR/AL) · HCM (obstructive / myosin inhibitors)
Granuloma / heart block / VT	Cardiac Sarcoidosis (FDG-PET, immunosuppression)
Familial / genetic DCM · ARVC · LVNC	Genetic Cardiomyopathies hub · ARVC · LV Non-Compaction
Iron / Fabry / storage / EMF	Infiltrative Cardiomyopathies hub
Group 2 pulmonary hypertension	Pulmonary Hypertension (don't give PAH drugs in HFpEF)

①

First moves for any new HF patient

a 5-step approach

#	Do	Why / how
1	Confirm HF + measure LVEF	Symptoms/signs + NT-proBNP + echo → set the phenotype (P.2). Phenotype decides the drugs.
2	Find the cause	Ischemic? Amyloid / HCM / sarcoid? Reversible (tachy, alcohol, thyroid, valve, PVCs)? (P.4) — many have specific therapy.
3	Start phenotype-based therapy	HFrEF → the four pillars (P.5). HFpEF → SGLT2i + finerenone + treat comorbidities. Mildly reduced (41–49%) → P.6.
4	Decongest & assess perfusion	Wet/dry · warm/cold; loop diuretic for congestion; protect the cardiorenal axis.
5	Risk-stratify & refer	NYHA, biomarkers, peak VO₂; recurrent admissions / inotropes / EF ≤ 35% → advanced-HF referral.

USE

Reference for trained clinicians — verify doses against current labeling and current guidelines. Built on the 2026 Second Universal Definition of HF, the 2022 AHA/ACC/HFSA treatment guideline, the 2021 revascularization guideline & cited trials.

Heart failure — 2026 universal definition

symptoms/signs + objective evidence

HF is a clinical syndrome of symptoms ± signs caused by structural/functional cardiac abnormality, corroborated by at least one of: elevated natriuretic peptides, or objective evidence of cardiogenic congestion (imaging or hemodynamics). HF is not a single diagnosis — always pursue the cause (P.4) and the EF phenotype (below), which drive therapy.

Burden: HF affects ~2.7% of US adults, projected to rise to 3.8% by 2050 as the population ages; and post-COVID, age-adjusted HF mortality remains elevated — the pandemic ended, but the HF epidemic continues.

ACC/AHA stages — progression & trajectories

stage drives prevention vs treatment

Stage	Definition	Focus
A · At risk	Risk factors (HTN, DM, CAD, obesity, CKD, cardiotoxins, FHx CM), no structural disease or symptoms	Risk-factor control — BP, SGLT2i in DM, GLP-1 RA in obesity, lipids, lifestyle; screen at-risk families
B · Pre-HF	Structural disease (↓EF, ↑filling pressure, valve disease) or ↑ biomarkers, no symptoms ever	LVEF ≤40%: ACEi and evidence-based beta-blocker. ARB in selected ACEi intolerance. ARNI is not a routine stage-B recommendation
C · Symptomatic	Structural disease + current or prior HF symptoms	GDMT by phenotype (P.5) + devices + diuretics; most of this guide
D · Advanced	Refractory symptoms despite GDMT — recurrent hospitalizations, inotrope-dependence, intolerance to GDMT	Refer for transplant / LVAD / palliative care → LVAD & Transplant cards

Stages are unidirectional — a patient who improves is still "Stage C," never back to A/B. Within Stage C, three trajectories are now recognized: Improvement (LVEF ↑ but structural abnormality persists), Remission (normalized LVEF + minimal symptoms + stable biomarkers), and Recovery (sustained normalization — a minority). "Stable HF" is a misnomer — always reassess trajectory.

EF phenotypes — 2026 classification

phenotype guides therapy

Phenotype	LVEF	Therapy headline
HFrEF	reduced	Full four-pillar GDMT (ARNI · BB · MRA · SGLT2i) — foundational outcome-improving HFrEF therapy (P.5). LVEF < 50% is unlikely to reflect normal function.
HFpEF	preserved	SGLT2i + finerenone + diuretics + aggressively treat comorbidities & the mimics (amyloid, HCM, constriction). See P.7–8.
HFimpEF	improved	Previously reduced, now improved — continue GDMT (TRED-HF: withdrawal → relapse). Not "recovered" unless sustained normalization.

The 2026 definition eliminates rigid LVEF cutoffs (≤40 / 41–49 / ≥50) and removes HFmrEF as a separate category. An LVEF of 41–49% ("mildly reduced") still exists clinically — treatment evidence for that range is on P.6. The 2022 AHA/ACC/HFSA treatment recommendations remain in force.

NYHA functional class — track it every visit

symptom burden, not stage

I No limitation; ordinary activity no symptoms.
II Slight limitation; comfortable at rest, ordinary activity → symptoms.

III Marked limitation; less-than-ordinary activity → symptoms.
IV Symptoms at rest / any activity.

NYHA fluctuates and guides several thresholds (e.g., CRT, transplant referral, palliative care). It is not the same as ACC/AHA stage — a Stage C patient can be NYHA I on GDMT.

Prognosis & the advanced-HF triggers

know when GDMT isn't enough

Risk tools & markers

Seattle HF Model · MAGGIC estimate 1-yr mortality — use for ICD timing & advanced-HF conversations.
Bad-trajectory markers persistently ↑ NT-proBNP, low SBP, ↑ creatinine, hyponatremia, recurrent admissions, GDMT intolerance.
Peak VO₂ < 14 (off β-blocker) / < 12 (on β-blocker) → transplant range.

"I NEED HELP" — refer to advanced HF

I IV inotropes · N NYHA IIIB–IV / persistent NPs · E end-organ dysfunction · E EF ≤ 35%.
D defibrillator shocks · H ≥ 1 HF hospitalization/yr · E edema / escalating diuretics · L low BP, high HR · P prognostic-med intolerance.

HFimpEF — do not de-escalate. Withdrawing GDMT from a recovered HFrEF patient (TRED-HF) frequently relapses. Continue all four pillars even when the EF normalizes.

The new-HF work-up — everyone gets these

2022 guideline core

Test	What it tells you
Natriuretic peptide	BNP / NT-proBNP — supports/excludes HF & prognosticates. Lower with obesity; higher with age, AF, renal dysfunction, ARNI (BNP only). Use serial trends, not a single value.
12-lead ECG	Ischemia/infarct, LVH, conduction (LBBB → CRT candidacy), arrhythmia, low voltage + pseudo-infarct (→ amyloid).
Echocardiogram	LVEF + phenotype, chamber size, wall thickness, diastolic function/filling pressures (E/e′), valves, RV function, pericardium, strain (apical sparing → amyloid).
Labs	CBC, CMP, iron studies (ferritin / TSAT), TSH, HbA1c, lipids, ± HIV/iron/TTR per clues. Check iron in every HFrEF (P.5).
CXR	Congestion, cardiomegaly, alternative pulmonary cause.
Ischemic eval	Assess for CAD in all new HFrEF with risk factors — coronary angiography (preferred when revasc would change management) or CTA / functional imaging (P.D, forthcoming).

Natriuretic peptides — read them right

rule-out is the strength

Acute (dyspnea in the ED)

Rule-out BNP < 100 / NT-proBNP < 300 pg/mL makes acute HF unlikely.
Rule-in (NT-proBNP, age-stratified) > 450 (<50 y) · > 900 (50–75) · > 1800 (>75 y).

Confounders

↓ Falsely low obesity (a "normal" value can still be HF).
↑ Falsely high age, renal failure, AF, PE/RV strain, sepsis.
ARNI raises BNP (substrate), lowers NT-proBNP — follow NT-proBNP on ARNI.

When to go further

advanced testing & referral

Add this test	When	Looking for
Cardiac MRI	Unexplained CM, suspected infiltration/inflammation, discordant data	Scar/LGE pattern → ischemic vs amyloid vs sarcoid vs myocarditis → cards
PYP + complete monoclonal screen	↑ wall thickness + low-flow, ≥ 65 y, apical sparing	AL screen first: serum/urine immunofixation + serum free light chains. PYP/SPECT can establish ATTR only after AL is excluded
FDG-PET	AV block in young/middle-age, VT, patchy CM	Cardiac sarcoidosis → Sarcoid card
Genetic testing	FHx CM/SCD, DCM/ARVC/LVNC/HCM phenotype	Familial CM → Genetic CM card
CPET (VO₂)	Advanced HF, transplant/LVAD eval	Peak VO₂ for prognosis / listing → Transplant card
Endomyocardial biopsy	Fulminant/rapidly progressive, suspected giant-cell/eosinophilic myocarditis	Specific myocardial diagnosis (selective)

Echo — what to extract beyond the EF

structure · filling · RV · valves

Domain	Key reads
Diastolic / filling	E/e′ > 14, LA volume index > 34, TR velocity > 2.8 m/s → elevated filling pressure (supports HFpEF).
RV function	TAPSE < 17 abnormal; S′, RV size, estimated RVSP — RV failure worsens prognosis & complicates LVAD.
Strain	Global longitudinal strain (early dysfunction); apical sparing pattern → amyloid.
Valves / pericardium	Functional MR & TR severity (→ TEER), AS incl. low-flow low-gradient, septal bounce / thickening → constriction.

Right-heart catheterization — when the numbers matter

don't guess in shock / advanced HF

When: uncertain volume/perfusion, cardiogenic shock, high filling pressures discordant with exam, advanced-therapy evaluation, or pulmonary-hypertension work-up. Read: RA, PCWP, PA, cardiac output/index, PVR, PAPi (RV function), cardiac power output (CPO < 0.6 W → high-risk in shock). → Use a specialist shock pathway for device selection.

HF is a final common pathway, not a diagnosis. Always identify the cause — many are reversible or have specific therapy, and the phenotype + cause decide treatment. Re-image / re-think when a patient doesn't respond to GDMT. The 2026 universal definition introduces a standardized classification of HF causes (below).

Universal classification of HF causes & where to take them

2026 pathogenic groups

Category	Causes / clues	Go to
Ischemic	Commonest cause of HFrEF. Prior MI, regional WMA, CAD. Assess viability & revascularization.	→ Part D (revasc)
Hypertensive	LVH, longstanding HTN → concentric remodeling, diastolic dysfunction, HFpEF.	→ Part C (HFpEF)
Valvular	AS/AR/MR/MS — primary valve disease causing LV dysfunction. Treat the valve.	→ Part E (structural)
Arrhythmia-related	Persistent tachycardia (AF/flutter/SVT) or frequent PVCs → reversible CM.	→ Part F (EP)
Infiltrative	Amyloid (↑ wall thickness, low voltage, apical sparing); Fabry; hemochromatosis; glycogen storage.	→ Cardiac Amyloidosis · Infiltrative CM hub
Infective	Chagas, HIV, viral myocarditis (acute/chronic).	→ Infective CM card
Inflammatory	Sarcoid (heart block, VT, patchy LGE/FDG); giant-cell; eosinophilic; autoimmune.	→ Cardiac Sarcoidosis
Toxic	Anthracyclines, HER2, ICI myocarditis, radiation; alcohol, cocaine/meth.	→ Specialist assessment · Toxic CM guide
Heritable / genetic	FHx DCM/SCD; lamin/titin/desmosomal; ARVC; LVNC; HCM.	→ Genetic CM hub · HCM
Metabolic / nutritional	Thyrotoxicosis, hypothyroid; beriberi (thiamine), selenium; obesity CM.	→ Metabolic · High-Output card
Pregnancy-related	Last month → 5 mo postpartum; bromocriptine in selected cases.	→ Peripartum CM card
Stress-induced	Takotsubo (apical ballooning, emotional/physical trigger).	→ Takotsubo card
Pericardial	Constrictive pericarditis, large effusion/tamponade.	→ Pericardial Tamponade card
Pulmonary / RV	Group 2 PH; RV failure from PE, CTEPH, or pulmonary disease.	→ PH card · RV Failure card
High-output	Anemia, AV fistula/shunt, Paget, severe liver disease.	→ Metabolic · High-Output card
Congenital	Residual shunts (ASD/VSD), single-ventricle physiology, Fontan failure.	→ ACHD-HF card

Reversible causes worth chasing in every new CM

"CHASE-IT"

Ischemia treat ACS/angina and assess CABG eligibility. Viability alone does not select a survival benefit.
Tachy/PVC rate/rhythm control, PVC ablation.
Valve repair/replace severe AS/MR.
Alcohol / toxins abstinence.

Thyroid treat hyper-/hypothyroidism.
Iron hemochromatosis (overload) — phlebotomy.
Nutritional thiamine (wet beriberi), selenium.
Sleep apnea treat OSA; rate-control AF.

Red flags → the targeted test

don't anchor on "idiopathic DCM"

Clue	Think / do
↑ wall thickness + low ECG voltage	Amyloid → serum/urine immunofixation + serum free light chains, then PYP/SPECT when appropriate
AV block or VT in young–middle age	Sarcoid → cardiac FDG-PET / MRI
Family history of DCM or SCD	Genetic CM → gene panel + cascade family screening
Frequent PVCs / persistent tachycardia	Arrhythmia-mediated CM → rate / rhythm control, PVC ablation (often reversible)
Peripartum window (last mo → 5 mo post)	PPCM → anticoagulate if EF < 35%; bromocriptine in selected cases
Fulminant / rapidly progressive	Giant-cell / eosinophilic myocarditis → biopsy; may need MCS

Takotsubo (apical ballooning after an emotional/physical trigger) usually recovers in weeks — supportive care, watch for LVOT obstruction & shock (avoid inotropes if obstructive). Tachycardia-/PVC-mediated CM can fully recover with rate/rhythm control — re-image at 3–6 mo before calling it permanent.

The four pillars are the foundation of outcome-improving HFrEF treatment. For the at-a-glance table — doses, actual trial AE rates, contraindications, and cost — use P.28, Discharge GDMT at a Glance. This page is the how-to: sequencing, titration, and the hard parts (low BP, CKD, hyperkalemia).

The four pillars — all are Class 1 for HFrEF outcomes

LVEF ≤ 40%

Pillar	Start → target	Headline benefit (vs comparator)
1 ARNI (or ACEi/ARB)	sac/val 24/26–49/51 → 97/103 BID	CV death/HF hosp ↓20% vs enalapril (PARADIGM-HF).
2 Beta-blocker	carvedilol / metop succ / bisoprolol → target	~34% lower mortality vs placebo (3 trials).
3 MRA	spironolactone / eplerenone 25 → 50 qd	Mortality ↓30% (RALES); CV death/HF hosp HR 0.63 (EMPHASIS).
4 SGLT2 inhibitor	dapagliflozin / empagliflozin 10 mg qd	Worsening HF/CV death ↓~25%, any diabetes status.

The utilization gap is lethal. Only 18% of eligible patients receive all four pillars; <1% reach target doses. Median time from diagnosis to quad therapy: 109 days. Modeling estimates full quadruple therapy could prevent >1.1 million HFrEF deaths/yr worldwide — at an ICER well under $50K/QALY. A dedicated GDMT clinic raised quad-therapy use from 21% → 88% in 16 wk.

Sequencing & titration — get all four on, fast

coverage beats up-titration of one

Principles

Start low doses of all four within days–weeks (in-hospital if possible) — STRONG-HF: rapid up-titration cut death/readmission.
Order is flexible; SGLT2i and low-dose BB/ARNI/MRA can start close together after stabilization and individual safety checks.
Up-titrate q1–2 wk to target or max tolerated; recheck BP · HR · K⁺ · Cr after each step.
Don't stop GDMT for asymptomatic ↓BP, mild ↑Cr (<30%), or HR ≥ 50.

When it's hard

BP is a budget symptoms and perfusion—not the cuff alone—decide whether it is overspent. If euvolemic, remove excess loop diuretic and non-outcome vasodilators before abandoning outcome-modifying therapy when feasible; persistent GDMT intolerance is itself an advanced-HF signal.
CKD / ↑K⁺ SGLT2i still indicated to low eGFR; both MRAs raise K⁺. Recheck labs and review interacting drugs. If K⁺ cannot remain <5.5, stop the MRA.
Bradycardia reduce BB before stopping; consider need for the others.
Gynecomastia switch spironolactone → eplerenone.

Then layer on — by indication

add-ons & devices

Add	When
IV iron	Check iron in every HFrEF. Ferritin <100 (or 100–300 with TSAT <20%) → IV iron improves symptoms / ↓ hospitalization.
Diuretic	Loop diuretic for congestion (symptom relief; lowest effective dose).
Other add-ons	Ivabradine (sinus ≥70 on max BB) · hydralazine+ISDN (self-identified Black NYHA III–IV) · vericiguat (recent worsening) · digoxin · IV iron → add-on reference P.23.
Devices	ICD if EF ≤35% & >40 d post-MI / >3 mo on GDMT; CRT if EF ≤35% + LBBB ≥150 ms → Part F (EP) P.16.
Refer (Stage D)	Recurrent admissions, inotrope need, intolerance to GDMT, low VO₂ → advanced HF eval → LVAD / Transplant / ECMO cards.

Target doses — titrate toward these

the trial-proven targets

Drug	Start	Target
Sacubitril/valsartan	24/26–49/51 BID	97/103 mg BID
Carvedilol	3.125 mg BID	25 mg BID (50 if > 85 kg)
Metoprolol succinate	12.5–25 mg qd	200 mg qd
Bisoprolol	1.25 mg qd	10 mg qd
Spironolactone / eplerenone	12.5–25 mg qd	25–50 mg qd
Dapagliflozin / empagliflozin	10 mg qd	10 mg qd (no titration)

Monitoring — after start & each up-titration

labs + exam

BMP (K⁺ / Cr) 1–2 wk after any ARNI/ACEi/MRA/diuretic change; Promptly evaluate K⁺ ≥5.5 or a substantial Cr rise. Hold/adjust therapy for dangerous hyperkalemia, AKI or hypoperfusion.
BP / HR each visit; Assess symptoms, orthostasis and perfusion. A BP/HR number alone does not establish titration safety.

Weight / congestion daily home weights; titrate diuretic to euvolemia.
NT-proBNP follow the trend (on ARNI, use NT-proBNP, not BNP).

LVEF 41–49% — the mildly reduced range. The 2026 universal definition eliminates "HFmrEF" as a formal category, but the treatment evidence for this EF range remains. Many patients are recovering HFrEF or early HFpEF. Trajectory decides behavior: a falling EF acts like HFrEF, a rising one like HFimpEF. Treat the lower-EF end (41–44%, or recently reduced) like HFrEF; use an SGLT2i unless contraindicated. Always hunt & treat the cause (P.4).

GDMT for LVEF 41–49% — what's proven

SGLT2i: ESC 2023 I, US 2022 2a

Therapy	Class	Evidence & use
SGLT2 inhibitor	ESC I US 2a	DELIVER & EMPEROR-Preserved enrolled EF > 40% (incl. mildly reduced EF): ↓ HF hospitalization / CV death (HR ~0.79–0.82). Use when eligible, with renal, volume and ketoacidosis safety checks.
ARNI / ACEi-ARB	2b	PARAGON-HF (EF ≥ 45%) narrowly missed overall — benefit greatest at the lower-EF end & in women. Reasonable, esp. EF 41–44%.
MRA	2b	TOPCAT: reasonable if eGFR > 30 & K⁺ < 5.0, esp. with elevated filling pressures. Finerenone — FINEARTS-HF included EF ≥ 40%.
Beta-blocker	2b	No dedicated EF 41–49% mortality trial; reasonable for post-MI, lower EF, or AF rate control.
Loop diuretic	1	For congestion — symptom relief.

Practical approach

trajectory & cause

Treat toward HFrEF if EF 41–44% or recently fell — full four-pillar consideration (P.5).
Always SGLT2i + loop diuretic for congestion + BP / AF / comorbidity control.
Re-image EF at 3–6 mo — recovering (→ HFimpEF, continue GDMT) or declining (→ HFrEF)?

Find the cause ischemic? infiltrative? arrhythmia-mediated? — the same don't-miss list (P.4).
Devices ICD/CRT thresholds are EF ≤ 35% — usually not met in this EF range; reassess if EF falls.
HFpEF overlap if EF rises ≥ 50%, manage as HFpEF (P.7–8).

Why these classes — the mildly reduced EF evidence

mostly subgroups & extrapolation

Trial	What it showed for EF 41–49%
DELIVER · EMPEROR-Preserved	Enrolled EF > 40% — consistent ↓ HF hospitalization across the EF spectrum, including the mildly reduced range. The firmest evidence.
PARAGON-HF (ARNI)	EF ≥ 45%; treatment-by-EF interaction — benefit concentrated at the lower-EF end & in women.
TOPCAT (spironolactone)	Neutral overall; signal in the lower-EF / Americas subgroup — supports MRA as reasonable.
Beta-blockers	No EF 41–49%–specific RCT; individual-patient meta-analysis suggested mortality benefit in sinus rhythm (not AF).

Bottom line: SGLT2i for eligible patients, layer ARNI/ACEi-ARB, MRA, and a beta-blocker (esp. sinus rhythm, post-MI, or EF at the lower end) — the weight of benefit rises as the EF falls toward 40%. Re-check the EF: many patients in this range move into HFrEF or HFpEF over time — assess trajectory.

Don't forget — secondary prevention & devices

check ischemia and blood pressure

Secondary prevention

Ischemia statin + antiplatelet; revascularize demonstrable ischemia (→ Part D).
Risk factors BP, glycemia, weight, smoking; treat AF, OSA, iron deficiency.
Rehab cardiac rehab improves function & outcomes.

Devices

ICD / CRT thresholds need EF ≤ 35% — not met in most patients with EF 41–49%; reassess if the EF declines.
Re-evaluate repeat echo at 3–6 mo & after revascularization — the EF (and the category) often changes.

HFimpEF — improved EF is not "cured." The 2026 definition distinguishes improvement (LVEF ↑ but structural abnormality persists), remission (normalized LVEF + minimal symptoms + stable biomarkers), and recovery (sustained normalization — a minority). In TRED-HF, withdrawing therapy from recovered DCM patients with a normalized EF caused ~40% to relapse within 6 months. Recovery reflects treatment effect, not a healed myocardium — do not de-escalate the four pillars just because the EF looks normal.

HFpEF = HF symptoms/signs + LVEF ≥ 50% + objective evidence of elevated filling pressure. A clinical diagnosis that is easily missed — "preserved EF" is not "no heart failure." Confirm filling pressures (NPs, echo, or exercise hemodynamics) and always exclude the mimics (P.9).

H2FPEF score — quick bedside probability

add the points

Pts	Variable	Threshold
2	Heavy	BMI > 30 kg/m²
1	Hypertensive	≥ 2 antihypertensive medications
3	Atrial fibrillation	Paroxysmal or persistent
1	Pulmonary hypertension	Echo PASP > 35 mmHg
1	Elder	Age > 60 y
1	Filling pressure	Echo E/e′ > 9

0–1 HFpEF unlikely · 2–5 intermediate → further testing · 6–9 HFpEF likely.

HFA-PEFF score — domains, cutoffs & interpretation

major = 2 · minor = 1 · max 6

Domain	Major — 2 pts	Minor — 1 pt
Functional	Age <75: septal e′ <7 or lateral <10. Age ≥75: <5 or <7 cm/s. Or avg E/e′ ≥15 or TR velocity >2.8 m/s	avg E/e′ 9–14; or GLS < 16%
Morphological	LAVI > 34 (SR) / > 40 (AF) mL/m²; or LV mass index ≥ 149♂/122♀ g/m² with RWT > 0.42	LAVI 29–34 (SR) / 34–40 (AF); or LVMI > 115♂/95♀; or RWT > 0.42; or LV wall ≥ 12 mm
Biomarker (sinus)	NT-proBNP > 220 or BNP > 80 pg/mL	NT-proBNP 125–220 or BNP 35–80 pg/mL
Biomarker (AF)	NT-proBNP > 660 or BNP > 240 pg/mL	NT-proBNP 375–660 or BNP 105–240 pg/mL

Take the highest hit per domain (major or minor, not additive); sum the 3 domains (functional · morphological · biomarker; biomarker uses the sinus or AF row) — max 6. ≥ 5 = HFpEF confirmed · 2–4 = intermediate → step F1 (diastolic stress echo or invasive exercise PCWP, §C) · ≤ 1 = unlikely. Steps: P pretest screen → E this score → F1 functional test → F2 establish etiology (screen mimics, P.9).

Invasive hemodynamics — rest & provocation

the gold standard when equivocal

Maneuver	Diagnostic of HFpEF
Rest	PCWP ≥15 mmHg
Fluid challenge (500 cc / 5 min)	PCWP > 18 mmHg
Passive leg raise (within 3 min)	PCWP > 19 mmHg
Exercise (position matters)	Supine PCWP ≥25, upright ≥20 mmHg, or PCWP/CO slope >2 mmHg·min·L⁻¹

Unmasks occult HFpEF when resting numbers look normal; defines the pulmonary-vascular response — Group 2 PH, not PAH (no PAH drugs). The ΔPCWP/ΔCO slope holds even with submaximal effort.

Reproduce the symptom, then explain it. HFpEF may look compensated at rest yet require an abnormal rise in left-sided filling pressure to deliver flow during ordinary activity. A wedge that rises with exercise explains why exertion feels awful; it does not by itself prescribe one drug. Treat congestion, BP, AF, ischemia, obesity, sleep-disordered breathing, CKD, and deconditioning, then build activity slowly and consistently. Provocative hemodynamics are specialized testing, not screening.

Biomarkers & the dyspnea differential

HFpEF NPs can be normal

Natriuretic peptides in HFpEF

Often lower than in HFrEF; up to ~⅓ of HFpEF have a "normal" BNP — a normal NP does not exclude HFpEF, especially in the obese.
Wall stress is lower (small, thick LV); AF raises NPs, obesity lowers them — interpret with the clinical picture.

Is it really HFpEF?

Exclude deconditioning, obesity, lung disease, anemia, CKD-volume, and the mimics (P.9) before committing to the label.
A diastolic stress echo (E/e′ & TR velocity with exercise) or invasive exercise PCWP settles equivocal cases.

Diastolic stress echo (positive): 2025 ASE: average E/e′ ≥14 (or septal ≥15) and peak TR velocity >3.2 m/s — a non-invasive surrogate for the exercise PCWP rise.

Grading diastolic function (when EF is normal)

Use the complete 2025 ASE algorithm

Parameter	Reference finding	Abnormal clue
E/e′ (average)	< 10	> 14
LA volume index	≤ 34 mL/m²	> 34
TR velocity	≤ 2.8 m/s	> 2.8
Septal e′	≥ 7 cm/s	<7: impaired relaxation

These are contextual echo clues, not a stand-alone grading rule. Low e′ indicates impaired relaxation, not necessarily high pressure. Integrate mitral inflow, E/e′, TR velocity and LA measures using the 2025 ASE algorithm. AF, valve disease and pacing require specific pathways.

What to start

2026 ACC pathway + 2022 US/2023 ESC classes

Therapy	Class	Evidence & use
SGLT2 inhibitor 10 mg qd	ESC I US 2a	The cornerstone — DELIVER · EMPEROR-Preserved: ↓ HF hospitalization (HR ~0.79–0.82).
Finerenone 20–40 mg	FDA ACC pathway	FINEARTS-HF (EF ≥ 40%): ↓ worsening-HF / CV death (RR 0.84). Watch K⁺.
MRA (spironolactone)	2b	TOPCAT: neutral overall, benefit in the Americas / higher-NP patients — reasonable if eGFR > 30, K⁺ < 5.0.
ARNI	2b	PARAGON-HF — reasonable in the lower-EF range & in women.
Loop diuretic	1	For congestion — but HFpEF is preload-sensitive; over-diuresis drops output. Lowest effective dose.

Treat the comorbidities — that is most of HFpEF

the highest-yield work

Target	Do
Obesity	STEP-HFpEF: semaglutide improved symptoms, weight & exercise capacity in obese HFpEF — strongly consider a GLP-1 RA.
Hypertension	Control to guideline BP targets — the dominant driver; avoid over-treating into hypotension.
Atrial fibrillation	Rate and consider rhythm control / ablation — AF is poorly tolerated (loss of atrial kick); anticoagulate per CHA₂DS₂-VASc.
CKD · diabetes	SGLT2i + finerenone do double duty; optimize glycemia.
OSA · ischemia · deconditioning	Treat sleep apnea; evaluate ischemia if present; cardiac rehab / exercise improves capacity.

Phenotype the HFpEF — the driver shapes therapy

more than any single drug

Phenotype	Lead with
Obese-metabolic	GLP-1 RA (STEP-HFpEF) + SGLT2i + weight management.
AF-driven	Rhythm control / ablation; rate control; anticoagulate per CHA₂DS₂-VASc.
CAD / ischemic	Evaluate & revascularize demonstrable ischemia (→ Part D).
Hypertensive	Aggressive BP control; SGLT2i + finerenone / MRA.
RH / PH-predominant	Volume management + treat the cause; not PAH drugs (Group 2 → PH card).
"Doesn't fit"	Hunt a mimic — amyloid, HCM, constriction (P.9).

Tried & neutral/negative in HFpEF: routine β-blocker (use only for AF/CAD), digoxin, nitrates (NEAT-HFpEF ↓ activity), PDE5 inhibitors (RELAX), and ACEi/ARB alone (CHARM-Preserved, I-PRESERVE, PEP-CHF). The evidence-based moves are SGLT2i, finerenone, ± MRA / ARNI.

Avoid: aggressive preload/afterload reduction & over-diuresis (a stiff, preload-dependent ventricle); PAH-specific drugs in Group 2 PH (→ Pulmonary Hypertension card); rate-control to bradycardia (need diastolic filling but also atrial kick).

Practical — start & monitor

K⁺ is the limiting factor

SGLT2i 10 mg qd, no titration; hold on sick days; expect a small reversible eGFR dip.
Finerenone eGFR at initiation ≥60: 20→40 mg/day. eGFR 25–<60: 10→20 mg/day. Do not start at K⁺ >5 or eGFR <25. Recheck K⁺/eGFR 4 weeks after starting and after dose changes.

MRA if used, require K⁺ < 5.0 & eGFR > 30; recheck day 3 & 1 wk.
Don't stack two MRAs (spironolactone and finerenone) — additive hyperkalemia.

Diuretics & decongestion — the HFpEF tightrope

a stiff ventricle is preload-dependent

Lever	How to use it
Loop diuretic	Lowest effective dose. Over-diuresis drops cardiac output on the steep diastolic pressure–volume curve → fatigue, pre-renal AKI, orthostasis. Titrate to euvolemia, not dryness.
Thiazide	Useful when hypertension co-exists — treats BP and volume together; watch Na⁺/K⁺.
SGLT2 inhibitor	Outcome benefit is not explained by diuresis alone. It does not replace loop diuretics for substantial congestion.
Monitoring	Daily weights, orthostatic vitals, Cr/lytes after changes; a modest Cr rise during effective decongestion can be functional. Reassess congestion, BP, urine output and perfusion before changing therapy.
Diuretic resistance	Sequential nephron blockade (add thiazide/acetazolamide); inpatient strategies in Part G.

When "HFpEF" doesn't add up — hunt a specific mimic. Several have disease-modifying therapy that GDMT won't provide. Re-examine the echo (wall thickness, strain, pericardium) and the clinical trajectory.

The HFpEF mimics

→ go to the focused card

Mimic	Clue	Confirm / go to
Cardiac amyloid	↑ wall thickness, low voltage, apical sparing, ≥ 65 y, neuropathy / carpal tunnel	PYP + light chains → Amyloid card
HCM	Asymmetric septal hypertrophy, SAM, dynamic LVOT gradient	Echo ± provocation → HCM card
Constrictive pericarditis	Prior pericarditis / cardiac surgery / chest XRT; surgically curable	See §B; CT/MRI pericardium
Restrictive CM	Biatrial enlargement, infiltration (amyloid / sarcoid / hemochromatosis)	CMR / biopsy → Infiltrative card
High-output	Anemia, thyrotoxicosis, AV fistula, beriberi	Treat the driver
Group 2 / RV-predominant	RV failure, PASP out of proportion	→ Pulmonary Hypertension card
Valvular	Severe AS (low-flow low-gradient), MR / MS	→ Part E (structural)

Constriction vs restriction — one is surgically curable

the classic separation

Feature	Constriction	Restriction
Ventricular interaction	Respirophasic septal shift; discordant LV/RV systolic pressures	Concordant; no septal bounce
Pericardium	Thickened / calcified (CT or MRI)	Normal
Tissue Doppler e′	Preserved / increased (annulus reversus)	Reduced
NT-proBNP	Often lower	Often markedly elevated
Treatment	Pericardiectomy (curative in selected)	Treat the infiltrative cause

Mixed pictures occur (e.g., post-radiation). When in doubt, invasive hemodynamics with simultaneous LV/RV pressure tracings (± respirometry) is the arbiter.

Restrictive cardiomyopathy — the causes to chase

small ventricles, biatrial enlargement

Group	Causes & clue
Infiltrative	Amyloid (commonest — PYP / free light chains), sarcoid (FDG-PET), hemochromatosis (iron studies / MRI T2*).
Storage	Fabry (α-galactosidase, ↑ wall + low native T1), Gaucher, glycogen-storage (Danon, PRKAG2).
Endomyocardial	Hypereosinophilic / Löffler, endomyocardial fibrosis, carcinoid, radiation, drugs (anthracycline, serotonergics).
Idiopathic / familial	Often sarcomeric (troponin); family history → gene panel (→ Genetic CM card).

High-output HF (warm, bounding pulses, wide pulse pressure, low SVR): chase anemia, thyrotoxicosis, AV fistula / large shunt, beriberi (thiamine), Paget disease, cirrhosis — treating the driver resolves the heart failure.

Thick walls — read the LVH before calling it "hypertensive"

↑ wall thickness is the shared clue

Cause	Discriminating features	Next test
Hypertensive LVH	Concentric, normal/high QRS voltage, HTN history; no apical sparing — regresses with BP control.	Clinical · serial echo
Amyloid	Voltage–mass mismatch (thick walls + low voltage), apical-sparing strain, granular myocardium, thick valves/septum.	PYP + light chains
HCM	Asymmetric septal hypertrophy, SAM, dynamic LVOT gradient, FHx / SCD.	Echo ± provocation
Fabry	Concentric ↑ wall + low native T1 on CMR; X-linked, angiokeratoma, neuropathy, proteinuria.	α-galactosidase A
Athlete's heart	Mild symmetric ↑ wall, large cavity, enhanced function, sinus bradycardia — physiologic.	Detraining · echo

Amyloid quick-screen (the commonest missed mimic): in HFpEF with ↑ wall thickness + any of — age ≥ 65, low-voltage / pseudo-infarct ECG, apical-sparing strain, bilateral carpal tunnel, neuropathy, or GDMT intolerance / low BP — send serum & urine immunofixation + serum free light chains and a ⁹⁹ᵐTc-PYP scan. → Cardiac Amyloidosis card.

Don't miss constriction — it is surgically curable. Suspect after pericarditis, cardiac surgery, or chest radiation; confirm with the §B features + pericardial imaging (CT / MRI) and, if needed, invasive hemodynamics. A pericardiectomy referral can transform a "refractory HFpEF" patient.

Ischemic cardiomyopathy (ICM) = LV systolic dysfunction predominantly caused by CAD — the single commonest cause of HFrEF. It carries a worse prognosis than non-ischemic CM at the same EF, and the etiology changes management (revascularization candidacy, antiplatelet, intensive secondary prevention). Evaluate for CAD in essentially every new HFrEF unless a clear non-ischemic cause is established.

Evaluating for CAD in new HF

define the substrate

Test	Best when	Notes
Invasive coronary angiography	Angina, prior MI, high pre-test probability	Reasonable in HF + angina or significant ischemia — defines anatomy & revascularization options in one step.
Coronary CT angiography	Low–intermediate probability	High negative predictive value — a normal CTA effectively excludes ICM and avoids an invasive study.
Stress imaging (SPECT / PET / stress echo / CMR)	Intermediate probability, ischemia & viability question	Quantifies ischemic burden and viable myocardium together; PET / CMR also give absolute flow / scar.

Hibernation, stunning & scar — the viability concept

will function recover after revasc?

Hibernating chronic hypoperfusion → dysfunctional but viable; recovers with revascularization.
Stunned transient post-ischemic dysfunction with restored flow — recovers spontaneously over days–weeks.

Scar transmural infarction — non-viable, will not recover; revascularizing it adds risk, not function.
Mixed most ICM is patchy. Viability informs recovery expectations but is not required to establish CABG benefit.

Viability testing — how each modality calls it

no single perfect test

Modality	Marker of viability
Dobutamine stress echo	Contractile reserve — a biphasic response (low-dose improvement then high-dose worsening) is the most specific sign of hibernation.
FDG-PET (perfusion–metabolism)	Flow–metabolism mismatch (↓ perfusion, preserved FDG uptake) = hibernating, viable. Highest sensitivity.
SPECT (rest / redistribution)	Preserved tracer uptake (> 50–60%) indicates viable myocardium.
Cardiac MRI (LGE)	Transmural scar extent — < 50% wall thickness predicts functional recovery; > 75% rarely recovers.

The STICH viability lesson: in the STICH substudy, the presence of viable myocardium predicted better overall survival but did not identify the patients who derived a survival benefit from CABG over medical therapy (no significant interaction). Viability testing alone should not gate the revascularization decision — weigh symptoms, ischemic burden, anatomy & surgical risk together.

Putting it together — act on the work-up

anatomy + ischemia + viability + symptoms

Favors revascularization

Angina or a large ischemic burden on stress imaging.
Substantial viable jeopardized myocardium + good distal targets.
Low EF + suitable anatomy → CABG preferred (STICHES, P.11).

Favors GDMT alone

No angina / no ischemia, predominantly transmural scar.
PCI for the EF alone — no survival/EF benefit (REVIVED, P.11).
Prohibitive surgical risk without a symptom target.

Recognize the ischemic-CM variants & mimics

not every low EF with CAD is chronic ICM

Pattern	Clue & implication
Chronic ICM	Multivessel CAD, prior MI / scar, regional wall-motion abnormality matching coronary territories.
Hibernation-predominant	Dysfunction out of proportion to scar — the territory recovers after revascularization.
Stunning	Acute, transient post-ischemic dysfunction (post-PCI, post-arrest) — recovers spontaneously.
Takotsubo (mimic)	Transient wall-motion pattern, often beyond one coronary territory. CAD may coexist. Integrate angiography and CMR.
Mixed etiology	CAD plus a second hit (toxin, tachycardia, valve) — treat both, or the EF won't recover.

Prognosis: ischemic CM has worse survival than non-ischemic CM at the same EF — be aggressive with GDMT & secondary prevention (P.12), and don't delay ICD evaluation once EF ≤ 35% persists after optimization (→ Part F).

The question is not "is there a stenosis" but "will revascularizing it improve survival or symptoms?" Three trials frame the answer in LV dysfunction & stable CAD — and they point in different directions for surgery vs PCI.

The defining trials

CABG ≠ PCI here

Trial	Population	Result
STICH / STICHES	EF ≤ 35%, CAD amenable to CABG	CABG + OMT vs OMT. All-cause mortality was not significant at 5 yr (STICH) but became significant at 10 yr — 58.9% vs 66.1%, HR 0.84 (STICHES). CV death & death/CV-hospitalization also ↓.
REVIVED-BCIS2	EF ≤ 35%, viable, extensive CAD	PCI + OMT vs OMT. No reduction in death or HF hospitalization (37.2% vs 38.0%, HR 0.99); no LVEF improvement; QoL converged by 24 mo.
ISCHEMIA	Stable CAD, mod–severe ischemia (mostly preserved EF)	Routine invasive vs conservative. No reduction in CV events; better angina-related QoL in those with baseline angina. Excluded EF < 35% & advanced HF.

What the 2021 revascularization guideline says

ICM with low EF

COR	Recommendation
1	CABG is recommended to improve survival in SIHD with multivessel CAD suitable for CABG and LVEF <35%.
2a	CABG is reasonable for survival in selected multivessel CAD with LVEF 35–50%. Diabetes, anatomy and surgical risk inform the Heart-Team choice.
Trial	REVIVED did not show survival/HF-admission benefit from PCI for stable severe ischemic LV dysfunction. This is not a Class 3 recommendation against CABG or revascularization for ACS/angina.

CABG vs PCI — the Heart-Team call

anatomy + risk + comorbidity

Favors CABG

Multivessel / left-main, high SYNTAX complexity.
Diabetes, low EF, viable jeopardized territory.
Durable, survival-supported in ICM (STICHES).

Favors PCI

Focal disease, suitable single/two-vessel targets.
High surgical risk, frailty, need for faster recovery.
But: no survival/EF benefit in low-EF ICM (REVIVED).

Read the trials carefully — their boundaries

don't over-extrapolate

Trial	What it does not establish
STICH / STICHES	Excluded significant left-main & severe angina; benefit needed long follow-up at experienced surgical centers — not a license to delay in symptomatic disease.
REVIVED-BCIS2	Tested PCI, not CABG — it does not refute surgical benefit; required viability, so it's silent on no-viability and acute settings.
ISCHEMIA	Excluded EF < 35%, recent ACS, left-main, unacceptable angina — do not apply its "conservative is fine" message to HFrEF.

Peri-procedural & timing in HF

the medical therapy carries the benefit

Continue GDMT through & after revascularization — it carries the survival benefit, not the stent.
Completeness of revascularization matters; treat the ischemic / viable territories.

ICD timing wait ≥ 90 d post-revasc & reassess EF before a primary-prevention implant (→ Part F).
High surgical risk → PCI of the culprit/ischemic territory or a hybrid approach; Heart Team decides.

Bottom line: in low-EF ICM, surgical (CABG) revascularization is the evidence-based survival play in selected patients with suitable anatomy & acceptable risk; PCI does not improve survival or EF when done for the LV dysfunction alone. Revascularize for angina, demonstrable ischemia, or large viable jeopardized territory — individualize through the Heart Team. → Part D feeds device timing in Part F (ICD ≥ 90 d post-revasc).

Once an ischemic etiology is established, GDMT doubles as secondary prevention — layer the four pillars onto the post-MI/CAD bundle. The biggest mortality gains in ICM come from consistent medical therapy, not the cath lab.

The ischemic-HF secondary-prevention bundle

layer all of it

Element	Target / class	Notes
Antiplatelet	1	Aspirin indefinitely; add a P2Y₁₂ inhibitor (DAPT) per ACS / stent timing.
High-intensity statin	1	Atorvastatin 40–80 / rosuvastatin 20–40; add ezetimibe ± PCSK9 to reach LDL goal (≤ 55 mg/dL very-high-risk).
Four-pillar GDMT	1	ARNI/ACEi-ARB · beta-blocker · MRA · SGLT2i — disease-modifying and protective post-MI (P.5).
ICD (primary prevention)	1	If EF ≤ 35% & NYHA II–III after ≥ 3 mo GDMT, ≥ 40 d post-MI / ≥ 90 d post-revasc (→ Part F).
Risk factors / rehab	1	BP, glycemia, smoking cessation, weight; cardiac rehab improves outcomes.

Angina in HFrEF — what's safe

avoid negative inotropes

Beta-blocker already a pillar — uptitrate for both HF and anti-anginal effect.
Nitrates (incl. H-ISDN) & ranolazine for residual angina.

Avoid non-dihydropyridine CCBs (diltiazem, verapamil) in HFrEF — negative inotropy worsens HF.
Persistent angina on OMT → reconsider revascularization (P.11).

When to escalate or refer

don't anchor on "just CAD"

Trigger	Action
Refractory angina despite OMT	Revascularization evaluation / Heart Team (P.11); consider refractory-angina therapies.
Recurrent ICD shocks / VT	EP referral — antiarrhythmics, ablation (→ Part F).
Stage D despite full GDMT	Advanced-HF referral for durable or temporary support and transplant assessment; inotropes are a bridge, not a destination.

Risk-factor targets

put numbers on the bundle

Target	Goal
LDL-C	≤ 55 mg/dL (very-high-risk ASCVD); high-intensity statin → add ezetimibe → PCSK9 inhibitor as needed.
Blood pressure	< 130/80 mmHg — much of the GDMT (ARNI/ACEi-ARB, MRA, BB) does double duty.
Glycemia	HbA1c individualized (~7%); prefer SGLT2i / GLP-1 RA — agents with proven CV benefit.
Tobacco	Complete cessation — the single highest-yield secondary-prevention change.
Activity	Cardiac rehab; build to ≥ 150 min/wk moderate activity as tolerated.

Longitudinal care — don't set & forget

the EF and the risk change

Reassess EF & symptoms after ≥ 3 mo GDMT — re-decide ICD candidacy (→ Part F).
Labs K⁺ / Cr with each RAASi/MRA change; lipids 4–12 wk after starting/uptitrating a statin.

Vaccinate influenza · pneumococcal · COVID — each lowers HF events.
Adherence & cost the limiting factor — use the access workflow on P.28; screen for depression.

LINK

This Part feeds the one-page medication, adverse-event, cost, and access view on P.28 and device detail in Part F (EP). Inpatient decongestion and cardiogenic-shock care are in Part G; advanced therapies require specialist assessment.

Valvular disease is both a cause and a consequence of HF. The pivotal split in MR is primary (degenerative — the valve is the problem) vs secondary / functional (the ventricle is the problem: annular dilation & leaflet tethering from LV remodeling). In HFrEF, secondary MR is common — it both reflects and worsens HF. GDMT and CRT come first — they can shrink functional MR by a grade.

Primary vs secondary MR — the fix differs

mechanism drives therapy

Type	Mechanism	The fix
Primary (degenerative)	Prolapse, flail, ruptured chordae, endocarditis — intrinsic leaflet / apparatus disease.	Mitral surgery (repair > replace); TEER if prohibitive risk.
Secondary (functional)	Annular dilation + leaflet tethering from LV / LA remodeling; valve is structurally normal.	GDMT + CRT first; TEER in the COAPT phenotype.
Ischemic secondary	Regional remodeling / papillary-muscle displacement post-MI.	Revascularize + GDMT; TEER selected (→ Part D).

TEER for secondary MR — COAPT vs MITRA-FR

opposite results, reconciled by selection

Trial	Result	Why
COAPT	TEER + GDMT ↓ HF hospitalization (HR 0.53) & all-cause mortality (HR 0.62) at 2 y.	Selected anatomy and severe MR after tightly supervised GDMT. Proportionality is an explanatory hypothesis, not a validated eligibility rule.
MITRA-FR	No reduction in death / HF hospitalization.	Different LV/MR burden, selection, treatment and procedural features. No single factor proves why the trials diverged.

Selection is multidimensional. RESHAPE-HF2 also reduced recurrent HF events in selected moderate-to-severe functional MR. Use symptoms, optimized therapy, anatomy and the current valve pathway, not a proportionality cutoff.

The COAPT phenotype — who to refer

match the trial that worked

Severe secondary MR (EROA ≥ 30 mm², RegVol ≥ 45 mL, or integrative severe).
Symptomatic NYHA II–IV despite maximally-tolerated GDMT + CRT if indicated.

LVEF 20–50%, LVESD ≤70 mm, PASP ≤70 mmHg and suitable anatomy under the 2020 US guideline pathway.
COR 2a TEER reasonable in the COAPT phenotype; life expectancy > 1 y.

Grading severe MR — the thresholds differ

secondary is "severe" at lower numbers

Parameter	Severe — primary	Severe — secondary
Vena contracta width	≥ 0.7 cm	≥ 0.7 cm
EROA	≥ 0.40 cm²	≥ 0.30 cm²
Regurgitant volume	≥ 60 mL	≥ 45 mL
Regurgitant fraction	≥ 50%	≥ 50%

Integrate qualitative, semi-quantitative & quantitative signs — no single number is definitive, and load (BP, sedation) changes secondary MR. Quantify after GDMT optimization.

Primary MR — when to operate & how

repair at a valve center

Symptomatic severe primary MR → surgery 1.
Asymptomatic + LVEF ≤ 60% or LVESD ≥ 40 mm → surgery 1.
Consider with new AF or PASP > 50 mmHg if durable repair is likely.

Repair > replace for degenerative MR — better LV & survival; do it at a high-volume center.
TEER can be reasonable for severe symptomatic primary MR with high or prohibitive surgical risk and suitable anatomy.

Don't refer functional MR before GDMT + CRT are maximized — and re-image after. Optimization alone often drops secondary MR a grade. Operating on the leak without addressing the ventricle treats the symptom, not the disease.

Severe AS + LV dysfunction: AVR is the only effective treatment — no medical therapy relieves fixed obstruction. A low EF is often afterload mismatch and recovers after AVR, so do not deny AVR for a low EF. Don't withhold GDMT, but the valve is the fix.

Defining & staging severe AS

symptoms decide timing

Stage	Definition
Severe (D1, high-gradient)	V_max ≥ 4 m/s, mean gradient ≥ 40 mmHg, AVA ≤ 1.0 cm².
D2 — LF/LG, reduced EF	AVA ≤ 1.0 but mean gradient < 40 & EF < 50% — classical low-flow low-gradient.
D3 — paradoxical LF/LG	AVA ≤ 1.0, gradient < 40, EF ≥ 50% but low flow (SV index < 35 mL/m²).

Symptoms (exertional dyspnea, angina, syncope) define when to act — and are often subtle. Once symptomatic, severe AS has a dismal untreated prognosis.

Low-flow low-gradient AS — the HF trap

is it really severe?

Reduced-EF LFLG (D2)

Dobutamine stress echo true-severe = AVA stays ≤ 1.0 & gradient rises ≥ 40; pseudo-severe = AVA opens > 1.0 with flow.
Flow reserve (↑ SV ≥ 20%) is prognostic — low reserve = higher risk but may still benefit.

Paradoxical preserved-EF (D3)

CT calcium score a high aortic-valve calcium score confirms true-severe when echo is equivocal.
Exclude error LVOT measurement, hypertension, AF — re-measure before committing.

TAVR vs SAVR — the evidence & indications

Heart-Team decision

Setting	Trial	Result
Intermediate risk	PARTNER 2	TAVR non-inferior to SAVR.
Low risk (balloon-exp.)	PARTNER 3	TAVR superior at 1 y (death/stroke/rehosp).
Low risk (self-exp.)	Evolut Low Risk	TAVR non-inferior for death / disabling stroke.

COR 1: AVR for symptomatic severe AS, and for asymptomatic severe AS with EF < 50%. Favor SAVR with bicuspid anatomy, concomitant CAD/other-valve surgery, or unfavorable TAVR access; choice is age/anatomy/lifetime-management driven.

Asymptomatic severe AS — when not to wait

EF and triggers force the hand

EF < 50% with severe AS → AVR 1, even if asymptomatic.
Symptoms on exercise testing (a safe, useful unmasker) → AVR.

Consider early AVR 2a if very severe (V_max ≥ 5 m/s), rapid progression, or markedly ↑ BNP.
Otherwise watchful waiting with serial echo — onset of symptoms is the trigger.

After TAVR — what to watch

conduction · leak · lifetime plan

Issue	Detail
Conduction disease	New LBBB / high-grade AV block → pacemaker in a minority (more with self-expanding valves); rhythm monitoring peri-procedure.
Paravalvular leak	Less common with current valves; more than mild PVL worsens outcomes — assess on echo.
Lifetime management	Valve durability & future valve-in-valve planning (esp. younger patients); antithrombotic = SAPT unless another OAC indication (e.g., AF).

Decompensated severe AS: be cautious with vasodilators & aggressive diuresis — a fixed obstruction is preload-dependent and hypotension can be catastrophic. Stabilize and expedite AVR. → shock care in Part G.

Beyond MR & AS, three structural problems shape HF care — tricuspid regurgitation, left-atrial-appendage closure for AF, and interatrial shunts. Each is a Heart-Team conversation layered on top of optimized GDMT.

Tricuspid regurgitation — the "forgotten valve"

drives congestion & cardiorenal disease

Secondary TR annular dilation from RV / PH / AF is the common form in HF; severe TR → RV volume overload, congestion, cardiorenal & hepatic dysfunction.
Under-treated isolated TR surgery historically carried high mortality.

TRILUMINATE T-TEER improved TR and QoL at 1 year. In the full 572-patient cohort, recurrent HF admissions fell by 2 years, without demonstrated survival benefit.
Select symptomatic severe TR after diuretic optimization; treat the driver (PH, left-heart disease, AF).

LAA closure (Watchman) — AF stroke prevention

when OAC isn't an option

Point	Detail
Why it matters in HF	AF accompanies ~⅓ of HF; many have high bleeding risk that limits long-term anticoagulation.
Evidence	PROTECT-AF & PREVAIL: percutaneous LAA closure non-inferior to warfarin for stroke prevention in nonvalvular AF.
Use	AF with a contraindication / intolerance to long-term OAC; weigh CHA₂DS₂-VASc vs HAS-BLED; needs short post-implant antithrombotic + TEE/CT follow-up.

Shunts — ASD / PFO & the atrial-shunt concept

close the harmful, beware the protective

Lesion	In the HF context
Significant ASD	Chronic L→R shunt (RV enlargement, Qp:Qs ≥ 1.5) → RV overload & HF — close; but not with severe PH / Eisenmenger physiology (R→L).
HFpEF atrial shunt	Devices that create an L→R shunt to offload the LA remain investigational — neutral overall in a large randomized trial; not standard care.
PFO	Relevant to stroke (paradoxical embolism), not a primary HF therapy.

TR severity & the RV consequence

grade it, then treat the driver

Severe TR — the call

Vena contracta ≥ 0.7 cm; dense triangular CW jet; systolic hepatic-vein flow reversal.
Dilated annulus (≥ 40 mm), RA / RV / IVC enlargement (newer "massive / torrential" grades = extreme).

Why it matters

→ RV volume overload, venous congestion, cardiorenal & cardiohepatic disease, ascites.
Treat the driver first — left-heart disease, PH, AF, RV failure (→ Pulmonary Hypertension card).

Structural referral quick-map

who goes to the Heart Team

Lesion	Refer when
Secondary MR	Severe + symptomatic on maximal GDMT / CRT — the COAPT phenotype (P.13).
Aortic stenosis	Severe + symptoms, or any severe AS with EF < 50% (P.14).
Tricuspid regurgitation	Severe + symptomatic after diuretic optimization; address the driver.
AF + high bleeding risk	Long-term OAC contraindicated → LAA closure evaluation.
Significant ASD	RV enlargement / Qp:Qs ≥ 1.5 without severe PH / Eisenmenger.

LINK

Structural disease overlaps every Part — fix the valve and optimize the ventricle: GDMT (Part B), revascularize ischemia (Part D), manage rhythm & devices (Part F), decongest (Part G). End-stage disease despite all of it requires advanced-HF referral.

Sudden cardiac death is a leading mode of death in HFrEF. An ICD prevents arrhythmic death — it does nothing for pump failure or symptoms. Optimize GDMT first (it lowers SCD and often improves EF, which reclassifies candidacy), then decide on the device.

Primary-prevention ICD — who qualifies

EF after ≥ 3 mo GDMT

COR	Indication (expected meaningful survival > 1 y)
1	EF ≤ 35%, NYHA II–III on ≥ 3 mo GDMT — ischemic or nonischemic.
1	EF ≤ 30%, NYHA I, ischemic (MADIT-II population).
timing	Wait ≥ 40 d post-MI & ≥ 90 d post-revascularization; reassess EF after GDMT first.

The evidence

ischemic benefit clearest

Trial	Result
MADIT-II	Post-MI, EF ≤ 30% → ICD ↓ mortality vs medical therapy.
SCD-HeFT	EF ≤ 35%, NYHA II–III (ischemic & nonischemic) → ICD ↓ mortality ~23%; amiodarone no benefit.
DANISH	Nonischemic CM → no significant all-cause mortality reduction overall, but halved SCD; benefit attenuated by age, CRT & modern GDMT.

Nonischemic primary-prevention ICD remains reasonable (esp. younger patients), but the absolute benefit is smaller in the CRT + four-pillar era — individualize.

Secondary prevention & device choice

survivors & the bridge

Secondary prevention 1 survived VF / unstable VT not from a reversible cause → ICD.
S-ICD no pacing/CRT capability — good for young, infection risk, difficult access, no pacing need.

WCD (wearable) a bridge during the waiting period; VEST missed its ITT endpoint (adherence) — selective use.
Don't implant early post-MI — DINAMIT showed no benefit at < 40 d.

Before you implant — device choice & counseling

shared decision

CRT-D vs ICD-alone

Coexisting CRT indication (LBBB, QRS ≥ 150) → CRT-D (P.17).
SCD cover only, no pacing need → transvenous ICD or S-ICD.

Counsel on

Inappropriate shocks, lead / pocket infection, generator changes.
End-of-life deactivation — discuss early & revisit.

When an ICD is not indicated

avoid futile implants

Situation	Why not
Reversible cause	Ischemia, electrolytes, drug, peripartum, tachy-CM — fix it; arrhythmic risk may resolve.
< 40 d post-MI / < 90 d post-revasc	No early survival benefit (DINAMIT); bridge with a WCD if high-risk.
NYHA IV, not a candidate	Not for CRT / transplant / LVAD → ICD unlikely to help; align with goals of care.
Life expectancy < 1 y	Competing non-arrhythmic mortality outweighs the benefit.

At each generator change, re-examine the indication — if the EF has normalized and the device has never fired, replacement is a fresh shared decision, not automatic.

An ICD is pure SCD insurance — it does not improve symptoms or pump function. Always maximize GDMT (± CRT) first & re-measure the EF; many patients improve out of the indication, and the SCD risk itself falls on therapy.

A wide LBBB makes the septum and lateral wall contract out of sync, wasting stroke work. CRT (biventricular pacing) resynchronizes them → reverse remodeling, fewer HF hospitalizations and lower mortality — the most EF-improving device therapy in HFrEF. Benefit tracks with QRS width & LBBB morphology.

CRT indications — QRS & morphology drive the class

on GDMT, EF ≤ 35%, sinus rhythm

COR	Indication (NYHA II–IV ambulatory, EF ≤ 35%, GDMT)
1	LBBB + QRS ≥ 150 ms — the strongest responders.
2a	LBBB + QRS 120–149 ms; or non-LBBB + QRS ≥ 150 ms.
2a	Expected to require > 40% ventricular pacing, or AF with a plan for near-100% BiV capture.
3	QRS < 120 ms — no benefit / harm (EchoCRT); do not implant for dyssynchrony imaging alone.

The evidence

mortality + reverse remodeling

Trial	Result
COMPANION	NYHA III–IV → CRT-P/CRT-D ↓ death or hospitalization; CRT-D also ↓ mortality.
CARE-HF	CRT (no defibrillator) ↓ all-cause mortality — resynchronization itself saves lives.
MADIT-CRT · RAFT	Mild HF (NYHA I–II) → CRT-D ↓ HF events; benefit concentrated in LBBB / QRS ≥ 150.

Maximize response & conduction-system pacing

capture % is everything

Best responders / "non-responder" check

LBBB + wide QRS + female + nonischemic respond best.
Aim > 98% BiV pacing; a non-responder → check pacing %, lead position, AF burden, scar.

Conduction-system pacing (His / LBBAP)

Physiologic; an alternative/adjunct when the CRT lead fails or for pacing-induced CM / high pacing burden.
Randomized outcome evidence still maturing — promising for non-responders.

Pacing-induced CM & the upgrade question

high RV pacing hurts

Scenario	Action
High RV pacing burden	Chronic RV pacing (> 20–40%) can cause LV dysfunction — anticipate in AV block with low-normal EF.
Pacing-dependent + EF falls	Upgrade to CRT or convert to conduction-system pacing (His / LBBAP).
New device, expected high pacing	Choose CRT or CSP up front rather than RV-only pacing.

Device choice & follow-up

capture % drives benefit

CRT-P vs CRT-D add the defibrillator when an ICD indication coexists; CRT-P reasonable in older / comorbid patients (P.16).
Target > 98% biventricular pacing — confirm at interrogation.

Non-responder optimize AV/VV timing, reposition / quadripolar lead, treat AF, address scar.
Reassess EF at 3–6 mo — reverse remodeling guides ICD & medication decisions.

CRT works fastest in nonischemic LBBB — expect reverse remodeling over 3–6 months; extensive lateral-wall ischemic scar blunts the response.

Uncontrolled AF defeats CRT — intrinsic conduction steals capture below the threshold for benefit. Rate-control aggressively, and consider AV-node ablation to guarantee near-100% biventricular pacing (P.18).

AF and HF cause and worsen each other — and both can be reversible drivers of cardiomyopathy (tachycardia-mediated). When a "new cardiomyopathy" coincides with a fast, sustained arrhythmia or a high PVC burden, treat the rhythm — the EF may fully recover (P.4).

AF in HF — rate, rhythm & anticoagulation

restore sinus when you can

Lever	In HFrEF
Anticoagulate	Per CHA₂DS₂-VASc (HF already scores a point) — DOAC preferred over warfarin in most.
Rate control	Beta-blocker first-line; digoxin adjunct. Avoid non-DHP CCBs (diltiazem/verapamil) & dronedarone in HFrEF.
Rhythm control	Increasingly favored; safe AADs in HFrEF = amiodarone, dofetilide. Avoid class IC (flecainide/propafenone) with structural disease.

Ablation & "ablate-and-pace"

sinus rhythm improves the EF

CASTLE-AF catheter ablation in HFrEF + AF ↓ death or HF hospitalization (HR 0.62) vs medical therapy & improved EF.
Select symptomatic HFrEF + AF, esp. if tachycardia-mediated CM is suspected.

AVN ablation + CRT ("ablate & pace") for refractory rapid AF when rhythm/rate control fails — guarantees ventricular-pacing capture.
Caveat creates pacing dependence — pair with CRT / conduction-system pacing (P.17).

Ventricular arrhythmias & PVC-induced CM

a curable cardiomyopathy

Problem	Approach
Scar-related VT	ICD + beta-blocker ± amiodarone; VANISH: catheter ablation beat escalated AADs for VT/shock recurrence.
VT / electrical storm	Treat ischemia & electrolytes, beta-blockade, sedation; antiarrhythmics, ablation, and temporary support if unstable.
PVC-induced CM	High burden (often > 15–20%) → suppress by ablation or meds; the EF frequently normalizes.

Antiarrhythmic safety in HFrEF

most are unsafe — know the few that aren't

Drug	In HFrEF	Note
Amiodarone	Safe	Rhythm/rate workhorse; watch thyroid / lung / liver / skin toxicity & interactions.
Dofetilide	Safe	In-hospital initiation; QT & renal dosing.
Sotalol	Caution	Negative inotropy + QT risk — avoid in significant HFrEF.
Dronedarone	Avoid	↑ mortality in HFrEF & permanent AF (ANDROMEDA / PALLAS).
Flecainide / propafenone	Avoid	Class IC — proarrhythmic with structural heart disease (CAST).

Tachycardia-induced CM — a reversible cause

treat the rhythm, recover the EF

Suspect when

Low EF but a not-very-dilated LV; persistent tachycardia or high PVC burden.
EF improves within weeks of rate / rhythm control.

Act

Rate / rhythm control, ablation; screen for it in every "new" CM (P.4).
Recovery can be complete — but relapses if the arrhythmia returns.

LINK

Arrhythmia is woven through HF — rhythm can cause the cardiomyopathy (reversible, P.4), defeat CRT (P.17), or end the patient (ICD, P.16). Recurrent shocks / VT storm or refractory AF → EP referral; hemodynamic collapse → urgent shock-team and advanced-HF assessment.

Acute decompensated HF is mostly a congestion problem — the job is decongestion to euvolemia without wrecking the kidneys. Profile every patient by congestion (wet/dry) and perfusion (warm/cold); the quadrant dictates therapy. Most ADHF is warm-and-wet.

Creatinine rises: reassess the patient before changing the diuretic

Original qualitative decision schematic. Weight or creatinine alone cannot distinguish successful decongestion from harmful volume removal. Consider residual congestion, BP, urine output, electrolytes and perfusion.

Motion follows the reasoning, not a measured physiologic rate.

The hemodynamic quadrants (Stevenson)

congestion × perfusion

Profile	Picture	Lead therapy
Warm & wet	Congested, perfusing (commonest)	IV loop diuretics ± vasodilator if hypertensive.
Cold & wet	Congested + hypoperfused	Diurese + inotrope if low output / hypotension; tMCS if shock (P.20).
Cold & dry	Hypoperfused, not congested	Cautious fluids; inotrope; hunt the cause.
Warm & dry	Compensated (the goal)	Optimize oral GDMT, plan discharge.

IV loop diuretics — dose it right (DOSE)

enough, early

Dose DOSE compared total daily IV furosemide at 1× versus 2.5× home oral furosemide-equivalent dose. Individualize the initial regimen and escalate by early response.
High-dose achieved more decongestion with only a transient, reversible ↑ Cr.

Check early a spot urine Na⁺ at 2 h (< 50–70 mmol/L = inadequate natriuresis → escalate).
Double the dose if response is poor rather than waiting — don't under-treat.

Diuretic resistance — add a second site

sequential nephron blockade

Add-on	Evidence & use
Acetazolamide	ADVOR: IV acetazolamide added to loop → more & faster decongestion (proximal tubule).
Thiazide (metolazone / HCTZ / chlorothiazide)	CLOROTIC: adding HCTZ → greater weight/fluid loss; watch K⁺, Na⁺, renal.
MRA (natriuretic dose) · SGLT2i	ATHENA-HF found no incremental acute decongestion benefit from spironolactone 100 mg/day. Use standard MRA/SGLT2 therapy for appropriate long-term indications.
Ultrafiltration	CARRESS-HF: UF not superior to stepped pharmacologic care in cardiorenal — reserve for true diuretic failure.

Track decongestion & the cardiorenal axis

euvolemia, not a creatinine

Monitor daily weight, strict I/O, JVP/edema exam, lytes & Cr; spot urine Na⁺ to gauge response.
A modest ↑ Cr during effective decongestion is not an automatic stop signal. Check residual congestion, BP, urine output and perfusion.

PA catheter not routine — ESCAPE showed no benefit; reserve for unclear hemodynamics / shock.
Don't stop GDMT reflexively — hold/adjust only for true intolerance.

Adjuncts & cautions

decongestion is the lever; the rest is selective

Lever	Use
IV vasodilator (NTG)	Hypertensive ADHF / flash pulmonary edema — rapid pre/afterload reduction; avoid if hypotensive or severe AS.
Inotrope	Only for low-output / cold profiles — not for warm-and-wet (↑ MVO₂, arrhythmia).
Tolvaptan	Symptomatic hyponatremia — aquaresis for the sodium, not an outcome therapy.
Avoid	NSAIDs, excess IV fluids, and reflexively stopping GDMT for a small Cr bump. Sodium / fluid restriction → P.45; oral sodium / hypertonic saline → P.46.

Under-diuresis > over-diuresis as the failure mode. Residual congestion at discharge is the strongest predictor of readmission — diurese to true euvolemia (flat JVP, no orthostasis), confirm a stable oral regimen for ≥ 24 h, then transition. → start/continue GDMT in P.21.

Cardiogenic shock = end-organ hypoperfusion from cardiac failure despite adequate (often elevated) filling pressure. Mortality is high & time-sensitive — recognize early, find the cause, and match support to trajectory. A shock team coordinates timely diagnosis and escalation. Outcome associations come mainly from observational programs.

SCAI SHOCK stages — a common language

stage drives escalation

Stage	Label	Picture
A	At risk	Large MI / decompensated HF, not yet hypoperfused.
B	Beginning	Hypotension or tachycardia, no hypoperfusion yet.
C	Classic	Hypoperfusion needing intervention (inotrope/pressor/MCS); ↑ lactate.
D	Deteriorating	Failing to respond — escalating support.
E	Extremis	Circulatory collapse / refractory arrest, often on CPR-ECMO.

Re-stage frequently; a modifier for cardiac arrest applies. Stage tracks mortality & guides urgency.

Work it up fast

confirm · classify · find the fix

Confirm hypoperfusion — lactate, urine output, mentation, cool extremities, narrow pulse pressure.
Classify LV vs RV vs mechanical vs mixed/distributive; echo immediately.

PA catheter useful here (unlike routine ADHF) — defines CI, filling pressures, guides support.
Find the fix acute MI → revascularize; mechanical (papillary rupture, VSD, tamponade, RV infarct) → targeted therapy.

Pharmacologic support

perfusion, not a number

Agent	Role
Norepinephrine	First-line vasopressor in CS — fewer arrhythmias / better survival than dopamine.
Dobutamine · milrinone	Inotropes for low output; milrinone spares the β-receptor (chronic BB, but renally cleared & vasodilatory).
Caution	Every inotrope/pressor raises MVO₂ & arrhythmia risk — use the lowest dose that restores perfusion; they bridge, not treat.

Temporary mechanical support

escalate by trajectory

IABP modest unloading — no mortality benefit in MI-shock (IABP-SHOCK II); not routine.
Microaxial pump DanGer Shock: improved survival in selected STEMI CS, at a cost of more complications.

VA-ECMO full cardiopulmonary support for selected biventricular or refractory shock; use current device guidance and specialist oversight.
Define the exit bridge to recovery, decision, durable LVAD or transplant — escalate before multi-organ failure.

TRAP

Read perfusion and trajectory, not blood pressure alone. Normal BP can coexist with cold, oliguric hypoperfusion; low BP may be predominantly vasodilatory. Temporary MCS is not a potency ladder: IABP, microaxial support, VA-ECMO, and combined strategies change preload, afterload, oxygenation, and ventricular loading differently. Match the device to LV/RV/biventricular phenotype, oxygenation need, trajectory, and exit strategy.

RV-predominant shock — a different playbook

don't flood the RV

Recognize

Causes: RV infarct, massive PE, post-cardiotomy, PH, post-LVAD RV failure.
High RA, elevated RA:PCWP, low PA pulsatility (PAPi), clear lungs with shock.

Manage

Optimize (don't overload) preload; keep coronary perfusion with norepinephrine.
Cut RV afterload — avoid hypoxia/hypercarbia/acidosis; inhaled pulmonary vasodilator; RV-MCS / VA-ECMO if refractory.

Track lactate clearance & end-organ markers (urine, mentation) — they reflect resuscitation better than blood pressure alone.

Don't drift. Shock that isn't improving on escalating pressors/inotropes within hours needs a shock-team decision on MCS — late cannulation after organ failure rarely rescues. Match the device to LV vs RV vs biventricular failure and verify current device guidance.

The admission is the opportunity. Start & up-titrate GDMT before discharge — once stable and decongesting — then follow closely. STRONG-HF reduced the composite of death or HF readmission with rapid titration and close follow-up. Discharge doses must remain tolerated and safe. Each hospitalization raises the stakes: in DAPA-HF, 2-yr death/HFH was 21% (never hospitalized) → 25% (>1 yr prior) → 34% (≤1 yr prior); SGLT2i benefit grew with recency (RRR 16→27→36%).

STRONG-HF — the high-intensity model

the evidence for doing this

STRONG-HF targeted RAS inhibitor, beta-blocker and MRA doses within 2 weeks, with visits at weeks 1, 2, 3 and 6. Death or HF readmission at 180 days was 15.2% versus 23.3%. SGLT2 therapy was not the randomized titration intervention.

Inpatient initiation — start all four, low & early

when stable, off inotropes

Pillar	In-hospital start
SGLT2 inhibitor	Start after stabilization within the drug-specific renal threshold. Check volume status, intake and ketoacidosis risk. No HF-dose titration. Often the first pillar in.
ARNI / ACEi-ARB	Start when BP allows; 36-h washout from an ACEi before sacubitril/valsartan (angioedema).
Beta-blocker	Start/continue once euvolemic & off inotropes — do not start in active congestion or low-output.
MRA	Start if K⁺ < 5.0 & eGFR > 30; recheck K⁺/Cr in days.

Safety while you sequence

BP · K⁺ · renal

Permit a small Cr rise (≤ ~30%) & relative hypotension if asymptomatic — don't abandon GDMT.
K⁺ manage with diet/binders rather than dropping the MRA/RAASi when feasible.

SGLT2i hold for major surgery, prolonged fasting or acute illness with poor intake. Restart only when stable and eating, without ketoacidosis risk.
½-dose both RAASi & BB beats full-dose-one-omit-another — get all four on board.

The discharge bundle & RV-failure note

close the loop

Before discharge	Confirm
Euvolemia	Near-dry weight, stable oral diuretic ≥ 24 h, flat JVP.
GDMT + plan	All four pillars started or documented why not; written up-titration plan (P.5).
Follow-up & labs	Visit within 7 days; labs scheduled; cardiac rehab referral; education + daily weights.
Access & cost	Affordability addressed — → P.28, discharge GDMT at a glance.
RV failure	Optimize preload, avoid hypoxia/acidosis (↑ PVR), inotropic RV support; refractory → MCS/RVAD (→ ECMO/LVAD cards).

Common readmission traps

address modifiable risks

Trap	Fix
Residual congestion	Discharge at true euvolemia, not merely "better" — the #1 driver of bounce-back.
GDMT not up-titrated	Use the admission; follow the STRONG-HF schedule (P.5).
No early follow-up	7-day visit + labs; a nurse phone check in between.
Cost / non-adherence	Address access & simplify the regimen (→ P.28).
Missed reversible cause	Re-confirm etiology (ischemia, AF, valve, tachy-CM) before discharge (P.4).

LINK

This closes the loop the guide opened: confirm the phenotype (P.2), start phenotype-based therapy (P.5–8), decongest (P.19), rescue shock (P.20), and leave on optimized, up-titrated GDMT with 7-day follow-up (STRONG-HF). Recurrent admissions despite all of it → advanced-HF referral for assessment of advanced therapies.

A chairside reference for initiating, titrating & troubleshooting each pillar — pairs with the sequencing logic on P.5 and the at-a-glance cost / trial-AE view on P.28. Principle throughout: low doses of all four beat the target dose of one.

RAS inhibition — ARNI ▸ ACEi ▸ ARB (use one only)

ARNI preferred

Agent	Start → target	Cautions & contraindications
Sacubitril/valsartan (ARNI)	24/26 or 49/51 BID → 97/103 BID	36-h washout from an ACEi (angioedema); assess symptomatic hypotension/perfusion rather than an automatic SBP hold; raises K⁺ less than ACEi; follow NT-proBNP (BNP rises on ARNI).
ACEi (enalapril, lisinopril)	low → target	Cough, angioedema, hyperK, ↑ Cr; CI pregnancy, bilateral renal-artery stenosis, prior angioedema.
ARB (candesartan, valsartan)	low → target	Use for ACEi cough; same renal/K cautions; the fallback when ARNI isn't feasible.

Never combine two RAS agents (ACEi + ARB, or either with ARNI) — only one at a time.

Beta-blockers — only three are evidence-based

start when euvolemic

Agent	Start → target	Notes
Carvedilol	3.125 BID → 25 BID (50 if > 85 kg)	α+β block — more BP effect.
Metoprolol succinate (XL)	12.5–25 qd → 200 qd	β1-selective; not tartrate (immediate-release).
Bisoprolol	1.25 qd → 10 qd	β1-selective.

Start only when euvolemic & off inotropes; titrate q ≥ 2 wk. In a mild flare reduce, don't stop. CI: high-grade AV block without a pacemaker, decompensated HF.

MRA & SGLT2i

the two easy wins

MRA

Spironolactone 12.5–25 → 25–50 qd; gynecomastia → switch to eplerenone.
Start if K⁺ < 5.0 & eGFR > 30; recheck K⁺/Cr day 3 & 1 wk; manage hyperK promptly. Discontinue MRA if K⁺ cannot be maintained <5.5.

SGLT2 inhibitor

Dapagliflozin / empagliflozin 10 mg qd, no titration; benefit independent of diabetes; initiate to low eGFR (≈ 20–25).
Hold for major surgery / poor intake (euglycemic DKA); genital mycotic infection; may trim the loop diuretic.

Intolerance & troubleshooting

adjust, don't abandon

Barrier	Move
Low SBP (< 90–100)	Stagger dose times; trim diuretic if euvolemic; use tolerated labeled doses and reassess perfusion. Do not routinely split unscored ARNI tablets. Adjust the responsible drug if symptomatic.
CKD / hyperkalemia	SGLT2i continue to low eGFR; both steroidal MRAs raise K⁺. Review interacting drugs and recheck labs. Binders may aid continuation, but improved clinical outcomes are uncertain.
Bradycardia / AV block	Reduce the BB before stopping; reassess other rate-active drugs; consider a device if pacing-dependent.
ACEi cough / angioedema	Cough → ARB or ARNI; angioedema → avoid ARNI & ACEi (ARB only, cautiously).

Drugs to avoid (or use with caution) in HFrEF

they worsen HF or outcomes

Drug / class	Why
Non-DHP CCB (diltiazem, verapamil)	Negative inotropy worsens HFrEF — avoid (amlodipine/felodipine are neutral if a CCB is needed).
NSAIDs / COX-2 inhibitors	Na⁺ & fluid retention, ↑ Cr, blunt diuretics & RAAS inhibitors.
Thiazolidinediones (pioglitazone)	Fluid retention → HF exacerbation & hospitalization.
Saxagliptin / alogliptin (DPP-4)	↑ HF hospitalization (prefer SGLT2i / GLP-1 RA for diabetes).
Class IC · dronedarone · sotalol	Proarrhythmic / ↑ mortality in HFrEF — for rhythm use amiodarone or dofetilide (P.18).

The commonest error is under-dosing — not over-dosing. Asymptomatic low BP, a < 30% Cr rise, and HR ≥ 50 are not reasons to stop GDMT. Keep all four pillars on board, even partially dosed, and re-titrate as the patient tolerates. → add-on therapies (ivabradine · H-ISDN · vericiguat · digoxin · IV iron): P.23.

Once the four pillars are titrated (P.5, P.22), layer these by indication — each has a specific niche, not a one-size add-on. Check & treat iron deficiency in every HFrEF.

Oral add-ons — match to the patient

after the pillars

Drug	Indication	Dose	Evidence & caveat
Ivabradine	Sinus rhythm, HR ≥ 70 on max-tolerated BB, EF ≤ 35%	5 → 7.5 mg BID	SHIFT: ↓ HF hosp. I_f-blocker — no effect in AF; watch bradycardia, phosphenes.
Hydralazine + ISDN	Self-identified Black, NYHA III–IV on GDMT; or ACEi/ARB-intolerant	titrate TID	A-HeFT: ↓ mortality added to GDMT; headache, TID adherence.
Vericiguat	Recent worsening (HF hosp / IV diuretic) on GDMT	2.5 → 10 mg qd	VICTORIA: ↓ CV death/HF hosp; sGC stimulator; mild hypotension.
Digoxin	Persistent symptoms; AF rate-control adjunct	level 0.5–0.9 ng/mL	DIG: ↓ hospitalization, neutral mortality; narrow window — toxicity with ↓ K⁺ / renal.

IV iron — treat the deficiency

check iron in every HFrEF

Define ferritin < 100, or 100–300 with TSAT < 20% — screen at diagnosis & at decompensation.
Use IV, not oral — oral iron is ineffective in HF (IRONOUT). FCM or ferric derisomaltose.

AFFIRM-AHF FCM at discharge after ADHF → ↓ HF hospitalizations.
IRONMAN narrowly missed its primary HF-admission/CV-death composite. Interpret alongside AFFIRM-AHF, HEART-FID and FAIR-HF2 (P.29).

Sequencing & monitoring the add-ons

pillars first, then layer

Add-on	Before adding / monitor
Ivabradine	Confirm sinus rhythm & max-tolerated BB first; monitor HR (target 50–60); stop if AF develops.
Hydralazine-ISDN	BP headroom; counsel on headache & the TID schedule (adherence is the limiter).
Vericiguat	SBP ≥ 100; avoid with other sGC stimulators / PDE5 inhibitors; check BP after titration.
Digoxin · IV iron	Digoxin — level + K⁺/renal (lower dose in CKD/elderly). IV iron — recheck iron studies, re-dose for ongoing deficiency.

Lower-yield / not routinely recommended

don't reach for these

Therapy	Verdict in HFrEF
Anticoagulation in sinus rhythm	Not routine — no net clinical benefit (bleeding offsets the stroke reduction). Anticoagulate for AF / VTE / thrombus only.
Statin started for HF alone	No mortality benefit in dedicated HF RCTs — use a statin only for a separate ASCVD indication.
n-3 PUFA (fish oil)	Only a modest effect — a reasonable optional adjunct, not core therapy.
Routine nitrates (no angina)	No benefit without angina; in HFpEF they can reduce activity.
Calcium channel blockers	Non-DHP are harmful (P.22); use amlodipine/felodipine only if needed for BP/angina.

Recommendations track the 2022 AHA/ACC/HFSA guideline — when in doubt, the proven survival therapy is the four pillars, fully dosed.

Add-ons do not substitute for the four pillars — maximize GDMT first, then layer by the specific trigger above. → devices (ICD/CRT) when EF stays ≤ 35%: Part F (P.16–17); advanced-HF referral if refractory: LVAD / Transplant cards.

Right heart catheterization (RHC) phenotypes heart failure — filling pressures, output, PVR, RV function. Routine PA-catheter use in general ADHF did not improve outcomes (ESCAPE), but RHC is invaluable in shock, unexplained dyspnea, PH work-up, pre-transplant/LVAD, and tailored therapy. The numbers only help if they are acquired correctly — standardization first.

▸

When to cath in heart failure

diagnostic & management

Diagnostic

Unexplained dyspnea / suspected HFpEF (rest ± exercise).
PH work-up & phenotype (pre- vs post-capillary).
Constriction vs restriction; suspected high-output.

Management

Cardiogenic shock / tailored therapy & escalation.
Pre-transplant / LVAD candidacy & RV assessment.
Refractory congestion; diuretic / inotrope titration.

Standardized acquisition — get the number right

end-expiration · validate the wedge

Zero the transducer at the mid-thoracic line (mid-RA); re-zero with position changes.
Read at end-expiration — not the digital mean (respiration shifts it); average only with large swings.

Validate PAWP waveform + occluded O₂ sat >90% or within 5% of systemic; half-inflate if hard to wedge; never over-wedge.
Thermodilution average reproducible injections. With severe TR, shunts or discordant results, verify technique and compare with measured-VO₂ direct Fick.

Normal resting values

know these cold

Site	Pressure (mmHg)	Flow / resistance
RA (mean)	2–6	CO 4–8 L/min
RV	15–30 / 2–8	CI 2.5–4.0 L/min/m²
PA	15–30 / 4–12 (mean 9–18)	SvO₂ 60–75%
PCWP (mean)	6–12	PVR < 2 WU · SVR 800–1200
LVEDP (if crossed)	8–12	LV gram / left-heart cath
Ao / systemic	≈ 100–140 / 60–90	MAP 70–105

Cardiac output — Fick vs thermodilution

pick the reliable one

Method	How	Caveat
Direct Fick	Measured VO₂ — gold standard	VO₂ rarely measured (equipment) → seldom used
Indirect Fick	Estimated VO₂ (Dehmer / LaFarge) — common	Inaccurate; not during exercise, obesity, age extremes
Thermo-dilution	Saline injectate, triplicate average	TR and shunts may affect accuracy. Estimated-VO₂ indirect Fick also has substantial error

⚑

Common pitfalls

avoid the artifact

Over-wedging gives an invalid tracing and risks PA injury, under/over-damping distorts the trace.
Wrong zero / non-supine shifts every value — re-zero at mid-thorax.

PADP ≠ PCWP in tachycardia, PH, or high PVR — don't substitute.
Discordant CO with severe TR or shunts needs reconciliation. Use measured VO₂ if direct Fick is available.

Measurement changes classification. Validate the wedge and document respiratory handling. Usually read at end-expiration. With large respiratory swings, average at least 3–4 respiratory cycles. Prefer thermodilution or direct Fick over assumed-VO₂ Fick when accuracy is critical.

Waveforms & derived indices turn pressures into a diagnosis & a prognosis — PH phenotype, RV reserve, and cardiac power. Advanced indices (PAPi, CPO, API) increasingly guide shock & advanced-HF decisions.

Waveforms & signatures

read the tracing

Finding	Meaning
a / v waves	a = atrial contraction (lost in AF; large with stiff RV / TS) · v = filling (giant v → MR on PCWP, TR on RA)
x / y descents	x = atrial relaxation · y = early diastolic emptying
Tamponade	Elevated & equalized diastolic pressures, blunted y, pulsus paradoxus
Constriction	Dip-and-plateau (√ sign), prominent x & y, Kussmaul, RV/LV interdependence, RAP ≈ PCWP
Restriction	LVEDP > RVEDP (≥ 5), PASP often > 50, concordant respiratory change

Derived indices — formula → use

phenotype & prognosis

Index	Formula	Use / threshold
TPG	mPAP − PCWP	< 12 normal
DPG	PADP − PCWP	No longer defines IpcPH versus CpcPH
PVR	TPG / CO (Wood units)	> 2 WU abnormal (×80 = dyn·s·cm⁻⁵)
PAPi	(PASP − PADP) / RAP	Low values raise concern. Cutoffs differ in RV-MI and LVAD cohorts
CPO	MAP × CO / 451 (W)	< 0.6 W → poor shock prognosis
API	(SBP − DBP) / PAWP	Low → adverse HF-CS outcomes
RAP/PCWP	RAP ÷ PCWP	> 0.63 → RV dysfunction
CO (Fick)	VO₂ / (Ca − Cv)O₂	indexed to BSA = CI

Putting it together — PH · transplant · RV

HF applications

PH in HF (Group 2)

PH = mPAP > 20; post-capillary PCWP > 15.
Ipc-PH (PVR ≤ 2) vs Cpc-PH (PVR >2 WU, without a DPG requirement) — Cpc worse.
HFpEF unmasking: resting PCWP ≥ 15 or supine exercise PCWP ≥25.

Transplant / LVAD / RV

Reversibility of PVR/TPG (vasodilator challenge) before transplant — fixed PVR is a red flag.
PAPi & RAP/PCWP > 0.63 flag RV-failure risk (post-LVAD / RV-MI).

▸

Hemodynamic phenotyping & power

profile → action

Congestion / perfusion

Warm-wet ↑PCWP, CI ok → decongest.
Cold-wet ↑PCWP, ↓CI → inotrope ± vasodilator.
Cold-dry ↓CI, low PCWP → fill cautiously.

Severity & RV

CPO < 0.6 W + rising lactate → decompensating shock; escalate.
Low PAPi + ↑RAP supports RV dysfunction, with context-specific thresholds. Consider RV support (→ ECMO / RVAD).

LINK

Cross-refs: the Pulmonary Hypertension guide (Group 2 detail) and current advanced-HF pathways for candidacy, RV-failure risk, temporary support, CPO, and PAPi. RHC is the connective hemodynamic backbone of the advanced-HF work-up.

The bedside exam still drives HF diagnosis & the volume/perfusion call. Anchor it on likelihood ratios (which findings actually move the needle) and the warm/cold · wet/dry profile — then correlate with NPs, echo, and (when needed) invasive hemodynamics (P.24–25).

Pressure applied to the lower leg and the persistent indentation after release, demonstrating pitting edema — **Look, press, release.** A persistent indentation confirms pitting. Edema alone does not identify its cause or measure cardiac filling pressure. Combine it with JVP, perfusion and the clinical context.
James Heilman, MD, Combinpedal, CC BY-SA 3.0. Unmodified.

Bedside likelihood ratios for HF

dyspnea in the ED

Finding	+LR (rules in)	−LR
Third heart sound (S₃ gallop)	≈ 11	0.88
Abdominojugular reflux	6.4	0.79
History of heart failure	5.8	0.45
Jugular venous distension	5.1	0.66
Initial clinical judgment	4.4	0.45
Rales · any murmur	2.8 · 2.6	—
Dyspnea on exertion	1.3	0.48 (absence helps exclude)

S₃ and a raised JVP / positive AJR are the highest-yield rule-in signs; no single finding is sensitive enough to exclude HF — combine them.

Read congestion & perfusion

the two axes

Congestion ("wet")

↑ JVP (estimate from the internal jugular at 45°), AJR / hepatojugular reflux, S₃, orthopnea/PND.
Edema, ascites, hepatic congestion; rales (insensitive — often absent in chronic HF); loud P₂; Valsalva square-wave.

Low perfusion ("cold")

Narrow / proportional pulse pressure (PP/SBP < 25%), cool extremities, pulsus alternans.
Sleepy/obtunded, symptomatic hypotension on ACEi, declining serum Na⁺, worsening renal function.

The Stevenson 2×2 → action

profile drives therapy

	Dry (no congestion)	Wet (congested)
Warm	Warm-dry — compensated; optimize oral GDMT	Warm-wet — diurese ± vasodilator
Cold	Cold-dry — cautious fluid; reassess; low output	Cold-wet — inotrope ± vasodilator + diuresis; consider tMCS

Profiles predict outcome & guide acute therapy (→ Part G P.19, ADHF). Cold-wet is the highest-risk profile — escalate (→ Cardiogenic Shock card).

Technique — how to elicit the signs

do it right

JVP internal jugular at 30–45°; height above sternal angle + 5 cm = CVP; > 8–9 cm H₂O is elevated.
AJR firm RUQ pressure 10–15 s → a sustained JVP rise > 3–4 cm = positive.

Valsalva a square-wave BP (sustained during strain, no overshoot) = high LV filling pressure.
Proportional PP (SBP − DBP)/SBP; < 25% → CI < 2.2 (low output).

Exam pearls: a positive AJR or visible JVP is the most reliable bedside read of elevated filling pressure; rales are insensitive in chronic HF (lymphatics adapt); the proportional pulse pressure < 25% flags low output; a Valsalva square-wave response indicates high LV filling pressure. Correlate with the invasive numbers on P.24–25.

The ED job in acute decompensated HF: stabilize, profile (perfusion × congestion × BP), and act in the first hour — then disposition. Most ADHF is warm & wet; the dangerous minority is cold (hypoperfused) or SCAPE (flash pulmonary edema).

▤

First-hour pathway

stabilize → profile → act

First-hour actions

dose it right

Lever	How
NIV (CPAP/BiPAP)	Early for respiratory distress / SCAPE — ↓ work of breathing & preload; avoids intubation.
IV loop diuretic	Naïve: furosemide 40 mg IV. On a home loop: give ≥ the total daily oral dose IV (high-dose ~2.5× = more decongestion, transient ↑Cr, no worse outcomes — DOSE); bolus or infusion comparable.
Vasodilator (NTG)	SBP high / SCAPE → sublingual then IV nitroglycerin drip (afterload ↓); not if hypotensive / preload-dependent.
Pressor / inotrope	Cold / hypoperfused → norepinephrine ± dobutamine; activate the shock team (P.20).
β-blocker	Don't start in active decompensation; don't abruptly stop chronic therapy unless cardiogenic shock — halve / continue.
Avoid	Reflexive large fluid boluses; routine nesiritide; stopping GDMT for an asymptomatic low BP / small Cr rise.

Disposition

level of care

ICU/CCU shock · SCAPE needing NIV/high-dose NTG · refractory hypoxia · unstable arrhythmia · troponin/ACS · need for close titration.
Floor / telemetry most warm-wet responding to IV diuretic.

Obs → home only if low-risk, rapid response, reliable 7-day follow-up (STRONG-HF early up-titration).
Before discharge near-euvolemia, stable oral regimen ≥24 h, GDMT started/optimized.

Special populations

tailor it

Population	Key points
Peripartum CM	Late-pregnancy / postpartum new HF; avoid ACEi/ARB/ARNI antepartum (use hydralazine/nitrate, β-blocker); anticoagulate if EF very low; bromocriptine considered with a team.
Cancer-therapy cardiotoxicity	Anthracycline / HER2 / ICI — start GDMT and coordinate hold/resume decisions with the treating specialist team.
CKD / dialysis	Diuretic resistance → sequential nephron blockade / UF or dialysis for refractory congestion; protect access; time fluid removal.
Frailty / elderly	Still benefits from GDMT — start low, avoid hypotension/AKI; deprescribe non-cardiac meds; align with goals of care.

Refer early — "I-NEED-HELP": Inotropes · NYHA IIIB–IV / persistent ↑NP · End-organ dysfunction · EF ≤35% · Defibrillator shocks · Hospitalizations >1 · Edema despite escalating diuretics · Low BP / high HR · Prognostic-med intolerance. Any of these → advanced-HF referral (LVAD · Transplant · VA-ECMO cards) before multi-organ failure.

Start all four pillars before discharge when clinically feasible. In the conventional HFrEF treatment frame (LVEF ≤ 40%), each pillar is Class 1; low doses across all four classes generally outrank maximizing only one. In-hospital initiation plus rapid up-titration and close follow-up in STRONG-HF reduced death or HF readmission. An all-generic regimen can be inexpensive, but coverage, dose, pharmacy, and assistance eligibility vary. Check SBP · HR · K⁺ · Cr; arrange reassessment in 1–2 weeks. Adverse-event comparisons below are trial-specific: ARNI versus enalapril, the other pillars versus placebo. Read within each row.

▦

The four pillars — HFrEF (LVEF ≤ 40%)

benefit · safety · access

Pillar · drug · start → target	Benefit (trial)	AE vs comparator — actual %	Don't start / hold	Illustrative cash
1 ARNI sacubitril/valsartan 24/26 or 49/51 → 97/103 BID. Symptomatic hypotension → reassess volume and concomitant drugs, then adjust dose.	PARADIGM-HF vs enalapril: CV death/HF hospitalization ↓20% (HR 0.80); NNT ≈ 21.	Hypotension 14% vs 9%; angioedema ~0.5% vs 0.2%. Lower than enalapril: hyperkalemia, renal dysfunction, cough.	36-hour ACEi washout; prior angioedema; pregnancy; SBP <100; K⁺ >5.2; caution with eGFR <30.	Generic available. Compare exact fill quotes. Entresto copay program discontinued.
2 Beta-blocker carvedilol 3.125→25 BID · metoprolol succinate 12.5→200 qd · bisoprolol 1.25→10 qd	~34% ↓ mortality each vs placebo (COPERNICUS · MERIT-HF · CIBIS-II); NNT 18–26.	Bradycardia, hypotension, fatigue; withdrawal and worsening HF were not increased versus placebo.	Active decompensation / shock / inotropes; initiate when euvolemic; avoid with high-grade AV block or symptomatic bradycardia.	generic ~$12; often on low-cost lists
3 MRA spironolactone 12.5–25→25–50 qd · eplerenone 25→50 qd	RALES (spironolactone) death ↓30%, NNT 9; EMPHASIS-HF (eplerenone) HR 0.63, NNT 13.	Hyperkalemia (K⁺ >5.5: 11.8% vs 7.2%). Gynecomastia 10% vs 1% with spironolactone; not increased with eplerenone.	K⁺ ≥5.0 or eGFR ≤30 at initiation; recheck K⁺ and creatinine at day 3 and 1 week.	spironolactone ~$12; eplerenone ~$18
4 SGLT2 inhibitor dapagliflozin / empagliflozin 10 mg qd	DAPA-HF / EMPEROR-Reduced: worsening HF/CV death ↓~25% (HR 0.74 / 0.75), NNT 19–21; benefit independent of diabetes.	Genital mycotic infection increased (single-digit %). Volume depletion, AKI, amputation, and DKA were not increased versus placebo in these HF trials.	Type 1 diabetes; recurrent DKA; eGFR below current initiation threshold; hold during sick days / prolonged fasting; reassess loop dose.	generic dapagliflozin ~$13; empagliflozin brand $0–10 copay / ~$600

Coverage before sequence: get a low dose of all four on board, then up-titrate every 1–2 weeks toward evidence-based targets. ACEi/ARB substitutes for ARNI only if ARNI is not feasible; never give an ACEi within 36 hours of an ARNI.

◐

HFpEF / mildly reduced EF — discharge therapy

congestion and comorbidities

Drug	Class	Evidence and practical notes
SGLT2 inhibitor — dapagliflozin / empagliflozin 10 mg qd	ESC I US 2a	The cornerstone across the EF spectrum. DELIVER and EMPEROR-Preserved: fewer HF hospitalization/CV-death events (HR ~0.79–0.82). Same dose and cautions as in HFrEF.
Finerenone — nonsteroidal MRA	FDA ACC pathway	FINEARTS-HF (LVEF ≥40%, N=6,001): fewer total worsening-HF events plus CV death (RR 0.84). 20–40 mg daily (start 10–20, titrate by eGFR/K⁺). Watch hyperkalemia (~2× placebo); do not initiate if K⁺ >5.0 or eGFR <25. U.S.-approved for HF with LVEF ≥40%, independently of diabetes or CKD.
Diuretic + comorbidity care	1	Loop diuretic for congestion; actively treat BP, AF, obesity, ischemia, and sleep apnea. Spironolactone (TOPCAT) and ARNI (PARAGON-HF) have more modest evidence, with benefit greatest near the lower end of the EF range.

Selected HFrEF adjuncts

add only for the matched indication

Add-on	When	Class	Trial and practical note
Ivabradine	Sinus rhythm, HR ≥70 on max-tolerated beta-blocker, EF ≤35%	2a	SHIFT: fewer HF hospitalizations. Label rates: bradycardia 10% vs 2.2%, phosphenes 2.8% vs 0.5%. No benefit in AF.
Hydralazine + isosorbide dinitrate	Self-identified Black, NYHA III–IV on GDMT	1	A-HeFT: mortality ↓43% (6.2% vs 10.2%). TID dosing burden; headache and hypotension.
Vericiguat	Recent worsening (HF hospitalization or IV diuretic), EF <45%	2b	VICTORIA: modest reduction in CV death/HF hospitalization (HR 0.90). Avoid with PDE5 inhibitors and riociguat; contraindicated in pregnancy.
Digoxin	Persistent symptoms despite GDMT; rate-control adjunct in AF	2b	DIG: no mortality benefit, fewer hospitalizations. Narrow therapeutic index; target 0.5–0.9; renal dose and monitor K⁺/Mg.
Loop diuretic	Congestion — symptom relief, not mortality benefit	1	TRANSFORM-HF: torsemide and furosemide had similar outcomes; choose by response, oral bioavailability, and access.

Cost and access — solve the plan before discharge

confirm at dispensing

Compare the complete prescription

Same product and fill compare strength, tablet count, days supplied, shipping and dispensing fees.
Insurance versus cash a discount used outside Part D does not count toward its deductible or annual cap.
Dated examples the linked patient handouts retain their September 5, 2026 price checks. Obtain a current quote before choosing a pharmacy.

Coverage

Commercial insurance verify current manufacturer support terms. Entresto’s former Co-Pay Program is discontinued. Other commercial offers generally exclude Medicare and Medicaid.
Uninsured / low-income check current manufacturer assistance, independent medication-access resources, and generic cash options; eligibility and pricing change.

Medication in hand

Close the access loop confirm coverage, prior authorization, dispensing pharmacy, and feasible pickup or delivery before discharge.
Use clinical pharmacy for medication reconciliation, access barriers, ongoing GDMT titration, and monitoring when available.

CHECK

Before signing the discharge: (1) all four pillars started—or a documented reason not; (2) starting doses plus a written up-titration and lab-recheck plan; (3) baseline K⁺ / creatinine / BP / HR; (4) affordability addressed with a workable coverage, cash, or assistance route; and (5) dispensing and pickup or delivery confirmed so the patient leaves with the medications. Any cash figures are estimates, not guaranteed prices. Adverse-event definitions differ between trials; compare within a row.

Treat the iron deficit, not only the hemoglobin. In symptomatic HFrEF/HFmrEF with iron deficiency, IV iron has an established functional and quality-of-life benefit; fewer HF hospitalizations are probable; a mortality benefit is unproven. Anemia is not required. Evidence is formulation-specific and does not make every IV iron product interchangeable.

Find it - frame the candidate - fix the cause

CBC + ferritin + TSAT

ScreenCBC, ferritin, TSAT at HF diagnosis, after decompensation, and periodically. Add reticulocytes, B12/folate, hemolysis testing, or smear when the anemia phenotype needs it.

DefineHF trial definition: ferritin <100 ng/mL, or 100-299 + TSAT <20%. Ferritin rises with inflammation; a normal-looking ferritin does not rescue a low TSAT.

PhenotypeBest-supported: symptomatic LVEF ≤45%, including stabilized acute HF. HFpEF data remain small and hypothesis-generating. Do not withhold therapy merely because Hb is normal.

InvestigateOccult GI or menstrual loss, anticoagulant-associated bleeding, malabsorption, low intake, CKD, inflammation, and malignancy. IV iron complements GDMT and source control; it replaces neither.

Evidence lane	Guideline / trial signal	What to tell the patient and team
US 2a B-R HFrEF	IV iron is reasonable with or without anemia.	Expect better functional status and quality of life. FAIR-HF and CONFIRM-HF improved symptoms, walk distance, or QoL.
ESC I A symptoms/QoL ESC IIa A HF hospitalization	Symptomatic HFrEF/HFmrEF; FCM or ferric derisomaltose for the hospitalization objective.	AFFIRM-AHF and IRONMAN narrowly missed their primary composites but support fewer HF admissions; the 2025 trial-level/IPD analysis strengthens that signal.
Post-guideline reality	HEART-FID and FAIR-HF2 were neutral for their prespecified primary tests.	Do not promise survival. Use the functional/QoL goal, probable hospitalization benefit, iron trajectory, treatment burden, and patient preference.
Not oral replacement for HF benefit	IRONOUT HF: oral iron minimally replenished stores and did not improve peak VO2.	Oral iron can still have a non-HF indication, but it is not the evidence-based substitute for IV iron in symptomatic iron-deficient HFrEF.

IV formulations - label, course, delivery, and the trap

verify current label + validated order

Product	U.S. label lane	Labeled delivery / course	HF role + formulation-specific trap
Ferric carboxymaltose Injectafer	HF ON-LABEL Adult NYHA II/III ID, improve exercise capacity; also IDA and NDD-CKD.	Use the HF grid, not the IDA shortcut. Day 1: 1,000 mg if Hb ≤14; 500 mg if Hb >14 to <15. Week 6: <70 kg: 500 only if Hb <10; ≥70 kg: 1,000 if Hb <10 or 500 if Hb 10-14. Weeks 12/24/36: 500 mg if still ID. Infuse up to 1,000 mg over ≥15 min.	Most HF-specific efficacy data and the only direct U.S. HF label here. Symptomatic hypophosphatemia: correct pre-existing low phosphate; check high-risk/repeat-course patients. No label data for Hb ≥15 or beyond week 36.
Ferric derisomaltose Monoferric	NO HF LABEL IDA after oral intolerance/poor response; NDD-CKD.	≥50 kg: 1,000 mg once; <50 kg: 20 mg/kg once; dilute in 100-500 mL NS and infuse ≥20 min.	IRONMAN is the large HF outcomes trial. One-visit convenience, but payer review may run through IDA/CKD rather than HF. Monitor hypersensitivity.
Ferumoxytol Feraheme	NO HF LABEL IDA after oral intolerance/poor response or CKD.	510 mg x2, 3-8 days apart; each over ≥15 min; reassess iron indices ≥1 month after dose 2.	Boxed anaphylaxis warning. Prior allergic reaction to any IV iron is a contraindication. Alters MRI for up to 3 months; coordinate cardiac MRI timing.
Iron sucrose Venofer	CKD LABEL IDA in CKD.	NDD-CKD: 200 mg x5 over 14 days; IV push over 2-5 min or infusion over ≥15 min.	Often preferred for price/step therapy, but has no large HF outcomes program and creates five visits for a 1-g course. Do not check serum iron within 48 hours.
Low-MW iron dextran INFeD	NO HF LABEL Documented ID after oral failure/intolerance.	Calculate deficit. First dose: 25 mg test over ≥30 sec, wait ≥1 h; current label gives ≤100 mg/day at ≤50 mg/min.	Boxed fatal-anaphylaxis warning. A negative test does not eliminate risk. One-visit 1-g total-dose infusion is not the current labeled regimen; delayed fever/myalgia can follow large doses.
Ferric gluconate Ferrlecit	HD + EPO IDA in hemodialysis with supplemental epoetin.	125 mg each dialysis x8; 100 mL NS over 1 h or undiluted at ≤12.5 mg/min.	A dialysis-course product, not a direct ambulatory HF pathway. Doses above 125 mg have more reported adverse events.

Reaction readiness: give every formulation where trained staff and resuscitation treatment are immediately available; follow the product's monitoring/observation instructions. Products differ in carbohydrate shell, maximum single dose, infusion volume/time, warnings, labeled population, evidence, and billing unit - do not substitute milligram-for-milligram without a new order.

Coverage + delivery - obtain an approval you can actually infuse

product + total mg + visits + site + expiration

The cleanest HF pathway: ferric carboxymaltose is the only formulation above with a direct U.S. HF indication. Other products may be preferred but reviewed under IDA/CKD criteria or plan-supported off-label evidence. Do not code anemia when anemia is absent.

Prior-authorization packet

Patient + phenotype plan/benefit, age, symptoms, NYHA class, LVEF, weight, CKD/dialysis status.
Timed labs CBC/Hgb, ferritin, TSAT; prior IV product/response/reaction; oral failure/intolerance only if true.
Exact request brand/generic, mg per dose, total course, visit count, infusion site, prescriber, and evidence/label lane.
Approval lock authorization number, approved product, total mg, visits, site, expiration, buy-and-bill vs specialty-pharmacy supply.

Medicare is not a blanket HF guarantee. Part B can cover qualifying provider-administered drugs when the service and drug are reasonable/necessary. NCD 110.10 creates a national rule for specified dialysis use; it does not create national HF-specific coverage. MAC and Medicare Advantage rules can differ.

Plan and site-of-care variation

Current policy examples Aetna recognizes an HF ferric-carboxymaltose pathway; UHC may add Hb and preferred-product gates. A code alone never guarantees payment.
Site steering hospital outpatient may be redirected to office/freestanding/home infusion. Document why higher-acuity care is needed: prior severe reaction, cardiopulmonary instability, fluid intolerance, or difficult access.
Total episode burden compare drug acquisition + facility fees + transportation + time: a lower-cost vial can become a five- or eight-visit course.

ORDER

Before signing: confirm the current label and validated order, deficiency and bleeding work-up, treatment goal, product-specific dose/course, dilution/rate, reaction plan, phosphate strategy when using repeat/high-risk FCM, MRI timing with ferumoxytol, authorization terms, and the owner/date for follow-up labs. Do not interpret immediate post-infusion serum iron/TSAT as durable repletion. Same patient + same labs can yield a different covered product under a different plan.

Device
Rule

Phenotype → mechanism → endpointChoose a device only after naming the problem it can change and the outcome its evidence actually supports.

Sequence before addingConfirm HF phenotype; optimize GDMT, rhythm, ischemia and valves; assess ICD/CRT; then consider adjunct devices.

Status is part of the prescriptionFDA HF indication, FDA indication for another condition, and research-only are not interchangeable.

Do not delay advanced-HF careRecurrent admissions, shock, inotrope dependence, end-organ dysfunction or escalating diuretic need should trigger advanced-HF referral.

Three regulatory and evidence lanes

United States · verify current IFU before implant

HF indication

Available for selected HF phenotypes

ICD/CRT for guideline-defined electrical risk; CardioMEMS/Cordella for pressure-guided management; Barostim and CCM for symptom/function endpoints in specific HFrEF groups.

Other indication

Relevant to HF, not an HF therapy claim

remedē has a non-HF CSA indication. Ultrafiltration is a fluid-removal technology with device-specific clearance. Neither has proven HF survival benefit.

Trial only

Mechanistic and investigational

Interatrial shunts, splanchnic ablation and most acute cardiorenal platforms remain investigational. Physiologic improvement is not clinical efficacy.

DRI²P²S decoded

A taxonomy-not a guideline or treatment bundle

D Dilators

Increase venous capacitance; redistribute stressed volume.

R Removers

Directly remove or algorithmically mobilize sodium and water.

I¹ Inotropes

Device-based neural stimulation intended to augment contractility.

I² Interstitials

Facilitate lymphatic/interstitial fluid drainage.

P¹ Pushers

Raise renal arterial pressure with descending-aortic pumps.

P² Pullers

Lower central or renal venous pressure.

S Selective

Deliver short-acting drugs directly to the renal arteries.

! Boundary

Most DRI²P²S examples are investigational acute-HF concepts.

What problem remains after optimized care?

U.S. labels summarized · verify current IFU

Unresolved phenotype	Candidate device lane	Expected evidence-based endpoint	Wrong expectation / stop sign
LVEF ≤35% with sudden-death risk after chronic GDMT.	ICD when full guideline criteria and survival horizon are met.	Reduce arrhythmic death.	Does not improve congestion, NYHA class or pump function by itself.
LVEF ≤35% + electrical dyssynchrony, strongest with LBBB/QRS ≥150 ms.	CRT / selected physiologic pacing.	Symptoms, remodeling, HF events, survival.	Do not bypass for BAT/CCM; ensure high effective BiV/physiologic capture.
NYHA II-III + prior-year HF admission and/or elevated NP; recurrent occult congestion.	CardioMEMS.	Fewer HF hospitalizations with an actionable pressure program.	No team workflow, no transmissions, or a promise of mortality reduction.
NYHA III at home on diuretic/GDMT, stable 30 days; comprehensive remote pathway.	Cordella.	Assessment/management to reduce HF hospitalization.	Single-arm evidence described as randomized superiority.
NYHA III or II-with-recent-III HFrEF; LVEF ≤35%; NT-proBNP <1600.	Barostim BAT.	Functional status, 6MWD, QoL.	Significant carotid/contraindicated anatomy; unaddressed CRT; promised survival benefit.
NYHA III; LVEF 25%-45%; not receiving CRT; symptom/function goal.	CCM.	6MWD, QoL, functional status.	CRT phenotype, unsuitable transvenous system or hard-outcome claim.
PSG-confirmed moderate-to-severe predominant CSA in an adult.	remedē.	CSA/sleep metrics and selected QoL benefit.	Predominant OSA, active infection, or labeling it an HF outcome therapy.
Severe secondary MR or severe symptomatic TR after phenotype-directed medical/device therapy.	Structural Heart Team for label- and anatomy-specific transcatheter/surgical options.	Device-specific symptom/HF-event and valve endpoints; selected COAPT-like secondary MR has randomized outcome benefit.	Calling every MR/TR an HF-device indication or intervening before GDMT, CRT/revascularization and RV/pulmonary assessment.
Stage D trajectory: recurrent shock/admissions, inotrope or end-organ decline.	Advanced-HF pathway: LVAD/transplant/palliative goals.	Survival, function and goal-concordant care in eligible patients.	Serial add-on devices that postpone destination decisions.

Order
Matters

CRT is not a competing symptom gadgetFor a qualifying electrical phenotype, CRT improves symptoms and can reduce HF events and death. Assess it before BAT or CCM.

ICD prevents arrhythmic deathIt does not treat congestion or improve exercise capacity. Discuss competing nonarrhythmic mortality and goals of care.

Reassess after optimizationUse chronic GDMT, recovery potential, timing after MI/revascularization and meaningful expected survival before primary-prevention implantation.

High-yield selection anchors

Confirm the complete guideline phenotype

Device	Canonical phenotype	Proven target	Do not oversimplify
Primary-prevention ICD	Selected ischemic/nonischemic cardiomyopathy with persistently low LVEF despite chronic GDMT and meaningful expected survival >1 year.	Sudden arrhythmic death.	Use cause, NYHA class, post-MI/revascularization timing, comorbidity and patient preference—not LVEF alone.
CRT	Strongest evidence: LVEF ≤35%, sinus rhythm, LBBB, QRS ≥150 ms, NYHA II-ambulatory IV on GDMT.	Symptoms, remodeling, HF events and survival.	Benefit varies by QRS morphology/duration; AF and pacing-indication pathways require capture strategy and EP judgment.
Conduction-system pacing	Selected need for physiologic pacing, CRT alternative/rescue, or anticipated substantial RV pacing.	Preserve/restore synchrony.	His-bundle and left bundle branch area pacing are not automatically interchangeable with trial-proven biventricular CRT.
Wearable defibrillator	Selected transient high-risk interval while recovery/definitive ICD candidacy is clarified.	Temporary shock protection.	Adherence is the therapy; routine use after every new low LVEF diagnosis is not supported.

Before referring for BAT or CCM

A six-gate check

1 · Verify phenotype Current echo, QRS morphology/duration, rhythm, NYHA class and objective limitation; correct reversible causes.

2 · Build GDMT Maximize tolerated evidence-based therapy; document barriers rather than labeling a patient “optimized” after one visit.

3 · Treat higher-value targets Revascularization/ischemia, severe valve disease, AF/tachycardia, iron deficiency, sleep-disordered breathing and uncontrolled BP.

4 · Decide ICD/CRT Complete EP assessment before an add-on device whose evidence is limited to symptoms/function.

5 · Name desired endpoint Hospitalization reduction (pressure-guided care) versus symptom/function improvement (BAT/CCM) versus CSA control (remedē).

6 · Check advanced-HF triggers Recurrent admissions, low output, hypotension, escalating diuretics, hyponatremia, organ dysfunction, intolerance of GDMT or peak VO₂ decline → advanced-HF referral in parallel.

BOTTOM LINE

Outcome-changing electrical therapy comes before adjunct symptom devices. “Narrow QRS” explains why CRT may not help; it does not by itself establish a BAT or CCM indication.

Four electrical chapters

Open the exact decision

ICD · P32	Primary vs secondary prevention, timing, transvenous vs subcutaneous system, programming, shocks and replacement decisions.
CRT · P33	Morphology/QRS matrix, AF and pacing pathways, CRT-P vs CRT-D, capture targets and structured nonresponse workup.
LBB-AP · P34	Candidate selection, HBP distinction, capture proof, programming, complications and evidence boundary.
WCD · P35	Bridge logic, LifeVest vs Kestra ASSURE, training/adherence workflow, evidence and a prespecified exit date.
Broader HF care	Use these device chapters alongside the comprehensive HF foundations, discharge GDMT and access chapter, and focused shock and advanced-HF guides.

ICD
Rule

Treat the substrate firstExclude a reversible cause; correct ischemia, electrolytes, drug toxicity and acute triggers; define etiology, scar/genetics and recovery potential.

Secondary prevention is not an EF ruleSurvived VT/VF arrest or sustained hemodynamically unstable/stable VT not due to a reversible cause generally supports ICD when meaningful survival >1 year is expected.

Primary prevention needs timeReassess after chronic GDMT and the applicable post-MI/revascularization recovery window; implanting early because the EF may recover can add harm without proven survival benefit.

An ICD is not HF therapyIt terminates malignant VT/VF. It does not decongest, improve exercise capacity or reverse remodeling unless the system also delivers indicated CRT.

Who crosses the implant threshold?

U.S. guideline anchors · individualize beyond EF

Lane	Core phenotype	Timing / gate	Boundary
Secondary prevention	Prior SCA from VT/VF or spontaneous sustained VT, unstable or stable, without a reversible cause.	Evaluate/treat ischemia and trigger; implant if recurrent arrhythmic risk remains and meaningful survival >1 year is expected.	Do not withhold solely because LVEF is >35%; do not implant for a completely reversible episode without persistent substrate.
Primary · NYHA II-III	Nonischemic DCM or ischemic heart disease; LVEF ≤35% despite chronic GDMT.	At least 40 days after MI; ordinarily reassess about 90 days after revascularization and 3 months after new cardiomyopathy/GDMT.	Benefit falls as pump-failure/noncardiac death competes with arrhythmic death; require meaningful expected survival >1 year.
Primary · ischemic NYHA I	At least 40 days post-MI, LVEF ≤30% on GDMT.	The VA/SCD guideline also uses ≥90 days after revascularization before routine primary-prevention implantation.	Earlier post-MI ICD trials did not improve total survival; a temporary-risk discussion is a WCD question, not permission for early ICD.
Disease-specific risk	Arrhythmogenic/genetic cardiomyopathy, HCM, sarcoid, LMNA/FLNC/DSP and other substrates may qualify above the generic EF threshold.	Use the disease-specific guideline, CMR scar, genotype/phenotype, syncope and documented VA.	“NICM” is not one risk state; DANISH tempers blanket EF-only claims but does not erase indicated individualized prevention.
Do not implant / defer	Active infection, unresolved reversible cause, expected good-functional survival <1 year, or no goal-concordant shock benefit.	Reassess after recovery; consider WCD only when a defined transient high-risk bridge is plausible.	Advanced directive, replacement and shock-deactivation planning start before the incision—not after recurrent shocks.

Choose the system—not just “an ICD”

Match future pacing and lead burden

System	Prefer when	Cannot provide / key tradeoff	Preimplant proof
Transvenous ICD	Brady pacing, ATP, CRT or likely future pacing is needed; recurrent monomorphic VT may benefit from ATP. Add an atrial lead only for a real pacing/diagnostic need.	Intravascular infection, occlusion, tricuspid interaction and long-horizon lead failure/extraction burden.	Venous access, infection risk, pacing indication, tricuspid anatomy and lifetime lead plan.
Subcutaneous ICD	ICD indication with poor venous access/high infection risk and no anticipated brady pacing, ATP or CRT.	No chronic brady pacing, ATP or CRT; screening failure and inappropriate sensing remain possible.	Surface-ECG screening in required positions; confirm no pacing/CRT need and discuss future conversion.
Extravascular ICD	FDA-approved substernal system can deliver defibrillation, cardioversion, ATP and pause-prevention without a transvenous lead.	No chronic brady pacing or CRT; pivotal evidence was single-arm and does not establish superiority over S-ICD or transvenous systems.	Anatomy, center expertise, pacing horizon and current Aurora EV-ICD labeling.
CRT-D	The patient independently meets both CRT and defibrillator goals.	More hardware and shock burden than CRT-P; CRT benefit does not automatically prove ICD benefit.	Use the CRT matrix on Page 33, then separately weigh ischemic/scar/genetic risk, age, comorbidity and preference.
No new generator / downgrade	Recovered EF, no prior VA, changing prognosis or goals may alter value at elective replacement.	Risk is not zero after EF recovery; replacement is a new decision, not an automatic battery exchange.	Review stored VA, prior therapies, EF/scar, competing mortality, lead status, infection risk and shock-deactivation preferences.

Program, monitor and respond

Every shock deserves a diagnosis

DELIVERY

Reduce avoidable therapy

EP-owned

Use disease- and device-specific high-rate and/or longer-detection strategies when appropriate; program ATP before/while charging for organized VT when safe. Do not copy one universal zone.
Turn on validated SVT discrimination; document detection zones, delays, ATP sequences, shock energies and post-shock pacing.
Remote monitoring: lead integrity, impedance, R-wave, thresholds, battery, arrhythmia burden and delivered/aborted therapies.

AFTER THERAPY

Interrogate before labeling

Urgent if unstable

A counter or alert is a device classification—not yet a rhythm diagnosis. Before anticoagulation, antiarrhythmic change, or judging ICD therapy, inspect stored atrial/ventricular electrograms and reconstruct onset, duration, detection/discriminators, therapy, termination, lead trends, and symptoms. Far-field sensing, SVT, T-wave/lead-noise oversensing, programming, or transmission artifact can imitate AF, VT, or device failure.
Inappropriate therapy: treat AF/SVT, oversensing or lead failure and reprogram. Appropriate therapy: treat trigger and substrate; recurrent VT/electrical storm needs urgent EP/HF care, AAD/ablation pathway and sympathetic control. Shock with instability, syncope, injury, or repeated therapies requires urgent assessment; absent/truncated electrograms may leave mechanism unresolved.
Pocket erosion/purulence, bacteremia, or lead/valve infection: obtain cultures, image as indicated, and involve an extraction team early; antibiotics around retained infected hardware are not source control. In selected higher-risk revisions, replacements, upgrades, or initial CRT-D procedures, an antibacterial envelope reduced major infection in WRAP-IT; it supplements prophylaxis, sterile technique, hematoma avoidance, and sound indication—and never justifies implanting through active infection.

Lifetime burden and evidence boundary

Benefit is not EF alone

Procedure / lead	Hematoma, pneumothorax/hemothorax, vascular injury, perforation/tamponade, infection/erosion, lead failure, venous obstruction, tricuspid interaction, generator changes and eventual extraction risk.
Evidence	MADIT-II and SCD-HeFT established selected primary-prevention benefit but predated ARNI/SGLT2 therapy. In DANISH, ICD reduced sudden death without reducing all-cause mortality; age and competing risk inform—not dictate—the decision.
Borderline / recovered EF	Confirm an EF near 35% with high-quality repeat imaging. “Appropriate” in an AUC is not obligatory; after recovery, new illness or battery depletion, reassess VA history, scar/genetics, prognosis and preferences.

CONSENT

Compare arrhythmic death risk with pump-failure and noncardiac death. Discuss shocks, inappropriate therapy, lead/generator procedures, driving restrictions, psychological burden, remote follow-up and the right to deactivate shock therapy later.

CRT
Rule

Strongest phenotypeLVEF ≤35%, sinus rhythm, true LBBB, QRS ≥150 ms and NYHA II-ambulatory IV symptoms on GDMT: CRT improves symptoms, HF events and survival.

QRS morphology changes benefitLBBB and wider QRS predict more benefit. Non-LBBB, shorter QRS and mild symptoms need a stricter evidence conversation.

Delivery is the treatmentA device labeled “CRT” is ineffective if fusion/pseudofusion, AF, PVCs, lead failure, scar or programming prevents near-continuous resynchronizing capture.

Pacing and defibrillation are separateChoose CRT-P versus CRT-D only after independently assessing SCD risk, age, scar/etiology, comorbidity, prior VA, lead burden and patient goals.

Guideline selection matrix

Do not collapse all wide QRS into one indication

Phenotype on GDMT	U.S. recommendation anchor	Bedside translation
LBBB ≥150 ms; LVEF ≤35%; sinus; NYHA II-ambulatory IV	Class I · A	Highest-certainty CRT phenotype; deliver BiV CRT unless an anatomy/functional problem requires a physiologic-pacing strategy.
LBBB 120-149 ms; LVEF ≤35%; sinus; NYHA II-ambulatory IV	Useful · 2a	Benefit is more selective; the 2023 physiologic-pacing guideline gives Class I support to BiV CRT in selected characteristics such as female sex.
Non-LBBB ≥150 ms; LVEF ≤35%; NYHA II-ambulatory IV	2a / 2b by class	More convincing in NYHA III/IV and IVCD than RBBB; counsel that response and outcome evidence are weaker than classic LBBB.
AF; LVEF ≤35%; pacing need or other CRT criteria	Useful · 2a	Proceed only with a plan for near-100% effective ventricular capture—rate control and often AV-node ablation when conducted AF defeats delivery.
High-grade/complete AV block; LVEF 36%-50%	Reasonable · 2a	Avoid chronic dyssynchronous RV pacing; BiV CRT or selected CSP depends on EF, pacing burden, anatomy and expertise.
LVEF ≤35% with anticipated >40% ventricular pacing	Useful · 2a	At new or replacement implant, use resynchronizing/physiologic pacing rather than create or worsen pacing-induced cardiomyopathy.
QRS <120 ms, or NYHA I-II non-LBBB <150 ms	No benefit	Do not implant routine CRT for mechanical dyssynchrony alone; exceptions arise from a separate brady/pacing pathway, not the narrow-QRS HF indication.

Implant-to-response pathway

Prespecify what “working” means

1 · Verify before implant 12-lead QRS morphology/duration, rhythm and burden, current LVEF/volumes/MR/RV, GDMT and recovery window; assess venous anatomy, scar/ischemia, infection and pacing/ICD needs.

2 · Deliver resynchronization Stable LV lead with acceptable capture margin, no clinically limiting phrenic stimulation and an electrically useful, non-scarred target when feasible; obtain a 12-lead paced ECG.

3 · Prove effective capture Interrogate percent pacing but also inspect ECG/EGMs for fusion and pseudofusion. Aim as close to 100% effective BiV/physiologic capture as possible; a fall below roughly 95%-98% demands an explanation.

4 · Reassess at 3-6 months NYHA/KCCQ or functional target, HF events, ECG/QRS, device diagnostics and echo remodeling; earlier review for symptoms, lead alert, phrenic stimulation, AF/PVCs or HF admission.

“Nonresponse” is a differential diagnosis

Troubleshoot before declaring biology

Failure domain	Find it	Correct it
Delivery	Low effective capture, AF with rapid conduction, frequent PVC/VT, lead dislodgment/high threshold, intermittent phrenic avoidance.	Rate/rhythm or PVC/VT therapy, AV-node ablation when appropriate, output/vector change, lead revision or alternative pacing site.
Electrical timing	Paced QRS not narrowed, fusion/pseudofusion, long intrinsic AV conduction, suboptimal AV/VV timing.	ECG-guided programming and device-specific optimization; avoid endless empirical changes without a measurable endpoint.
Substrate	Posterolateral scar, severe RV failure/PH, untreated ischemia, severe MR, progressive cardiomyopathy.	Reimage/revascularize when indicated; reconsider lead target, CSP or hybrid HOT/LOT-CRT; address valve/advanced-HF pathway.
Not the device	Residual congestion, absent GDMT, anemia/iron deficiency, OSA/CSA, lung disease, frailty or poor adherence.	Treat the competing limiter. Reverse remodeling without symptom gain—or symptom gain without remodeling—still needs an endpoint-specific interpretation.

Procedure and evidence boundaries

A response device still has procedural cost

BiV-specific hazards	Failed/unstable CS lead, coronary-sinus dissection or perforation, phrenic stimulation, high threshold/accelerated battery depletion, ineffective capture, infection and venous occlusion.
CRT-P vs CRT-D	CARE-HF proved morbidity/mortality benefit from CRT without a defibrillator; COMPANION and RAFT support CRT strategies but do not make CRT-D automatic. No adequately powered head-to-head trial proves universal CRT-D superiority.
Durability	Continue GDMT and reassess symptoms, admissions, EF/volumes/MR and device delivery. EF recovery does not automatically erase resynchronization need; generator exchange, upgrade or downgrade is a fresh goal/risk decision.

RESCUE

If a coronary-sinus LV lead cannot be placed or effective BiV CRT cannot be achieved, HBP or LBB-AP is a guideline-supported crossover. Do not let a technically present but ineffective LV lead end the resynchronization plan.

LBB-AP
Rule

Name what was capturedLBB-area pacing (LBBAP/LBB-AP) includes true left bundle branch pacing (LBBP) and LV septal pacing. A deep septal lead alone does not prove conduction-system capture.

HBP is differentHis-bundle pacing is more proximal and can correct block, but commonly has higher thresholds, smaller R waves and more late threshold rise than LBB-AP.

BiV CRT remains the benchmarkClassic LBBB/HFrEF has mature mortality/HF-event evidence for BiV CRT. Three 2026 RCTs gave conflicting CSP results; routine first-line substitution is not supported.

Expertise is part of the indicationUse a trained implanting program able to document capture, manage septal complications, program the chosen port configuration and cross over to BiV CRT when needed.

Where LBB-AP fits

Separate established rescue from emerging alternative

Clinical problem	Role	Evidence boundary / decision
BiV CRT candidate; CS lead cannot be implanted or is anatomically/electrically suboptimal.	Rescue · 2a HBP with BBB correction or LBB-AP is reasonable.	Document why effective BiV CRT failed and prove CSP capture/correction; a bailout is not a failed implant.
LVEF ≤35%, sinus, LBBB ≥150 ms, NYHA II-IV on GDMT.	Alternative · 2b CSP may be considered; BiV CRT is Class I.	Use shared decision-making: less long-term/hard-outcome evidence, operator dependence and extraction uncertainty versus physiologic activation and simpler lead geometry.
Pacing indication; LVEF 36%-50%; anticipated ventricular pacing roughly 20%-40% or more.	Prevent PICM CPP with BiV CRT, HBP or LBB-AP is reasonable.	Expected pacing burden, baseline QRS/LBBB, EF and anatomy determine strategy; normal-EF LBB-AP is a selective Class 2b pathway, not mandatory.
Established pacing-induced cardiomyopathy or HF worsened by substantial RV pacing.	Upgrade BiV CRT recommended; CSP revision can be beneficial.	Confirm high RV pacing is causal, then compare venous/lead burden, tricuspid valve, scar and probability of capture.
AV-node ablation, infra-His block or high pacing dependence.	Pacing strategy LBB-AP often favored over HBP.	Plan backup pacing when loss of capture would be catastrophic; a functioning defibrillator indication still requires a defibrillator-capable system.
HF with non-LBBB or LVEF >35% and no pacing indication.	Uncertain	Do not generalize a narrow-QRS/pacing cohort or QRS narrowing to proven HF survival benefit.

Prove capture at implant

12-lead ECG at baseline, high and low output

Claim	Required documentation	Supportive LBB-capture findings
LV septal pacing	Deep septal lead position by fulcrum sign/contrast/echo/CT plus an RV-conduction-delay pattern in V1 (rare exceptions).	This establishes LV septal activation—not necessarily LBB capture.
LBB-area pacing with LBB capture	LV septal criteria plus at least one validated capture criterion and output-dependent threshold testing.	Nonselective→selective LBB or septal transition; abrupt V6 RWPT change with loss/capture; short/constant V6 RWPT; V6-V1 interpeak >44 ms; potential-to-stimulus V6 RWPT concordance. No single metric is universally sensitive.
LBBB correction / CSP-CRT	Record native and paced 12-lead ECG, QRS width/morphology, correction threshold, myocardial threshold, sensing and device configuration.	Show reproducible BBB correction or a prespecified electrical target; distinguish capture from anodal fusion or LV septal-only pacing.

Program and surveil the deep septal lead

Port-aware · capture-aware · complication-aware

PROGRAM

Preserve the intended physiology

At discharge

State which port contains the CSP lead and whether an RV/CS backup lead exists; tailor AV/VV timing and upper rates to the clinical goal.
Program output with a margin above the documented LBB/correction threshold—not merely myocardial capture. Verify any automatic capture algorithm against the actual paced morphology.
Store a reference 12-lead ECG and EGM; record R wave, threshold/pulse width and impedance for trend comparison.

Interrogate physiology, not only battery

Early + longitudinal

Before discharge and at follow-up: 12-lead paced morphology at working and threshold outputs, capture/correction, sensing, impedance and percentage pacing.
At 3-6 months for HF/CRT: symptoms/function, HF events and echo remodeling; review AF/PVC burden and effective pacing.
Rising threshold, impedance change, altered V1/V6 morphology or loss of correction → imaging/EP review; decide reprogram, revise, add backup or cross over.

Complications unique enough to name

Do not call every septal lead routine RV pacing

Acute	Septal perforation into LV, septal hematoma/dissection, coronary injury/fistula, new RBBB or complete AV block, lead dislodgment, tamponade and usual pocket/venous complications.
Late	Threshold rise/loss of LBB capture, delayed perforation, tricuspid interaction, lead fracture/dislodgment and uncertain long-term extraction behavior of deep septal leads.
Evidence	2026 RCTs conflict: HeartSync-LBBP favored LBBP for death/HF hospitalization, whereas LEFT-BUNDLE-CRT did not establish LBBAP noninferiority and PhysioSync-HF favored BiV on a hierarchical HF endpoint. Populations, capture definitions and CSP mix differ; BiV remains first-line for conventional LBBB CRT while larger confirmatory trials mature.

Expert-center operational anchors

Targets, not outcome-validated cutoffs

Implant target	Common consensus targets are bipolar R wave >4 mV and capture <1.5 V at 0.5 ms (ideally <1 V), but the final report must separately name myocardial, LBB and BBB-correction thresholds.
Cadence / hybrid	In-person paced 12-lead commonly at 4-8 weeks and 6 months, then individualized remote follow-up; echo at about 3-12 months. LOT-CRT can target marked residual electrical delay, but CSPOT was a 48-patient acute mechanistic study—not outcomes evidence.
Learning-curve signal	MELOS reported HF-indication success 82.2%, overall complications 11.7% and acute LV septal perforation 3.67% in early multicenter practice. These technique-era cohort rates are not contemporary patient-specific risk estimates.

WCD
Rule

Write the bridge in one sentence“Temporary VT/VF protection while ___ is treated/reassessed by ___ date; then implant, discontinue or revise the plan.” No endpoint means indefinite low-value wear.

Hours worn are the doseThe device cannot detect or shock when it is off, poorly fitted, disconnected or without skin contact. Fit, alarms, skin and battery workflow are clinical care.

Shock onlyA garment WCD does not provide bradycardia pacing, ATP, CRT, HF treatment or automatic EMS activation; it cannot treat asystole/PEA. It does not replace an indicated implant.

No head-to-head outcome trialLifeVest and Kestra ASSURE have separate FDA files and performance evidence; there is no randomized brand-versus-brand clinical-outcome comparison.

Use only for a plausible transient-risk lane

2017 VA/SCD guidance · current FDA label at fitting

Bridge scenario	Why temporary?	Exit criterion
ICD explanted for infection in a patient with prior SCA/sustained VA	High recurrent risk while bacteremia/device infection clears and reimplant strategy is defined; WCD is reasonable (Class 2a).	Source control, extraction completion, cultures/antibiotic plan and reimplant/no-reimplant decision.
Early after MI (<40 days) or revascularization (<90 days) with low EF	ICD is deferred while LV recovery and competing early nonarrhythmic mortality evolve; WCD may be reasonable (Class 2b).	Repeat EF and full primary-prevention assessment at the guideline time point.
New NICM, myocarditis or secondary cardiomyopathy	Etiology treatment/GDMT may recover EF; arrhythmic risk is heterogeneous, not automatic at every EF ≤35%.	Prespecified 6-12 week/approximately 3-month clinical and imaging reassessment, sooner for sustained VA.
Awaiting transplant, temporary systemic infection or another short contraindication to implant	Risk and implant feasibility can change; bridge must not delay advanced-HF destination planning.	Transplant/implant eligibility, infection clearance, prognosis or goal change.
Patient refuses ICD	Both U.S. labels include adults who refuse an implantable defibrillator, but refusal may persist and WCD burden may be indefinite.	Document informed refusal, revisit goals and alternatives; do not imply equivalent long-term evidence or pacing capability.

LifeVest vs Kestra ASSURE

Verify the dispensed model and latest IFU

Feature	ZOLL LifeVest family	Kestra ASSURE WCD	Clinical interpretation
FDA indication	Adults at risk for SCA who are not candidates for, or refuse, an ICD. Selected earlier 3000/3100/4000 labeling also covers pediatric patients meeting size/weight criteria; do not transfer legacy specifications to Model 5100.	Adults at risk for SCA who are not candidates for, or refuse, an implantable defibrillator; active implanted defibrillator is contraindicated.	The label is broad; guideline-supported bridge logic and patient-specific benefit still determine prescription.
Detection / divert	Continuous surface ECG; alarms precede therapy and a conscious patient uses the trained response controls to withhold a shock.	Four EGM channels plus marker channel; alerts precede therapy and a conscious patient can press the Alert Button to divert.	Teach the actual controls hands-on. A response control is for a conscious patient—not a caregiver or bystander.
Therapy	Model 4000 FDA labeling describes physician-programmable 75-150 J and up to five shocks; current-model settings must come from its IFU.	Original FDA-approved IFU specifies a nonconfigurable 170-J biphasic shock.	Energy/controls are model-specific engineering—not comparative-outcome evidence and not a stand-alone reason to choose a system.
Evidence	Large registries plus VEST, the only major randomized WCD outcome trial: after recent MI with LVEF ≤35%, WCD assignment did not significantly reduce arrhythmic death, the primary endpoint.	FDA approval used ACE-DETECT ambulatory detection/safety and ACE-CONVERT induced-conversion studies; ACE-DETECT was 30-day, single-arm and therapies were disabled because participants had active ICDs.	No head-to-head trial. Do not translate detection/conversion performance into proven mortality benefit.

Prescription-to-discontinuation workflow

The service is part of the device

1 · Prescribe precisely Indication, VT/VF thresholds/settings, bridge reason, expected wear, emergency instructions, treating EP/HF owner and fixed reassessment date.

2 · Fit and teach-back Correct garment/electrode contact, battery exchange/charging, bathing plan, alert recognition, response control, data transmission and emergency response; involve caregiver only with patient permission.

3 · Review early and repeatedly Contact in 24-72 hours for fit/skin/alarm problems; monitor daily wear time, noise/contact alerts, arrhythmias and transmissions. Refit or change strategy if adherence is not realistically achievable.

4 · Shock / syncope / no-wear Shock or syncope → emergency services/clinical assessment plus recording review; a bystander never presses the response control. Repeated alarms, skin injury, falls, anxiety or inability to hear/respond → urgent service and prescriber reassessment.

5 · Exit on time Repeat EF/risk assessment after recovery window; implant when indicated, discontinue when risk resolves, or document a goal-concordant alternative. Never auto-renew without re-phenotyping.

Numbers worth remembering—without cross-trial ranking

Different devices · generations · designs

VEST · n=2,302	Recent MI, LVEF ≤35%: arrhythmic death 1.6% vs 2.4% (RR 0.67; 95% CI 0.37-1.21; P=.18), so the primary endpoint was neutral. Median wear including zero-wear days was 18 h/day; appropriate shock 1.3%, inappropriate shock 0.6%.
ACE-DETECT · n=130	Thirty-day single-arm ASSURE study in people already protected by active ICDs; WCD therapy/actual shock alarms were disabled. Four qualifying ICD-recorded VT/VF episodes were detected; three false-positive shock-alarm markers occurred across 3,501 patient-days; median wear 23 h/day.
Translation	VEST studied an earlier LifeVest generation and ACE-DETECT did not test survival or ambulatory shocks. Do not infer comparative efficacy, comfort, adherence or alarm superiority between brands.

SAFETY

Wear continuously except when the model-specific instructions require removal (such as bathing); keep equipment dry and out of MRI/flammable or oxygen-enriched environments. Skin irritation, alarm burden, inappropriate shock, missed therapy while unworn, falls and bystander shock are material harms to discuss.

Shared consent script

Five questions before “yes”

1 · What problem? Congestion warning, electrical risk, dyssynchrony, functional limitation or central apnea?

2 · What outcome? Death, hospitalization, symptoms, exercise function, sleep events-or a research physiologic endpoint?

3 · How certain? FDA indication plus randomized outcomes, symptom-focused trial, single-arm performance goal or early feasibility only?

4 · What burden? Access, incision, carotid/venous/PA anatomy, leads, antithrombotics, charging/transmission, programming, labs and clinic capacity.

5 · What alternative? More GDMT, CRT/ICD, valve/rhythm care, advanced-HF therapy, sleep therapy, trial participation or no implant.

STOP

Never stack devices because symptoms persist without re-phenotyping. Persistent NYHA III may represent congestion, low output, dyssynchrony, ischemia, valve disease, pulmonary disease, anemia, sleep disorder, frailty-or all of them.

Barostim selection and exclusion screen

Use the current IFU-not trial memory

Proceed when	Pause / do not implant until resolved	Reason
True persistent NYHA II-with-recent-III or III limitation despite maximally tolerated contemporary GDMT.	A correctable valve/ischemic/rhythm cause or a guideline CRT phenotype has not been addressed.	BAT evidence is adjunctive; it should not displace therapies with stronger outcome evidence.
LVEF and NT-proBNP meet the current label; goals center on symptoms/function.	Expectations are mortality prevention or proven HF-hospitalization reduction.	The long-term randomized primary morbidity/mortality endpoint was neutral.
Carotid anatomy is suitable after required imaging and surgical review.	Current IFU contraindication such as >50% carotid disease/ulcerative plaque, unsuitable high bilateral anatomy, baroreflex failure/autonomic neuropathy, uncontrolled symptomatic bradyarrhythmia or relevant material allergy.	Implant and stimulation risk are anatomy- and physiology-dependent.
Patient can attend titration and surveillance; infection risk is acceptable.	Advanced-HF trajectory, active infection, very limited non-HF survival or inability to complete follow-up.	Do not add procedural burden without a realistic symptom/function horizon.

CCM practical referral screen

Narrow lane · deliberate goals

Positive fit	Persistent NYHA III limitation; LVEF 25%-45%; contemporary GDMT; not receiving CRT; patient seeks better function/QoL and accepts an additional transvenous system.
Recheck first	CRT/physiologic pacing indication, reversible cause, uncontrolled arrhythmia, high PVC burden that may limit therapy delivery, active infection or advanced-HF instability. Current formal contraindications: mechanical tricuspid valve or inability to obtain venous access.
Consent words	“CCM may improve exercise capacity and quality of life; survival benefit is not established.” Risks include arrhythmia, venous/pneumothorax/perforation or pericardial injury, infection, tricuspid/lead/device complications and interaction with other CIEDs.
Follow	Symptoms/KCCQ, 6MWD, volume, rhythm, percent therapy delivery, charging behavior, lead/device parameters, GDMT and advanced-HF triggers.

DO NOT CONFUSE

Barostim, CCM and CRT act through different mechanisms and carry different evidence. “Device for HFrEF” is not a therapeutic class.

FDA
Lane

Current U.S. indicationWireless PA pressure/heart-rate monitoring in NYHA II or III HF with HF hospitalization in the prior year and/or elevated natriuretic peptides.

Intended outcomeHemodynamic data guide HF management to control PA pressure and reduce HF hospitalizations-across LVEF phenotypes.

Formal contraindicationInability to take dual antiplatelet or anticoagulant therapy for 1 month after implant.

Evidence without marketing compression

Randomized trials · open-label management

550CHAMPION participants; NYHA III + prior HF admission

0.706-month HF hospitalization HR in CHAMPION

0.88GUIDE-HF overall primary HR; P=.16

+7.1312-month KCCQ between-group points in MONITOR-HF

Trial	What it established	Interpretation guardrail
CHAMPION	PA pressure-guided management reduced HF hospitalization at 6 months and over longer follow-up.	All participants were implanted; the randomized intervention was access to pressure data and resulting care.
GUIDE-HF	Overall primary composite was neutral (HR 0.88; P=.16). Prespecified pre-COVID analysis suggested benefit (HR 0.81; P=.049).	Lead with the overall neutral trial; pandemic-era interaction analyses are supportive, not a substitute primary result.
MONITOR-HF	In 348 NYHA III patients with prior admission, KCCQ improved and HF hospitalizations fell; device/system complication-free 97.7%.	Dutch open-label RCT; supports effectiveness in a contemporary system, not a proven mortality effect.

FDA
Lane

Current U.S. indicationMeasure, record and transmit PAP in NYHA III HF patients at home on diuretics and GDMT who have been stable on GDMT for 30 days.

PurposeDevice output aids clinician assessment and management with the goal of reducing HF hospitalizations.

Limits ≠ trial enrollmentNot for emergency use or real-time monitoring. The FDA indication does not restate the trial’s prior-admission/elevated-NP entry criteria.

CardioMEMS and Cordella are not copy-paste labels

Check current IFU, payer and program requirements

Feature	CardioMEMS	Cordella
U.S. HF population	NYHA II-III; prior-year HF hospitalization and/or elevated NP.	NYHA III; at home on diuretic + GDMT; stable on GDMT 30 days.
Measurement posture	Supine PAP/heart rate using home electronics.	Seated PAP plus connected BP, HR, weight, SpO₂ and symptom platform.
1-month antithrombotic screen	Contraindicated if unable to take dual antiplatelet or anticoagulant therapy for 1 month after implant.	Likewise screen ability to receive the required 1-month dual antiplatelet or anticoagulant therapy.
Evidence frame	Multiple randomized management trials; overall GUIDE-HF result neutral, with supportive pre-COVID and MONITOR-HF findings.	PROACTIVE-HF became open-label single-arm with a historical performance goal after FDA input.
What not to claim	No established mortality benefit; a sensor without a responsive team is not therapy.	Performance-goal superiority is not randomized head-to-head superiority over usual care or CardioMEMS.

PROACTIVE-HF: read the design before the number

456 implanted · single arm

Design

Prospective, multicenter, open-label, single-arm after the original randomized design changed with FDA input.

Eligible trial patients had NYHA III symptoms plus recent HF hospitalization and/or elevated NP.

6 months

HF hospitalization/all-cause death rate 0.15 events/patient vs prespecified historical performance goal 0.43.

Device/system complication-free 99.2%; sensor-failure-free 99.8%. Hemoptysis occurred in 3%, including two intubations.

2 years

Low reported event rates and sustained health-status improvement were observed across LVEF strata.

Still single-arm: association and durability do not establish randomized causal benefit.

Candidate and implant checklist

Referral is a shared HF-implant decision

FIT

Best operational phenotype

Not LVEF-limited

Recurrent or difficult-to-detect congestion despite skilled longitudinal HF care.
Patient/caregiver can transmit reliably; team can review and act consistently.
Renal function, BP and medication flexibility permit pressure-responsive therapy.
PA anatomy and right-heart catheterization are suitable; implant risk is acceptable.

PAUSE

Low-value or unsafe setup

Fix the system first

No clinic ownership, alert protocol or medication-response pathway.
Nonadherence that cannot be made workable; no realistic transmission plan.
Advanced HF trajectory where the sensor would delay transplant/LVAD/palliative decisions.
Cannot receive required post-implant antithrombotic therapy, active infection or unsuitable anatomy.

ENDPOINT

The supported benefit is fewer HF hospitalizations when pressure-guided management is actionable; survival benefit is not established. Do not interpret a single pressure without trend, waveform quality, renal function, perfusion and the bedside phenotype.
Implant risks: PA injury/perforation or hemoptysis, access bleeding, arrhythmia, thrombus/embolus and infection.

One-page referral packet

Reduce repeat visits and inappropriate implants

CLINICAL

Phenotype and trajectory

Current, dated

NYHA class, HF admissions/urgent IV therapy, KCCQ or MLWHFQ, 6MWD/CPET when available.
Etiology, current LVEF/RV/valves, QRS morphology/duration, rhythm and device history.
BP, orthostasis, volume/perfusion profile, creatinine/eGFR, K, Na, hemoglobin/iron and NP.
Advanced-HF red flags and competing noncardiac prognosis.

TREATMENT

What has been tried

Dose + barrier

Four-pillar GDMT doses/tolerance; diuretic regimen; adherence and access barriers.
Revascularization, valve, AF/PVC, CRT/ICD and sleep evaluation status.
Desired endpoint stated in the patient’s words.
Ability to transmit/charge, attend titration, obtain labs and manage an additional implant.

Operational protocol

Trend → verify → act → close loop

1 · Own it Name the reviewer, review latency, backup, permitted actions, and escalation route; then establish the euvolemic target range and baseline BP, kidney/electrolyte state, diuretic dose, and confounders.

2 · Daily acquisition Teach a repeatable posture/time; monitor transmission reliability and waveform/data quality—not only numeric alerts.

3 · Verify the signal Review multi-day trend, symptoms, BP, weight, perfusion, creatinine/electrolytes, adherence, NSAIDs, diet, infection and arrhythmia.

4 · Act early Adjust diuretic/vasodilator/GDMT or evaluate another cause using a written protocol; avoid reflex diuresis for every isolated rise.

5 · Close the loop Communicate the change, define labs and reassessment, document response and escalate when pressure remains high or renal/perfusion limits appear.

TRAP

A sensor becomes an intervention only through an owned care loop. Audit alert volume, true/missed events, workload, harm, and outcomes. Pressure is not volume: PAP can move with vascular tone, rhythm, and ventricular interaction; discordant pressure, weight, and symptoms demand reassessment—not automatic dose escalation.
Implant risks: PA injury/perforation or hemoptysis, access bleeding, arrhythmia, thrombus/embolus and infection.

The one-minute device consult

A reproducible handoff

Phenotype “This patient has ___ HF, LVEF ___, NYHA ___, QRS/rhythm ___, with unresolved ___ despite ___.”

Trajectory “In 12 months: ___ admissions/urgent IV visits; BP ___; eGFR ___; NP ___; KCCQ/6MWD/peak VO₂ ___.”

Device lane “We are considering ___ for the endpoint ___; its status is FDA-HF / other-indication / investigational.”

Alternatives checked “GDMT, ICD/CRT, ischemia/valve/rhythm/sleep and advanced-HF options were addressed as follows: ___.”

Operational fit “Anatomy, antithrombotics, infection, other CIEDs, transmission/charging, follow-up and patient goal are ___.”

Decision boundary “We will judge success by ___ at ___; stop/escalate for ___; the fallback/advanced-HF plan is ___.”

Barostim
Label

WhoNYHA III or II with a recent history of III, symptomatic despite GDMT (medications and devices), LVEF ≤35% and NT-proBNP <1600 pg/mL.

What the label supportsImprovement in functional status, 6-minute hall walk and quality of life.

CRT nuanceCurrent labeling no longer states the older explicit CRT exclusion; the pivotal evidence base excluded patients with a Class I CRT indication.

What is implanted

Extravascular · vascular surgery + HF program

Pulse generator + carotid sinus lead

HF indication

The electrode is secured to one carotid sinus after intraoperative mapping. Electrical activation of the baroreflex is intended to reduce sympathetic and augment parasympathetic signaling.

It does not pace, resynchronize or defibrillate. Existing CIED function and interactions must be checked.

Programmed, titrated therapy

Longitudinal

Stimulation is titrated over follow-up for tolerability and effect. Neck sensation, cough, voice change, hypotension/bradycardia symptoms and device/lead status require structured review.

Assess benefit with NYHA class, KCCQ/MLWHFQ, 6MWD and patient goals-not a single BNP value.

BeAT-HF: durable symptoms, neutral hard endpoint

Open-label · nonimplanted control

323patients in long-term primary analysis

3.6 ymedian follow-up per patient

0.94CV death/HF morbidity rate ratio; P=.82

97%system/procedure-related major neurological or cardiovascular event-free

What to say: BAT produced durable improvements in quality of life, 6MWD and NYHA class, while the randomized primary CV mortality/HF morbidity endpoint did not differ (rate ratio 0.94, 95% CI 0.57-1.57). A hierarchical win ratio favored BAT, but it does not overturn the neutral primary endpoint.

CONSENT

“This is a symptom/function device.” Discuss uncertain hard-outcome benefit, alternatives, continued GDMT and advanced-HF surveillance. Implant risks include stroke/TIA or embolization, carotid/major bleeding injury, cranial/cervical nerve injury, infection/hematoma, device migration/erosion, reoperation and death, plus stimulation symptoms.

CCM
Label

WhoNYHA III despite GDMT, not receiving CRT, with LVEF 25%-45%.

What it improves6-minute hall walk, quality of life and functional status.

Label updateThe 2024 indication removed the former normal-sinus-rhythm language. Do not deny referral by quoting the original 2019 label.

Mechanism and hardware

Nonexcitatory signal during refractory period

Two RV septal leads + generator

HF indication

CCM delivers high-voltage, nonexcitatory signals during the absolute refractory period. It aims to modify calcium handling and contractile biology without initiating a beat.

It is not pacing, CRT, antitachycardia therapy or defibrillation.

EP integration is mandatory

CIED check

Review venous access, tricuspid anatomy, lead burden, rhythm, PVC burden and existing pacemaker/ICD. Test sensing and device-device interaction at implant and follow-up.

Programming, charging and adherence determine delivered therapy.

FIX-HF-5C: efficacy signal and limits

160 patients · unblinded · 24 weeks

Trial phenotype	Result	What it does not establish
NYHA III/IV, QRS <130 ms, LVEF 25%-45%; 74 CCM vs 86 medical therapy.	Bayesian between-group peak VO₂ difference +0.84 mL/kg/min; MLWHFQ, NYHA class and 6MWD favored CCM.	Small, short, unblinded symptom/function trial; not powered to prove mortality or durable HF-hospitalization benefit.
Seven device-related events were reported; composite CV death/HF hospitalization numerically favored CCM.	Findings supported the functional indication and warrant individualized benefit-risk discussion.	A nominal secondary hard-outcome P value is not a definitive outcome claim.

Do not turn a trial criterion into a label: FIX-HF-5C enrolled QRS <130 ms; the current indication says “not receiving CRT” and does not specify a QRS cutoff.

Scope
Guardrail

FDA indicationImplantable phrenic nerve stimulation for moderate-to-severe central sleep apnea in adults-not an HF indication.

Diagnose the breathing disorderUse polysomnography to distinguish central from obstructive events and optimize HF, opioids, altitude and other contributors.

Promise the measured endpointExpect fewer central events, better oxygenation/sleep metrics and selected QoL benefit. HF mortality/hospitalization benefit is not established.

Where it fits

Sleep medicine + EP + HF

Transvenous neurostimulation

CSA indication

A generator, stimulation lead adjacent to a phrenic nerve and sensing lead coordinate diaphragmatic breaths during sleep. Therapy is activated and titrated over follow-up.

It does not splint an obstructed pharynx; do not use it as an OSA device.

Pivotal randomized phase

Sleep endpoint

All patients were implanted; active therapy vs deferred activation increased the proportion achieving ≥50% AHI reduction at 6 months, with concordant sleep/QoL improvements.

The trial was not designed to prove HF outcome benefit.

Safety and modern labeling

Avoid stale MRI advice

Issue	Current practical point	Action
Infection	Active infection is a contraindication.	Defer implant; manage infection and reassess candidacy.
MRI	FDA-approved 2025 labeling is MR Conditional under specified conditions; the device automatically enters therapy-disabled magnet mode during exposure.	Use the exact model-specific MRI manual; do not repeat the obsolete blanket “MRI contraindicated” statement.
Other CIEDs	Potential sensing/stimulation interactions and cumulative lead/venous burden.	Coordinate implant side, test interaction, and interrogate all systems.
Implant / revision	Lead dislodgement or revision, venous/access complications, therapy discomfort and interaction with other implanted systems.	Consent explicitly; coordinate implant planning and longitudinal device surveillance.
Diathermy	Therapeutic diathermy is prohibited with an implanted system.	Flag the implant across procedural records; use model-specific safety instructions.
Phenotype	Mixed/obstructive apnea, unstable HF, opioid effect and hypoventilation require different or additional treatment.	Sleep specialist reviews PSG and causal contributors before referral.

Do not conflate phrenic stimulation with ASV

Different device · different evidence

remedē

Implantable stimulation

CSA indication

Restores diaphragmatic effort during central events.
Evidence centers on AHI, oxygenation, sleep architecture and QoL.
Do not market exploratory HF-subgroup analyses as definitive HF outcomes.

ASV

Mask-based servo ventilation

Device-specific evidence

SERVE-HF in symptomatic HFrEF with LVEF ≤45% and predominant CSA found increased all-cause and CV mortality despite AHI control.
ADVENT-HF used a different peak-flow-triggered algorithm: no CV benefit, but no ASV safety signal.
2025 AASM guidance conditionally permits ASV for HF-related CSA with low-certainty evidence; in HFrEF use shared decision-making, experienced-center monitoring and symptom/QoL goals.

REFERRAL

Moderate-to-severe predominant CSA + persistent burden after cause-directed care → multidisciplinary evaluation. The endpoint is CSA control and sleep-related benefit, not treatment of low LVEF itself.

Evidence ladder: place each claim on the right rung

As of July 2026

Rung	Examples in this guide	Permissible conclusion
FDA HF indication	CardioMEMS, Cordella, Barostim, CCM; ICD/CRT through guideline/device indications.	Use only for the current labeled/guideline phenotype and the supported endpoint.
Other indication	remedē for adult moderate-to-severe CSA.	Treat the indicated condition; do not convert it into an HF morbidity/mortality claim.
Neutral randomized	BeAT-HF hard primary endpoint; GUIDE-HF overall primary; REDUCE LAP-HF II; RELIEVE-HF overall; REBALANCE-HF.	Lead with the neutral result; label supportive secondary/subgroup findings as exploratory or symptom-specific.
Randomized pilot	FASTR/Reprieve.	Proof of concept; pivotal confirmation required.
Single-arm feasibility	Aortix, preCARDIA, Doraya, eLym/DELTA-HF design.	Technical/physiologic signal only; no causal clinical efficacy.
Concept / preclinical	Remaining DRI²P²S neural, lymphatic, pusher, puller and selective platforms.	Biologic plausibility and trial rationale-not patient care.

Research
Only

Interatrial shuntsLower LA pressure by creating left-to-right flow, but add right-sided volume and may create device/shunt complications.

Splanchnic ablationTargets venous capacitance and stressed-volume redistribution; randomized sham-controlled HFpEF evidence was neutral.

No routine implant laneResponder hypotheses, open-label improvement and physiologic plausibility do not establish standard care.

Interatrial shunts: what the randomized trials found

Sham/placebo controls matter

Trial	Population / design	Primary message	Safety / subgroup caution
REDUCE LAP-HF II	621 HFpEF/HFmrEF patients; randomized, blinded, sham-controlled.	No overall efficacy at 2 years: win ratio 1.01 (95% CI 0.82-1.24).	Major adverse cardiac events 6.9% vs 2.7%; more ≥30% RV-volume increase. “Responder” signal remains hypothesis under study.
RELIEVE-HF	508 symptomatic HF patients across LVEF; randomized to shunt vs placebo procedure.	Overall hierarchical effectiveness endpoint neutral: win ratio 0.86; P=.20.	Opposing exploratory signals by LVEF-possible fewer events in HFrEF and more events when LVEF >40%-must not become routine selection rules.

Greater splanchnic nerve ablation

Redistribution hypothesis · neutral phase 2

Concept

Increase splanchnic venous capacitance to shift blood away from the central compartment during exertion.

Mechanistically related to the DRI²P²S “dilator” idea; not the same as carotid BAT.

Trial

REBALANCE-HF randomized 90 HFpEF patients with invasive exercise PCWP elevation to endovascular right GSN ablation vs sham.

Primary PCWP difference -0.03 mmHg (P=.95); no secondary clinical benefit at 12 months.

Meaning

Safe/feasible in this trial, but physiologic and clinical efficacy were not demonstrated in the broad enrolled phenotype.

Use only in research; do not substitute an uncontrolled responder analysis.

How to discuss a trial

Precision without therapeutic misconception

Say	“This device is designed to lower filling pressure. Whether that improves how people feel or prevents HF events remains uncertain.”
Ask	What is randomized? Is there a sham? What is the primary endpoint? How are RV function, shunt direction/patency, stroke, device events and HF events adjudicated?
Avoid	Calling a post hoc subgroup “the indicated responder,” equating procedural success with efficacy, or quoting open-label before/after KCCQ change as causal.
Protect	Continue proven HF therapy, preserve advanced-HF options, and define an exit/closure plan if harm or futility appears.

BOTTOM LINE

No interatrial shunt or splanchnic-ablation strategy is routine HF care in this guide. Refer only through a protocol with explicit equipoise and safety surveillance.

Selective intrarenal therapy

Local drug delivery is not an HF outcome claim

Bifurcated renal-artery infusion

Not routine ADHF

Benephit-type catheters deliver agents such as fenoldopam directly to both renal arteries, aiming for high local exposure with lower systemic effect.

Historical 510(k) clearances for renal-artery infusion and post-market registry use do not establish an ADHF treatment indication or benefit.

Evidence boundary

Historical registry

Be-RITe! reported high cannulation success and observational safety in cardiovascular procedures at risk for contrast nephropathy.

It was not a randomized ADHF outcomes trial. Arterial access, embolization, dissection and drug effects remain relevant.

First
Principles

Congestion is heterogeneousTotal fluid excess, venous redistribution, renal venous hypertension, low perfusion and interstitial/lymphatic failure may coexist.

Treat acute HF firstIdentify trigger; define wet/dry and warm/cold; use evidence-based diuresis/vasodilation/support; reassess response and perfusion.

Shock exits this mapHypoperfusion, rising lactate, altered mentation or escalating vasoactive support requires shock-team/MCS logic-not experimental decongestion by acronym.

Seven mechanistic classes

Original 2020 concept · rapidly evolving examples

Class	Physiologic target	Representative concepts	Core risk / evidence gap
D · Dilators	Venous capacitance / stressed-volume redistribution.	Splanchnic nerve modulation.	Hypotension, orthostasis; randomized HFpEF efficacy neutral.
R · Removers	Sodium/water extraction or controlled diuresis.	Ultrafiltration, Reprieve, peritoneal/other fluid-removal platforms.	Access, electrolyte/renal injury; rate and refill mismatch.
I¹ · Inotropes	Autonomic stimulation to augment contractility.	Cardiac nerve plexus / autonomic stimulation systems.	Arrhythmia, myocardial demand; early feasibility.
I² · Interstitials	Thoracic-duct/lymphatic decompression.	WhiteSwell eLym.	Vascular/lymphatic injury; no definitive efficacy trial.
P¹ · Pushers	Raise renal arterial pressure / flow.	Aortix, Reitan-type and other descending-aortic pumps.	Large-bore access, hemolysis/vascular risk; small single-arm studies.
P² · Pullers	Lower SVC/CVP or renal venous pressure.	preCARDIA, Doraya, renal-vein unloading concepts.	Thrombosis/venous injury/hemodynamic instability; feasibility evidence.
S · Selective	High local renal drug exposure with less systemic effect.	Bifurcated renal-artery infusion catheters.	Arterial access/embolization; no established ADHF outcome benefit.

Map the dominant bottleneck

One patient may occupy several columns

Bedside/hemodynamic clue	Likely bottleneck	Research class that targets it	Conventional action before research
Rapid pulmonary congestion, modest weight gain, preserved pressure.	Central redistribution / reduced venous capacitance.	D	Vasodilator when appropriate, diuresis, treat trigger and BP.
Large total excess; poor natriuresis despite adequate perfusion.	Diuretic resistance / sodium retention.	R	Confirm dose delivery; intensify loop; sequential nephron blockade; monitor urine, electrolytes and renal function.
High CVP, ascites, renal dysfunction disproportionate to forward flow.	Renal venous congestion.	P²	Decongest, address RV/TR/PH, measure hemodynamics when uncertainty changes care.
Low output with high filling pressure but not shock.	Forward-flow/renal perfusion limitation.	I¹ / P¹	Define cause, valve/rhythm/ischemia; advanced-HF assessment; selective inotrope/support only for accepted indications.
Persistent edema/extravascular lung water despite intravascular response.	Interstitial refill / lymphatic bottleneck.	I²	Time, effective decongestion, albumin/oncotic and renal/hepatic context; avoid over-removal.

DRI²P²S is descriptive. It organizes investigational mechanisms; it does not certify regulatory status, prove a class effect, or tell a clinician which device to deploy.

CHECKPOINT

Before any device trial: confirm true congestion, define perfusion and BP, quantify diuretic response, identify a measurable target, and write rescue/withdrawal criteria.

Endpoint ladder for early devices

Do not skip a rung in interpretation

Technical: deploy, activate, remove; verify target engagement.

Necessary, never sufficient.

Physiology: CO, filling pressure, lymph flow, extravascular lung water, urine/sodium output.

Can improve while clinical outcomes do not.

Safety: access/bleeding, nerve/lymphatic injury, arrhythmia, hypotension, hemolysis, renal injury.

Needs denominator, adjudication and adequate follow-up.

Patient benefit: dyspnea, function, renal recovery, length of stay, worsening HF, readmission, survival.

Requires controlled trials and multiplicity-aware analysis.

CONSENT

“The device is being tested to see whether changing this physiology helps patients.” Do not say the mechanism is established treatment.

Dilators: increase venous capacitance

Mostly redistribution-not fluid removal

Splanchnic nerve modulation

Investigational

The splanchnic bed is a large venous reservoir. Modulation aims to reduce centrally stressed volume and filling pressures without first removing liters of fluid.

Chronic/endovascular HFpEF ablation in REBALANCE-HF was feasible but did not improve exercise PCWP or clinical outcomes versus sham.

Clinical guardrail

Neutral RCT

Do not infer benefit from an acute pressure fall, an open-label subgroup or the effectiveness of pharmacologic vasodilators.

Monitor BP, orthostasis, renal perfusion and symptoms; trial eligibility is not an indication.

Removers: three different propositions

Commercial availability ≠ routine HF benefit

Strategy	Mechanism / evidence	Practical place	Do not claim
Ultrafiltration	Direct isotonic plasma-water removal. UNLOAD: more 48-hour fluid/weight loss and fewer 90-day HF rehospitalizations vs IV diuretics.	Selected refractory congestion with a multidisciplinary rate/access plan; not reflex first-line treatment.	CARRESS-HF in congestion + worsening renal function found stepped pharmacologic care superior for renal preservation, equal weight loss and fewer serious adverse events.
Reprieve	Automated urine-output/sodium monitoring, individualized diuretic titration and matched sodium-chloride replacement. FASTR randomized 100 (96 treated): 24-hour natriuresis 1082 vs 423 mmol; IV-diuretic duration 46 vs 88 hours.	Investigational. Safety composite 31% vs 39% (P=.42); creatinine change did not differ. At 72 hours cumulative natriuresis remained higher, while net fluid loss, net sodium excretion and weight loss were similar.	Exploratory, multiplicity-unadjusted findings; faster early decongestion did not prove shorter stay or 90-day benefit. Pivotal confirmation required.
Other removal platforms	Peritoneal/ascites-directed or implantable pumping concepts address selected chronic fluid burdens.	Kidney/HF specialty or protocol-specific use according to the actual indication and evidence.	Do not generalize one device’s clearance, cirrhosis indication or pilot physiology to ADHF outcomes.

Rate is a treatment variable

Removal must match refill and perfusion

1 · Define target Total vs intravascular excess, BP/perfusion, CVP/PCWP when measured, renal trajectory and desired net balance.

2 · Track response Hourly urine/effluent, cumulative sodium/fluid, weight, symptoms, BP, creatinine/BUN, Na/K/Mg/CO₂ and hemoconcentration.

3 · Detect mismatch Hypotension, worsening perfusion, cramps, rising lactate or disproportionate renal injury suggest removal faster than vascular refill or a wrong phenotype.

4 · Stop/escalate Reduce/stop device therapy, reassess hemodynamics and cause; shock or progressive low output moves to an advanced-HF/shock pathway.

LANGUAGE

“Faster” is not automatically “safer,” and “more liters” is not an outcome. Report renal function, symptoms, adverse events and post-discharge HF events alongside decongestion.

Do Not
Conflate

I¹ is not CCMDRI²P²S “inotropes” are investigational autonomic/neural concepts intended for acute hemodynamic support-not FDA-labeled chronic CCM.

I² is not ultrafiltrationInterstitial devices aim to restore lymphatic drainage rather than directly pump plasma water from a venous circuit.

Both are earlyFeasibility and short-term physiology cannot establish survival, readmission or durable renal benefit.

I²

Interstitial / lymphatic decompression

Target the refill bottleneck

WhiteSwell eLym

Early feasibility

A transcatheter system creates a localized low-pressure zone near the thoracic-duct outflow to facilitate lymph drainage into the venous circulation.

The target is interstitial/extravascular fluid-not simply another venous ultrafiltration circuit.

DELTA-HF

55 enrolled · 40 treated

Of 71 screened, 55 enrolled in the procedure pathway; 15 were not treated (12 because venous anatomy was incompatible). All 40 treated had successful deployment, activation and removal. Weight fell 3.6 kg during therapy and 6.8 kg by discharge; renal function was stable; primary safety freedom was 95%.

One vascular complication caused mediastinal hematoma and death; another patient required low-dose inotrope for hypotension. Investigational; uncontrolled efficacy.

I¹

Device-based neural inotropy

Cardiac autonomic stimulation concepts

Concept	Intended physiology	Why it is attractive	What must be proven
Cardiac nerve-plexus modulation e.g., Cardionomic concepts	Stimulate selected autonomic pathways to augment LV contractility and output.	Potential titratable hemodynamic effect without continuous systemic catecholamine exposure.	Reproducible target capture, arrhythmia safety, myocardial demand, device interaction, organ perfusion and patient-centered outcomes.
Neural stimulation systems e.g., NeuroTronik concepts	Modulate cardiac autonomic tone to improve contractility/hemodynamics.	Mechanism-specific, potentially on/off and dose-programmable.	Sham-controlled efficacy, durable benefit and safe selection across ischemic substrate, AF and implanted CIEDs.

Status: investigational. Names in a review are examples of a mechanism class, not endorsements, approvals or evidence of a class effect.

P¹

Pushers: increase renal arterial pressure

Descending-aortic pumps

Platform	Mechanism	Best published signal	Status / caution
Aortix	Catheter-deployed intra-aortic entrainment pump above the renal arteries; intended to augment renal perfusion and unload the heart.	CRS pilot: 18 completed therapy; mean 4.6 days, net fluid loss 10.7 L; CVP/PCWP and creatinine fell from baseline.	Feasibility only: 37 total serious adverse events occurred in 11/21 patients; 18 were device-, procedure- or study-related (9 of those 18 in 2 patients). No randomized efficacy.
Other P¹ concepts	Reitan-type rotary pumps and other descending-aortic flow augmentation platforms.	Early device-specific feasibility and hemodynamic studies.	Do not transfer Aortix results across platforms or equate renal-pressure augmentation with renal recovery.

P²

Pullers: lower venous back-pressure

SVC, IVC or renal-vein targets

Platform	Target / evidence	Interpretation
preCARDIA	Intermittently regulates SVC flow to reduce cardiac preload. VENUS-HF early-feasibility: 60 treated; 98.3% free of device/procedure major events; among 52 with ~24 h therapy, RA pressure fell 23%, PCWP 18%, and urine output increased.	Single-arm HF evidence. K221294 clears selective IVC/SVC flow control for specified perioperative/emergency-hemorrhage applications and blood-pressure monitoring; FDA states HF-treatment safety/effectiveness is not established.
Doraya	Adjustable infrarenal IVC flow regulator. First-in-human n=9 showed lower CVP and improved diuresis. U.S. EFS n=16: 100% implant success, no device-related SAE, one procedure SAE; urine output rose 80→224 mL/h during dwell.	Still uncontrolled feasibility; no hard-outcome claim or HF authorization.
Renal-vein unloading	Transcatheter concepts directly lower renal venous pressure to improve the transrenal gradient.	Mechanistically attractive, early development; anatomy, thrombosis, hemolysis and renal-injury safety require device-specific evidence.

The transrenal gradient is not the whole kidney

Interpret physiology in context

PROMISE

Why pressure manipulation may help

Selected phenotype

Renal perfusion depends on both arterial inflow and venous/interstitial back-pressure.
High CVP can impair filtration even when systemic arterial pressure appears adequate.
A local device may change renal hemodynamics without full systemic inotrope/vasodilator exposure.

LIMIT

Why a pressure win may fail

Whole-patient biology

Intrinsic CKD, tubular injury, neurohormonal sodium retention and low filtration area do not disappear.
Arterial/venous access, anticoagulation, thrombosis, hemolysis and limb/organ ischemia may offset benefit.
Urine output can rise without durable kidney recovery or fewer HF events.

TRIAL LANE

No P¹ or P² platform in this atlas is routine U.S. ADHF care. Consider only in an appropriate clinical trial with hemodynamic phenotype, independent safety oversight and rescue criteria.

Non-negotiable safety statements

Put these in the note and consent

Label	Verify the current FDA indication, contraindications and model-specific IFU on the date of decision; device labels change.
Endpoint	State whether evidence supports survival, hospitalization, symptoms/function, sleep metrics, physiology or only feasibility.
Sequence	An adjunct device never replaces GDMT, indicated CRT/ICD, valve/rhythm care or timely advanced-HF referral.
Research	DRI²P²S is an investigational classification. Trial participation requires informed consent, protocol eligibility and independent safety monitoring.
Escalate	Shock, hypoperfusion, escalating inotrope/vasopressor need, refractory congestion with organ dysfunction or repeated admissions → advanced-HF/shock team now.

FINAL RULE

Do not let a compelling mechanism outrun the evidence. Match the device to the phenotype, quote the correct endpoint, build the follow-up system, and keep destination therapy in view.

Make the advice fit the patient. Avoid excessive salt while preserving nutrition. Routine aggressive sodium or fluid restriction has not demonstrated better major outcomes in stable ambulatory HF. Congestion, serum sodium, renal function, thirst and intake should guide an individualized plan.

What to recommend — and what the numbers mean

sodium ≠ salt

Clinical setting	Practical interpretation
Dietary sodium	The 2022 US guideline considers avoiding excessive sodium reasonable to reduce congestive symptoms in stage C HF (2a, C-LD). ESC self-care advice avoids >5 g salt/day ≈2 g sodium/day. These are pragmatic recommendations, not a trial-established survival threshold; <1.5 g sodium/day is not obligatory for every patient.
Stable / euvolemic HF	Reconsider an automatic 1.5 L/day fluid cap when it causes substantial thirst or treatment burden. FRESH-UP supports flexibility in selected stable outpatients; liberal advice did not mean forcing large fluid volumes. Maintain attention to congestion and changes in intake.
Severe HF / hyponatremia	ESC guidance permits considering 1.5–2 L/day in selected severe HF or hyponatremia. US guidance calls the benefit uncertain in advanced HF with hyponatremia (2b, C-LD). Establish the cause and volume state; reassess thirst, weight, electrolytes and kidney function. Stable-outpatient trials do not settle this setting.
Counseling details	State the units: 1 g sodium ≈2.5 g salt; 100 mmol sodium ≈2.3 g sodium. Review labels and major dietary sources; avoid making food unpalatable or compromising calories/protein. If limiting fluids, define what counts and revisit the plan during heat, vomiting or other losses.

Trial evidence and the limits of inference

neutral superiority ≠ equivalence

Study	Test and main result	Expert reading
SODIUM-HF 2022 · n=806	NYHA II–III outpatients: counseling toward <1500 mg sodium/day vs usual care. At 12 months, death / CV hospitalization / CV ED visit: 60/397 vs 70/409; HR 0.89 (0.63–1.26), P=.53.	Achieved medians 1658 vs 2073 mg/day: only 415 mg apart. KCCQ +3.38 points was a modest, unblinded secondary signal; walk distance was neutral. Tests a counseling strategy, not the effect of complete adherence or large salt excess.
FRESH-UP 2025 · n=504	Stable NYHA II–III outpatients: liberal fluid advice vs ≤1500 mL/day. At 3 months, KCCQ overall difference +2.17 points (−0.06 to 4.39), P=.056; primary endpoint not met.	Median intake 1764 vs 1480 mL/day. Less thirst with liberal advice was supportive after failed hierarchical primary testing. Few safety events cannot establish equivalence; recent decompensation, Na <130 mmol/L and eGFR <30 were excluded.
Aliti 2013 Acute HF · n=75	Combined 800 mg sodium + 800 mL fluid/day restriction during hospitalization did not improve weight loss or clinical congestion; thirst increased.	Small, short trial of a bundled extreme regimen. It cannot separate sodium from fluid effects or determine the optimal intake during acute decongestion.
OSPREY-AHF Acute HF · n=65	Blinded oral NaCl vs placebo during IV diuresis also tests less-restricted vs restricted sodium exposure. Joint weight–creatinine endpoint: P=.33.	Challenges the assumed benefit of routine inpatient restriction. Actual dietary intake was uncertain; the small superiority trial did not establish noninferiority. Details on P.46 and in the evidence critique.

READOUT

Reduce excess, tailor limits, and reassess. OSPREY links the restriction question to oral supplementation during active diuresis. Keep inpatient and outpatient evidence distinct; continue to P.46 for route, dose and interpretation.

Route, setting and endpoint matter. Oral salt during IV diuresis, a single outpatient hypertonic-saline infusion, and repeated inpatient rescue are different interventions. OSPREY-AHF and SALT-HF did not demonstrate better primary outcomes; neither establishes a routine salt-supplementation strategy.

Contemporary randomized evidence

studied regimens · not an order set

Trial / setting	What was tested / found	How to read it
OSPREY-AHF 2023 · inpatient	Double-blind oral NaCl 2 g three times daily versus placebo during IV furosemide; 70 consented, 65 analyzed. At 96 h, the joint weight–creatinine endpoint was neutral (P=.33). Weight loss: 4.0 vs 4.6 kg (P=.57); creatinine rise: 0.04 vs 0.15 mg/dL (P=.30).	Also a blinded test of less-restricted vs restricted oral sodium exposure, informing routine inpatient restriction. The scheduled dose added up to ≈2.36 g sodium/day; prescribed diet does not establish actual intake. Secondary sodium/BUN signals do not prove renal protection. Small and underpowered, with no noninferiority margin: a neutral result leaves clinically meaningful benefit or harm possible. See the expert critique for diet-order clarification.
SALT-HF 2024 · ambulatory	167 patients with worsening HF; one IV furosemide infusion with hypertonic versus normal saline. Three-hour urine volume: 1099 vs 1103 mL; P=.963. Most congestion measures and 30-day clinical events were neutral.	Does not support routine addition to this outpatient strategy. A single infusion and short follow-up cannot settle repeated rescue treatment in refractory hospitalized patients. An isolated secondary weight signal does not reverse the neutral primary endpoint.

Why the wider literature needs scrutiny

benefit signals ≠ established outcomes

Evidence	Contribution and limitation
Licata 2003 / SMAC-HF 2011	Older randomized studies reported substantial benefits but assigned different sodium diets during and after admission and different loop delivery (infusion vs bolus). Long-term results cannot isolate the effect of inpatient hypertonic saline.
Griffin 2020 · US rescue series	40 patients, 50 admissions, 58 treatment episodes. Improved before/after diuretic and biochemical trajectories support feasibility. Concurrent therapies, regression to the mean and repeated observations limit causal inference; there was no randomized comparator.
2023 meta-analysis / trial-sequential analysis	Ten trials suggested better decongestion surrogates, with heterogeneity and low-to-moderate certainty. Mortality and HF readmission were not pooled. Sequential analysis addresses random error, not bundled interventions or bias. The search preceded SALT-HF.
SALT-HF chloride analysis · 2025	Post hoc: 148 patients with chloride data, only 16 with chloride <96 mmol/L. A continuous interaction suggested a urinary sodium-concentration signal, but the low-chloride subgroup estimate crossed zero; urine-volume benefit did not vary correspondingly. Hypothesis-generating, not a treatment threshold.

READOUT

No established survival benefit or routine hypertonic-saline protocol follows from this evidence. Keep the possible rescue role separate from proven decongestion strategies (P.19). Open the chapter’s evidence base for compact study breakdowns, funding, source papers and expert critique.

Contents

Cross-guide map — when to open a focused guide

First moves for any new HF patient

Heart failure — 2026 universal definition

ACC/AHA stages — progression & trajectories

EF phenotypes — 2026 classification

NYHA functional class — track it every visit

Prognosis & the advanced-HF triggers

The new-HF work-up — everyone gets these

Natriuretic peptides — read them right

When to go further

Echo — what to extract beyond the EF

Right-heart catheterization — when the numbers matter

Universal classification of HF causes & where to take them

Reversible causes worth chasing in every new CM

Red flags → the targeted test

The four pillars — all are Class 1 for HFrEF outcomes

Sequencing & titration — get all four on, fast

Then layer on — by indication

Target doses — titrate toward these

Monitoring — after start & each up-titration

GDMT for LVEF 41–49% — what's proven

Practical approach

Why these classes — the mildly reduced EF evidence

Don't forget — secondary prevention & devices

H2FPEF score — quick bedside probability

HFA-PEFF score — domains, cutoffs & interpretation

Invasive hemodynamics — rest & provocation

Biomarkers & the dyspnea differential

Grading diastolic function (when EF is normal)

What to start

Treat the comorbidities — that is most of HFpEF

Phenotype the HFpEF — the driver shapes therapy

Practical — start & monitor

Diuretics & decongestion — the HFpEF tightrope

The HFpEF mimics

Constriction vs restriction — one is surgically curable

Restrictive cardiomyopathy — the causes to chase

Thick walls — read the LVH before calling it "hypertensive"

Evaluating for CAD in new HF

Hibernation, stunning & scar — the viability concept

Viability testing — how each modality calls it

Putting it together — act on the work-up

Recognize the ischemic-CM variants & mimics

The defining trials

What the 2021 revascularization guideline says

CABG vs PCI — the Heart-Team call

Read the trials carefully — their boundaries

Peri-procedural & timing in HF

The ischemic-HF secondary-prevention bundle

Angina in HFrEF — what's safe

When to escalate or refer

Risk-factor targets

Longitudinal care — don't set & forget

Primary vs secondary MR — the fix differs

TEER for secondary MR — COAPT vs MITRA-FR

The COAPT phenotype — who to refer

Grading severe MR — the thresholds differ

Primary MR — when to operate & how

Defining & staging severe AS

Low-flow low-gradient AS — the HF trap

TAVR vs SAVR — the evidence & indications

Asymptomatic severe AS — when not to wait

After TAVR — what to watch

Tricuspid regurgitation — the "forgotten valve"

LAA closure (Watchman) — AF stroke prevention

Shunts — ASD / PFO & the atrial-shunt concept

TR severity & the RV consequence

Structural referral quick-map

Primary-prevention ICD — who qualifies

The evidence

Secondary prevention & device choice

Before you implant — device choice & counseling

When an ICD is not indicated

CRT indications — QRS & morphology drive the class

The evidence

Maximize response & conduction-system pacing

Pacing-induced CM & the upgrade question

Device choice & follow-up

AF in HF — rate, rhythm & anticoagulation