Korean Society of Heart Failure Guidelines for the Palliative Care and Hospice for Heart Failure Patients

Abstract

Heart failure (HF) is a leading cause of mortality and morbidity in South Korea and imposes substantial physical, emotional, and economic burdens on patients, families, and society. Despite the high symptom burden and complex care needs of patients with HF, palliative care and hospice services remain underutilized owing to cultural, institutional, and knowledge-related barriers. This guideline from the Korean Society of Heart Failure (KSHF) provides evidence-based recommendations to integrate palliative and hospice care across the HF disease trajectory, with the aim of improving quality of life, alleviating suffering, and supporting holistic care for patients and their caregivers. By clarifying the role and scope of palliative care in HF, and suggesting models for collaborative care between cardiology and palliative teams, these recommendations seek to bridge the gap between HF and palliative care services in South Korea and to promote goal-concordant, patient-centered decision-making.

Graphical Abstract

INTRODUCTION

Heart failure (HF) is one of the leading causes of death in South Korea, and its prevalence continues to rise. In both men and women, mortality and hospitalization rates related to HF are increasing, thereby imposing a greater economic burden on the healthcare system and society.¹,²⁾ Regardless of its etiology, HF is a chronic disease that requires lifelong management, and patients may experience recurrent episodes of symptom exacerbation such as dyspnea, pain, and fatigue throughout the disease course, similar to patients with cancer.³⁾ According to the World Health Organization, palliative care is defined as “an approach that improves the quality of life (QoL) of patients and their families facing problems associated with life-threatening illness, through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual.” In South Korea, palliative care has been implemented for various diseases, including cancer, chronic obstructive pulmonary disease (COPD), acquired immune deficiency syndrome, and chronic liver cirrhosis.

However, compared with other countries, there are still diseases in South Korea that are not yet widely recognized as indications for palliative care, and HF is one of them. In the United States, patients with HF account for approximately 8% of all palliative care recipients, and those diagnosed with “HF” based on the International Classification of Diseases-10 revision code represent the fifth most common diagnosis among patients enrolled in palliative care services.⁴⁾ In addition, the Centers for Medicare and Medicaid Services provide eligibility criteria for the use of palliative care in patients with chronic cardiac disease. The American Heart Association (AHA) and the European Society of Cardiology (ESC) also address palliative care and hospice for HF patients in their respective HF guidelines.⁵⁾ While palliative and hospice care are actively utilized for HF patients in Western countries, they are not yet fully established in South Korea owing to various factors, including systemic and regulatory barriers.⁶⁾

The notion that palliative care should be initiated only at the terminal stage of disease is a common misconception. In patients with HF, the role of palliative and supportive care ideally begins at the time of diagnosis, gradually expands as the disease progresses, and extends through the bereavement period.⁷⁾ As in patients with advanced cancer, those with advanced or end-stage HF experience similar symptoms, including dyspnea, fatigue, pain, depression, and anxiety. With disease progression, they may become less responsive to conventional treatments and experience an increasing symptom burden. Therefore, as in cancer care, it is desirable to integrate palliative care and hospice throughout the entire continuum of care for HF. Nevertheless, there are several barriers to palliative care in HF that differ from those in cancer. In general, patients with HF and their caregivers often do not perceive HF as a life-threatening illness comparable to cancer and tend to overestimate the patient’s life expectancy. They may also believe that palliative care or hospice will hasten death. Furthermore, given the natural course of HF, prognostication is challenging, which makes clinicians reluctant to discuss prognosis, life expectancy, death, and end-of-life care. As a result, timely conversations about palliative care are often delayed or not initiated. This guideline aims to improve the QoL and relieve suffering among patients with HF and their families (caregivers) by delineating the role of palliative care and hospice across the entire HF care continuum and by providing practical, clinically applicable recommendations for use in real-world practice.

SCOPE AND TIMING OF PALLIATIVE CARE FOR PATIENTS WITH HF

Target population

In patients with HF, palliative care is not limited to a specific (terminal) stage but can be initiated from the time of diagnosis in the form of concurrent supportive care based on an assessment of symptom, functional, and emotional burden. However, because the clinical course of HF is characterized by repeated episodes of deterioration and recovery, it is difficult to predict the end-of-life phase and referral to palliative care is often delayed. Therefore, it is necessary to identify candidates for palliative care by comprehensively considering clinical course, symptom burden, recurrent hospitalizations, functional decline, and changes in treatment goals.

For the purposes of this guideline, advanced HF is defined in accordance with the 2018 position statement of the Heart Failure Association of the ESC⁸⁾—subsequently incorporated into the 2021 ESC HF Guidelines⁹⁾—as a condition characterized by the simultaneous presence of all four of the following criteria, despite optimization of guideline-directed medical therapy (GDMT):

1) Severe and persistent symptoms of HF, corresponding to New York Heart Association (NYHA) functional class III (advanced) or IV;

2) Severe cardiac dysfunction, defined by at least one of the following: left ventricular ejection fraction (LVEF) ≤30%, isolated right ventricular failure, severe uncorrectable valvular abnormalities or congenital heart disease, or restrictive/non-dilated cardiomyopathy with severely impaired filling;

3) Episodes of fluid retention and/or reduced cardiac output requiring hospitalization, intravenous diuretics or inotropes, or repeated outpatient interventions to prevent clinical deterioration;

4) Severe impairment of exercise tolerance, defined as inability to perform any significant exercise, 6-minute walk test distance <300 m, or peak oxygen consumption <12 mL/kg/min.

This definition is broadly consistent with American College of Cardiology (ACC)/AHA Sstage D HF as defined in the 2022 AHA/ACC/Heart Failure Society of America Guideline for the Management of Heart Failure—the most advanced and refractory stage of HF requiring specialized interventions.¹⁰⁾ It is important to recognize that NYHA class alone is insufficient to define advanced HF, as NYHA class III/IV encompasses a heterogeneous spectrum ranging from patients who may still respond to therapy optimization to those with truly refractory end-stage disease.

The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profiles provide a complementary framework for characterizing severity in patients already being evaluated for mechanical circulatory support (MCS) therapy.¹¹⁾ INTERMACS profiles 1–4 broadly correspond to patients who are hemodynamically unstable (profile 1: critical cardiogenic shock), inotrope-dependent (profile 2), clinically stable on inotropes or with frequent decompensations (profile 3), or ambulatory with severe symptoms at rest or with minimal exertion (profile 4). However, as INTERMACS criteria were originally developed for MCS device registries and do not cover patients' full clinical history or those ineligible for MCS, they should be used as a severity classifier within—not as a substitute for—the broader clinical definition of advanced HF outlined above.

When reviewing international criteria for defining the target population for palliative care in HF, the Centers for Medicare and Medicaid Services in the United States present, as one of the criteria for terminal status in cardiac disease requiring hospice care, the following: 1) the patient is receiving appropriate treatment for the underlying heart disease and is not a candidate for or has declined surgical interventions, and 2) the patient has NYHA class IV symptoms even at rest. In the AHA guidelines, hospice referral is suggested as a means to improve QoL in patients whose life expectancy is estimated to be less than 6 months.⁵⁾

In South Korea, a Delphi study was conducted among experts to determine appropriate timing for referral of HF patients to specialist palliative care and hospice services, and consensus was reached on several criteria related to comorbidities and complications, current HF therapies, healthcare utilization, HF severity, life expectancy, symptom burden, activities of daily living, and psychosocial needs.¹²⁾ By integrating these international guidelines and domestic expert consensus, this guideline defines its target population as patients with advanced HF according to the Korean Society of Heart Failure (KSHF) guideline, who correspond to INTERMACS profiles 1–4, show refractoriness to GDMT, and are not candidates for heart transplantation or left ventricular assist device therapy (Table 1).

Table 1. Practical definition of advanced HF.

Criterion	Definition	Practical examples	Note
1. Severe persistent symptoms	NYHA class III (advanced) or IV despite optimized GDMT	Dyspnea at rest or minimal exertion; fatigue limiting ADL	Must persist despite optimization
2. Severe cardiac dysfunction	LVEF ≤30%; or isolated RV failure; or severe uncorrectable valvular/congenital disease; or restrictive non-dilated CM	Echocardiographic evidence of severely impaired systolic or diastolic function	≥1 criterion required
3. Fluid retention/low output episodes	Recurrent hospitalization; IV diuretics/inotropes; outpatient IV therapy	≥2 HF hospitalizations within 3 months
4. Severe exercise intolerance	Unable to exercise; 6MWT <300 m; pVO2 <12 mL/kg/min	Unable to walk one block; severely limited ADL
INTERMACS 1–4 (complementary)	Profiles 1–4 in MCS candidates	Profile 1: cardiogenic shock; profile 4: ambulatory at rest	Supplementary severity classifier
Life expectancy		Clinician-estimated life expectancy ≤6 months

The definition was modified from consensus-based criteria for referral of patients with HF to specialist palliative care or hospice by KSHF.¹²⁾ All four advanced HF criteria must be present simultaneously.

HF = heart failure; NYHA = New York Heart Association; GDMT = guideline-directed medical therapy; ADL = activities of daily living; LVEF = left ventricular ejection fraction; RV = right ventricular; CM = cardiomyopathy; 6MWT = 6-minute walk test; pVO₂ = peak oxygen consumption; INTERMACS = Interagency Registry for Mechanically Assisted Circulatory Support; MCS = mechanical circulatory support.

Timing along the continuum of care

Palliative care and supportive care can be provided in parallel with disease-modifying therapy, and as the disease progresses, it is desirable that symptom control and goals of care are gradually expanded. Hospice represents a form of structured palliative care at the end-of-life stage, and the timing of transition should be discussed within a multidisciplinary team, considering the continuation of disease-directed treatment, the patient’s values and preferences, remaining prognosis, and available caregiving resources.

DEVELOPMENT PROCESS OF THE PALLIATIVE CARE GUIDELINE FOR PATIENTS WITH HF

Target users of the guideline

The target users of this guideline are clinicians managing patients with HF.

Composition and establishment of the guideline development committee

In March 2025, the Central Hospice Center in the National Cancer Center commissioned the KSHF to develop the draft guideline for “Palliative Care and Hospice for Patients with HF.” After discussion by the KSHF board of directors, the decision was made to proceed with guideline development led by the policy director, involving 15 society members and 2 external committee members from the Korean Society for Hospice and Palliative Care. Additionally, 1 guideline development methodology expert (Kim HJ), who serves on the Clinical Practice Guideline Committee of the Korean Academy of Medical Sciences, participated as an advisor for literature search, systematic literature review, and establishing guideline development methodology.

Selection of key questions

There were no existing guidelines providing an appropriate level of palliative care recommendations for patients with HF. However, the guideline development committee reviewed existing review articles and the clinical practice guideline for end-of-life care to select nine key questions. Clinical experts from related societies and methodology experts participated in the selection of key questions and review of recommendation statements. Key questions were formulated as sentences based on Population, Intervention/Index test, Comparator, Outcome elements, and their feasibility for development was evaluated before finalization (Table 2).

Table 2. Nine key questions and PICO elements.

1. Is palliative care beneficial for patients with end-stage HF?

P: Patients with end-stage HF; I: Palliative care; C: Usual care; O: Symptom relief, improvement in QoL

2. Is oxygen therapy beneficial for patients with end-stage HF who complain of dyspnea?

P: Patients with end-stage HF; I: Oxygen therapy; C: Usual care; O: Relief of dyspnea, improvement in QoL

3. Is diuretic administration beneficial for patients with end-stage HF who complain of congestive symptoms?

P: Patients with end-stage HF; I: Diuretic administration; C: Usual care; O: Relief of congestive symptoms, improvement in QoL

4. Is inotrope administration beneficial for patients with end-stage HF who complain of congestive symptoms?

P: Patients with end-stage HF; I: Inotrope administration; C: Usual care; O: Relief of congestive symptoms, improvement in QoL

5. Is opioid administration beneficial for patients with end-stage HF who complain of dyspnea or pain?

P: Patients with end-stage HF; I: Opioid administration; C: Usual care; O: Pain relief, relief of dyspnea, improvement in QoL

6. Is benzodiazepine administration beneficial for patients with end-stage HF who complain of dyspnea?

P: Patients with end-stage HF; I: Benzodiazepine administration; C: Usual care; O: Relief of dyspnea, improvement in QoL

7. Is pharmacological treatment beneficial for relieving depression and anxiety in patients with end-stage HF?

P: Patients with end-stage HF; I: Pharmacological treatment; C: Usual care; O: Symptom relief, improvement in QoL

8. Is non-pharmacological intervention beneficial for relieving depression and anxiety in patients with end-stage HF?

P: Patients with end-stage HF; I: Non-pharmacological intervention; C: Usual care; O: Symptom relief, improvement in QoL

9. Is deactivation of an implantable cardioverter defibrillator beneficial in patients with end-stage HF?

P: Patients with end-stage HF; I: Implantable cardioverter defibrillator deactivation; C: Usual care; O: Improvement in quality of dying, reduction in family members' post-bereavement depression

PICO = Population, Intervention/Index test, Comparator, Outcome; HF = heart failure; QoL = quality of life.

Development methods and literature selection

Since there were no existing clinical practice guidelines for palliative care in patients with HF, this guideline was developed de novo, and systematic reviews were conducted according to the methodology proposed by Cochrane.¹³⁾ For each key question, one working expert was assigned to select keywords, which were then refined and finalized through discussions with the guideline development expert. Searches were performed without restrictions on year, language, or publication status for all key questions, and results were aggregated using EndNote. Major domestic and international databases, including Ovid-MEDLINE, Ovid-Embase, Cochrane Library, and KoreaMed, were searched, with the final search strategies provided in Supplementary Figure 1. The final retrieved articles underwent a primary screening process independently reviewed by the guideline development chair and committee members, based on titles and abstracts. Subsequently, the assigned committee members for each key question conducted secondary screening, documenting exclusion reasons after full-text review.

Evidence synthesis and quality assessment

Evidence quality assessment consisted of two parts: quality evaluation of individual primary studies and overall assessment of synthesized evidence. For the final selected studies per key question, the assigned committee members performed quality assessments. Tools used included ROB 2.0 for randomized controlled trials, ROBINS-I for non-randomized studies, QUIPS for prognostic studies, AMSTAR for systematic reviews, and AXIS for cross-sectional studies. Results were summarized and recorded using a pre-defined Excel template.

Derivation of recommendations

Evidence levels were established by reviewing major international grading systems such as the Scottish Intercollegiate Guidelines Network and Grading of Recommendations, Assessment, Development and Evaluation (GRADE), as well as existing domestic clinical practice guidelines, and adapting them into a modified 4-tier system: high, moderate, low, and very low (Table 3).

Table 3. Definition of level of evidence.

Level of evidence	Definition
High	Results derived from randomized controlled trials or observational studies with comparators, with minimal methodological concerns and high consistency of results.
Moderate	Results derived from randomized controlled trials or observational studies with comparators, but with some risk of bias or potential inconsistency.
Low	Results from limited observational studies with or without comparators, with low reliability due to serious issues with bias, inconsistency, or imprecision.
Very low	Results from observational studies without comparators, expert opinions, or reviews, with very low reliability due to serious issues with bias, inconsistency, or imprecision.

It is important to note that recommendation strength is not solely determined by evidence certainty. Consistent with GRADE methodology, the guideline development committee considered the full constellation of factors outlined in Table 4—including benefit-harm balance, patient values and preferences, clinical applicability, and resource implications—when determining the final strength of each recommendation. In particular, for interventions supported by low or very low certainty evidence, strong or directional recommendations may still be appropriate when the benefit-harm balance is clearly favorable, when clinical consensus is robust, or when patient-centered values strongly support a particular course of action. Such discordance between evidence certainty and recommendation strength is not a methodological inconsistency but is explicitly recognized as appropriate within internationally established guideline frameworks, including the GRADE “strong recommendation, low certainty” (1C) scenario (Table 4).

Table 4. Definition of strength of recommendation.

Strength of recommendation	Definition
Strong recommendation	Benefits outweigh risks in most situations; strongly recommended even after considering patient values and resources.
Conditional recommendation	Benefit-risk balance may vary by clinical context or patient/societal values; selectively recommended through individualized decision-making.
Not recommended	Risks may outweigh benefits; advised against after considering clinical context or patient/societal values.
Inconclusive	Evidence is very insufficient or judgments on benefits/harms conflict; defer decision on intervention and await additional evidence.

RECOMMENDATIONS AND EXPLANATIONS FOR EACH KEY QUESTION

KQ 1. Is palliative care beneficial for patients with end-stage HF?

Compared with oncology, the integration of palliative care into HF management has been delayed, and a systematic evaluation of different delivery models—consultative, inpatient, and home-based/telehealth—and their supporting evidence is therefore necessary. Patients with advanced HF experience substantial physical and emotional burden due to refractory symptoms and complex treatment decisions; although palliative care consistently improves QoL and symptom control, survival benefits remain inconsistent across mostly small studies with imprecise estimates and potential publication bias, leading to an overall low certainty of evidence but a strong recommendation in favor of its use (Table 5).¹⁴,¹⁵⁾

Table 5. Palliative care in patients with advanced HF.

Categories	Details
Recommendation	Palliative care for patients with advanced HF is recommended.
Strength of recommendation	Strong recommendation
Level of evidence	Low
Key rationale beyond evidence	Consistent QoL benefit; no harms; expert consensus supports integration

HF = heart failure; QoL = quality of life.

Across multiple studies, palliative care interventions have been shown to reduce symptom burden and improve overall QoL in patients with HF. These benefits arise from not only the control of physical symptoms such as pain and dyspnea, but also from multidisciplinary approaches that clarify treatment goals, address psychosocial factors, facilitate advance care planning (ACP), and promote discussions regarding life-sustaining treatment and preferred place of death.¹⁴,¹⁶,¹⁷⁾ Importantly, key elements such as value-based treatment preferences, ACP, and discussions about preferred place of death—areas often insufficiently addressed in conventional care—have demonstrated increased documentation rates following palliative care interventions. Furthermore, patients receiving palliative care tend to have lower rates of hospital readmission and reduced use of non-beneficial intensive medical interventions. This suggests that palliative care may help achieve an appropriate balance between treatment intensity and QoL, which is crucial in the advanced stage of diseases. Despite these positive findings, consistent survival benefits have not been demonstrated. The heterogeneity in study populations and sample sizes likely contributes to this inconsistency.

The outcomes according to palliative care delivery models are summarized below.

KQ 1.1. Consultative palliative care

Consultative palliative care involves the HF care team maintaining primary responsibility for treatment, while a palliative care team provides support for symptom management and shared decision-making. This model offers timely support without substantially increasing treatment burden for patients or duplicating primary HF care responsibilities.¹⁸⁾ The SWAP-HF trial, which implemented a social worker–led consultative intervention initiated during hospitalization and continued in the outpatient setting, demonstrated effective support for ACP. In the intervention group, concordance between physician and patient prognostic understanding reached 94%, compared with 26% in the control group. Documentation of care preferences in the electronic medical record was also significantly higher in the intervention group. Although no statistically significant differences were observed in QoL, depression, or anxiety scores at 6 months, consultative interventions were associated with improved communication and better alignment of treatment goals with patients’ values.¹⁵,¹⁹⁾

KQ 1.2. Inpatient palliative care

Inpatient palliative care involves a multidisciplinary team—including physicians, nurse practitioners, and social workers—intervening early during hospitalization to assess symptoms and align treatment preferences. The core feature of this model is early intervention at the time of admission for acute decompensation, enabling timely support for critical decisions such as do-not-resuscitate (DNR) orders or hospice referral. Studies report that inpatient palliative care reduces symptom burden and improves QoL by systematically addressing severe symptoms from the early phase of hospitalization.²⁰,²¹⁾ Some studies have also observed reductions in non-beneficial intensive care utilization and high-intensity interventions, as well as decreased healthcare resource use.¹⁶,²²⁾ However, most available evidence derives from small, single-center studies. Consistent reductions in mortality or long-term readmission rates have not been demonstrated, and therefore survival benefit remains uncertain.²³⁾

KQ 1.3. Home-based palliative care

Home-based palliative care involves specialist teams providing home visits or telehealth-based monitoring and self-management support. Clinical and economic outcomes vary depending on the intensity and delivery format of the intervention (in-person visits versus remote counseling). In the Swedish PREFER study, high-intensity integrated home visits by a multidisciplinary team significantly improved quality-adjusted life years by 0.25 compared with usual care and improved symptoms such as nausea.²⁴⁾ From an economic perspective, reductions in hospital admissions and emergency transfers outweighed personnel costs, resulting in approximately €61,000 in cost savings per patient.¹⁶⁾ In contrast, the ENABLE CHF-PC trial in the United States, which employed a nurse-led telephone coaching model, did not demonstrate significant improvements in QoL, depression, or anxiety at 16 weeks. However, significant reductions in pain intensity and pain-related functional interference were observed. Additionally, improvements in QoL were noted among patients with lower baseline QoL at study entry.¹⁴,²⁵,²⁶⁾ Unlike the home-visit model, the telehealth model did not demonstrate statistically significant reductions in hospitalization days or emergency department visits. Overall, home-based palliative care contributes to improved QoL and symptom control. Intensive models incorporating direct home visits have demonstrated potential cost-saving benefits in selected healthcare systems.

Basis for recommendation

1) Level of evidence

Most studies evaluating palliative care interventions in advanced HF are small in scale, show inconsistent results regarding hard clinical outcomes such as survival, and are subject to potential publication bias. Therefore, the overall certainty of evidence is rated as low (Supplementary Tables 1 and 2, Supplementary Figures 2, 3, 4, 5).

2) Benefits and harms

Clear benefits have been demonstrated in physical and emotional symptom relief, improving QoL, clarification of treatment goals, and facilitation of ACP. These effects are relatively consistent across delivery models. No significant harms or adverse consequences attributable to palliative care interventions have been clearly reported. Therefore, the overall balance of benefits and harms favors palliative care.

3) Values and preferences

Patients with HF and their families frequently express unmet needs for information about prognosis, end-of-life planning, and communication regarding care goals. Palliative care addresses these needs and facilitates treatment decisions aligned with patients’ values and preferences.

4) Resources and costs

Although palliative care requires multidisciplinary personnel and initial resource investment, it may reduce overall healthcare expenditures through shorter hospital stays and decreased use of non-beneficial high-intensity interventions. Particularly, home-based models appear to demonstrate favorable cost-effectiveness.

5) Equity, acceptability, and feasibility

Palliative care promotes equitable access to symptom-based care by focusing on patient needs rather than prognosis alone, thereby potentially improving health equity. It is generally perceived positively by patients and families due to its emphasis on symptom relief, QoL, and communication support. Healthcare providers and institutional stakeholders may also accept integration of palliative care within the continuum of HF management. However, feasibility may vary depending on institutional infrastructure, availability of dedicated palliative care personnel, and the level of policy support.

Although the certainty of evidence is low, the committee concludes that a strong recommendation is appropriate in this context, consistent with the GRADE 1C scenario (strong recommendation, low certainty evidence). The rationale is as follows: 1) the benefit-harm balance is clearly and consistently favorable across all three delivery models; 2) no palliative care intervention has been associated with meaningful harm in any study reviewed; 3) patients with advanced HF and their families consistently express unmet needs for prognostic communication and goal-concordant care that palliative care specifically addresses; and 4) expert consensus within the multidisciplinary guideline development committee, including both HF and palliative care specialists, unanimously supported integration of palliative care. Taken together, these factors justify a strong recommendation despite low evidence certainty.

KQ 2. Is oxygen therapy beneficial for patients with end-stage HF who complain of dyspnea?

Dyspnea is one of the most distressing symptoms in patients with advanced HF. Oxygen therapy has traditionally been used empirically for symptom relief; however, its effectiveness in non-hypoxemic patients remains uncertain. In the systematic review addressing this key question, five small experimental studies were identified.²⁷,²⁸,²⁹,³⁰,³¹⁾ These studies included one randomized controlled trial and several crossover trials, and they evaluated dyspnea, exercise duration, and QoL using heterogeneous oxygen concentrations, delivery methods, and assessment time points. Overall risk of bias was moderate to high because of lack of blinding, selective reporting, and small sample sizes; therefore, the overall level of evidence was judged to be low (Table 6).

Table 6. Oxygen therapy in patients with advanced HF.

Categories	Details
Recommendation	Routine oxygen therapy for dyspnea in patients with advanced HF is not recommended.
Strength of recommendation	Not recommended
Level of evidence	Low
Key rationale beyond evidence	No symptom or survival benefit demonstrated; potential harm (CO2 retention)

HF = heart failure.

Across these studies, exercise duration tended to increase slightly with oxygen therapy, but the differences were not statistically significant and were unlikely to represent clinically meaningful improvement. Similarly, dyspnea scores showed no significant difference between oxygen and control groups at most time points, and only one study reported a small short-term QoL benefit without sustained longer-term improvement.³¹⁾ Taken together, the expected benefits of oxygen therapy in patients with advanced HF are small and uncertain.

Although oxygen therapy rarely causes serious acute adverse events, it imposes small but persistent physical, psychological, and practical burdens.³²,³³,³⁴⁾ Nasal cannula or masks may cause mucosal dryness, skin irritation, and discomfort, while oxygen equipment can restrict mobility and daily activity. In addition, oxygen may be perceived as an essential life-saving treatment, creating psychological dependence even when clinical benefit is uncertain. Therefore, oxygen should not be initiated solely because a patient reports dyspnea; clinicians should first document oxygen saturation, clarify the likely mechanism of breathlessness, and consider alternative strategies such as diuretic optimization, positioning, non-pharmacologic support, and individualized symptom-focused care.³⁵⁾ Routine oxygen therapy also requires equipment, personnel support, and ongoing resource use, especially in home-based settings. In the absence of clear evidence of meaningful benefit, indiscriminate use in non-hypoxemic patients is unlikely to be cost-effective and may reduce resource efficiency.³⁶,³⁷⁾ Appropriate implementation may therefore be supported by institutional protocols that require confirmation of hypoxemia, assessment of reversible causes of dyspnea, and structured reassessment after initiation. When dyspnea is primarily related to congestion or fluid overload, management should prioritize diuretic optimization, sodium and fluid restriction, and positioning strategies.³⁸,³⁹,⁴⁰⁾ Other non-pharmacologic interventions, including airflow therapy, environmental modification, breathing techniques, anxiety management, and psychological support, may also be helpful. Although the evidence base for some alternatives is limited, these approaches more directly target symptom mechanisms and patient experience than routine oxygen therapy in non-hypoxemic individuals.

Basis for recommendation

1) Level of evidence

Studies evaluating oxygen therapy in advanced HF are small and limited by risk of bias, imprecision, and indirectness. Overall certainty of evidence is low to very low (Supplementary Tables 3 and 4, Supplementary Figures 6, 7, 8, 9).

2) Benefits and harms

Benefits in exercise duration, dyspnea relief, and QoL are small or uncertain. In contrast, ongoing discomfort, equipment burden, psychological dependence, and resource utilization are present. The overall net benefit is minimal or uncertain.

3) Values and preferences

Patient perceptions of oxygen therapy vary considerably. Decisions regarding its use should be individualized through shared decision-making rather than applied routinely.

4) Resources, equity, and feasibility

Routine oxygen therapy consumes equipment and personnel resources without clear benefit. Reducing unnecessary use may improve resource efficiency and equity. Implementation barriers include entrenched clinical practice patterns and psychological reliance, but institutional protocols and education can facilitate appropriate use.

In conclusion, routine oxygen therapy for dyspnea in patients with advanced HF is not recommended. Oxygen therapy has traditionally been used empirically for dyspnea relief; however, the available evidence does not support routine use in non-hypoxemic patients with advanced HF. The rationale for this recommendation is the absence of demonstrated symptom benefit combined with the potential for harm through hypercapnia in patients with concurrent respiratory compromise. In hypoxemic patients (SpO₂ <90%), supplemental oxygen remains clinically appropriate and is not addressed by this recommendation.

KQ 3. Is diuretic administration beneficial for patients with end-stage HF who complain of congestive symptoms?

Patients with advanced HF frequently experience recurrent congestion and dyspnea, resulting in substantial physical distress. At this stage, the therapeutic goal often shifts from prolonging survival to alleviating symptoms related to fluid overload and maintaining QoL.⁴¹,⁴²⁾ Diuretics are central to symptom control in this context. However, evidence regarding optimal diuretic strategies specifically in end-stage or palliative care settings remains limited. Seven randomized controlled trials were identified, enrolling patients with acute decompensated HF and compared various strategies, including bolus versus continuous intravenous infusion, high-dose versus low-dose intravenous therapy, and add-on therapies (Table 7).

Table 7. Diuretics in patients with advanced HF.

Categories	Details
Recommendation	Diuretic therapy for patients with advanced HF who experience symptoms of congestion can be recommended.
Strength of recommendation	Conditional recommendation
Level of evidence	Very low
Key rationale beyond evidence	Pathophysiological rationale; standard clinical practice; individualized dosing required

HF = heart failure.

Representative trials included the DOSE trial (high vs. low dose; bolus vs. continuous infusion),⁴³⁾ the DRAIN study (continuous infusion vs. bolus in high-risk acute decompensated HF),⁴⁴⁾ Palazzuoli et al.⁴⁵⁾ (continuous vs. bolus), ADVOR study (acetazolamide add-on),⁴⁶⁾ CLOROTIC study (thiazide add-on),⁴⁷⁾ and the QUEST and EVEREST trials (tolvaptan add-on).⁴⁸,⁴⁹⁾ Because these studies were conducted predominantly in acute care populations rather than explicitly in end-stage patients, their findings must be applied indirectly. Consequently, the overall certainty of evidence was rated as very low. Nevertheless, given the essential role of diuretics in relieving congestion and dyspnea, a conditional recommendation is made, emphasizing individualized treatment according to patient status and goals of care.

The overall effects of diuretic therapy were evaluated by comparing “intensified strategies” (continuous infusion, high-dose therapy, or combination regimens) with “standard strategies” (intermittent bolus, lower doses, or monotherapy). Outcomes included weight reduction, decongestion, dyspnea relief, and adverse events such as renal dysfunction and electrolyte abnormalities.

In patients with advanced HF, intensified diuretic strategies (continuous infusion, high dose, or combination therapy) may provide greater short-term weight reduction and decongestion compared with standard strategies. However, these benefits are accompanied by increased risks of renal dysfunction, electrolyte disturbances, and hypotension, potentially destabilizing frail patients. Given that the primary goal in palliative settings is symptom relief rather than survival prolongation, intensified diuresis cannot be assumed to be universally superior solely based on greater short-term fluid removal. Treatment decisions should be guided by the patient’s goals—whether prioritizing aggressive symptom control or minimizing adverse effects and treatment burden.

KQ 3.1. Continuous infusion vs. intermittent bolus

Results comparing continuous infusion and intermittent bolus administration have been inconsistent across patient populations. In the large DOSE trial, no significant differences were observed between the two approaches in terms of symptom improvement, decongestion markers, 60-day mortality and rehospitalization.⁴³⁾ These findings suggest that, in general acute HF populations, neither strategy demonstrates clear superiority. In contrast, smaller studies in higher-risk populations, including the DRAIN study and the Palazzuoli et al.,⁴⁵⁾ reported that continuous infusion resulted in greater urine output at 72 hours, more pronounced weight reduction, and larger decreases in BNP levels.⁴⁴⁾ These findings suggest that in selected high-risk patients, continuous infusion may provide short-term decongestive advantages, although evidence remains inconsistent. Overall, rather than favoring one strategy universally, the choice between continuous infusion and bolus administration should be individualized based on disease severity and clinical risk.

KQ 3.2. High-dose vs. low-dose intravenous therapy

In the DOSE trial, high-dose intravenous therapy (2.5 times the baseline oral dose) produced greater diuresis and weight loss compared with low-dose therapy (equivalent to baseline oral dosing). High-dose therapy was associated with improved short-term decongestion and symptom relief.⁴³⁾ However, this intensified strategy was also associated with more frequent transient worsening of renal function, including rises in serum creatinine. Furthermore, no significant differences were observed in longer-term outcomes such as 60-day mortality or rehospitalization. Thus, while high-dose diuretic therapy may be conditionally appropriate when rapid symptom relief is prioritized, it should be used cautiously with careful monitoring for renal dysfunction.

KQ 3.3. Add-on combination strategies

Adding acetazolamide, thiazides, or tolvaptan to loop diuretics has been shown to enhance short-term decongestion. In ADVOR trial, addition of acetazolamide to standard loop diuretic therapy significantly increased rates of successful decongestion within three days.⁴⁶⁾ CLOROTIC trial showed benefits of additional hydrochlorothiazide in weight reduction and diuretic response, but also revealed association with higher incidence of worsening renal function.⁴⁷⁾ Tolvaptan add-on therapy demonstrated short-term benefit in improvement of congestion markers, dyspnea, and serum sodium levels, but it did not demonstrate long-term benefit in mortality or rehospitalization in QUEST and EVEREST trials.⁴⁸,⁴⁹⁾ Overall, combination strategies tend to improve short-term symptom relief but do not improve long-term prognosis. Depending on the agent, risks of renal dysfunction or electrolyte imbalance may increase.

Basis for recommendation

1) Level of evidence

Direct comparative studies specifically in end-stage HF are lacking. Most evidence derives from acute decompensated HF populations and must be applied indirectly. Differences in study populations, small sample sizes, and inconsistent outcomes contribute to substantial uncertainty. The overall certainty of evidence is therefore rated as very low (Supplementary Tables 5 and 6, Supplementary Figures 10, 11, 12).

2) Benefits and harms

Diuretics clearly provide short-term benefits in weight reduction, decongestion, and dyspnea relief. However, differences between strategies are modest, and long-term prognostic benefits are limited. Intensified regimens may increase the risk of renal dysfunction and electrolyte imbalance, necessitating careful risk–benefit assessment.

3) Values and preferences

Patients with advanced HF often prioritize relief of dyspnea, edema reduction, and preservation of daily functioning over survival extension. However, some patients may value renal preservation or minimizing medical intervention. Shared decision-making is therefore essential.

4) Resources and equity

Standard diuretic strategies are relatively inexpensive and widely accessible, supporting favorable resource use and equity. In contrast, continuous infusion or high-dose strategies may require hospitalization and closer monitoring, increase healthcare resource utilization, thereby limiting feasibility in resource-constrained settings.

Although the overall certainty of evidence is very low due to the absence of randomized comparator studies in the end-stage HF population, the committee notes that diuretic therapy for symptom relief is supported by strong pathophysiological reasoning and is a foundational component of standard HF management. The conditional recommendation reflects the limited formal evidence and the need for individualized dosing decisions, but the direction of recommendation reflects broad expert consensus. These non-evidence factors were determinative in setting the recommendation direction.

KQ 4. Is inotrope administration beneficial for patients with end-stage HF who complain of congestive symptoms?

Patients with advanced HF frequently experience severe symptoms related to reduced cardiac output and persistent congestion, including refractory dyspnea. In selected patients, short-term inotropic therapy may be considered to alleviate symptoms and improve QoL. Ten randomized controlled trials evaluating inotropic agents (including dobutamine) and five non-randomized observational studies examining continuous intravenous inotropic therapy were reviewed to assess efficacy and safety.⁵⁰,⁵¹,⁵²,⁵³,⁵⁴,⁵⁵,⁵⁶,⁵⁷,⁵⁸,⁵⁹,⁶⁰,⁶¹,⁶²,⁶³,⁶⁴⁾ Overall, inotropic therapy demonstrated short-term improvements in functional capacity and relief of congestion-related symptoms. These findings suggest that in patients whose primary treatment goal is QoL improvement rather than survival prolongation, inotropes may provide clinically meaningful benefits. However, important risks have been reported, including central venous catheter–related bloodstream infection, arrhythmia, and drug dependence.⁶⁰,⁶¹⁾ Long-term survival benefit remains uncertain, and potential adverse effects on mortality cannot be excluded. Accordingly, the overall certainty of evidence was rated as low. Given the balance of potential benefits and harms, a conditional recommendation was made, emphasizing shared decision-making in patients prioritizing symptom relief (Table 8).

Table 8. Inotropic therapy in patients with advanced HF.

Categories	Details
Recommendation	Inotropic therapy for patients with advanced HF who experience symptoms of congestion can be recommended.
Strength of recommendation	Conditional recommendation
Level of evidence	Low
Key rationale beyond evidence	Hemodynamic rationale; symptom palliation; individual risk-benefit assessment required

HF = heart failure.

Meta-analytic findings indicate that inotropic therapy is associated with short-term improvements in functional status and symptom burden. Reported benefits include reduction in HF-related hospitalization,⁵⁰,⁵¹,⁵⁸⁾ improvement in dyspnea and functional class,⁵⁶,⁵⁹⁾ decrease in left ventricular end-diastolic volume and increase in LVEF.⁵⁶,⁵⁷,⁵⁹⁾ These improvements suggest that inotropic therapy can enhance short-term activity tolerance and relieve congestion. In patients with advanced disease whose goals focus on symptom palliation rather than life prolongation, these short-term benefits may be clinically meaningful.

Because inotropic therapy often requires continuous intravenous administration via a central venous catheter, several potential harms have been reported. Reported harms include central line-associated bloodstream infection (approximately 15%),⁶¹⁾ drug-related adverse events such as hypotension and arrhythmias (approximately 20–25%), increased risk of drug dependence, and hemodynamic instability.⁶⁰,⁶¹⁾ Although evidence regarding mortality is inconsistent, concerns remain regarding potential long-term harm, including increased myocardial oxygen consumption and arrhythmogenic risk.

Overall, the balance of benefits and harms may be characterized as small to moderate net benefit when symptom relief is prioritized, but uncertain or potentially unfavorable when long-term survival is the primary goal. Short-term beneficial effects of inotropic therapy can be meaningful in patients with adverse HF who prioritize symptom relief, although it arises potential risk including infection, catheter management, arrhythmogenic burden. Given the variability in patient values and limited direct data on patient preferences, treatment decisions should be guided by shared decision-making.

Continuous intravenous inotropic therapy requires substantial healthcare infrastructure, including central venous catheter placement and maintenance, infection prevention systems, regular clinical monitoring and accessibility to home nursing services. From an equity perspective, access is more readily available in resource-rich settings such as large urban hospitals, whereas it may be limited in rural areas, among patients living alone, or in socioeconomically vulnerable populations.

In terms of cost, home-based maintenance therapy may reduce hospitalization expenses; however, prolonged treatment may increase overall costs due to complications such as infection and rehospitalization. Adequate infrastructure is also essential, including specialized staff training, standardized infection control protocols, and established home nursing and emergency response systems. Accordingly, the practical implementation of inotropic therapy may be constrained by available resources and the patient’s caregiving environment.

Several structural and clinical barriers may hinder the implementation of inotropic therapy. The most significant obstacle is the burden associated with central venous catheter management. Risks such as catheter-related bloodstream infection, mechanical complications, and arrhythmias necessitate the availability of specialized personnel, robust infection control systems, and emergency response infrastructure. In institutions where such infrastructure is limited, safe implementation of this intervention is challenging. In addition, the absence of adequate equipment for continuous intravenous administration and insufficient home-based monitoring systems may preclude safe long-term treatment.

The caregiving environment is another critical factor. In patients who live alone or lack sufficient caregiver support, maintaining therapy may be difficult and the risk of complications may increase. Patient and caregiver anxiety regarding infection, the burden of device management, and the need for ongoing communication with healthcare providers may further reduce treatment acceptability. In particular, variable home-based palliative care systems across institutions and regions may result in disparities in access, thereby limiting implementation in certain regions or facilities. Finally, although inotropic therapy may offer short-term benefits, its potential harms are not negligible, and the overall level of evidence remains limited. This contributes to clinical uncertainty among healthcare providers regarding intervention selection, which may further impede adoption in practice.

To ensure effective application of the recommendation in clinical practice, standardized practical tools, educational materials, and user-adapted versions of the guideline are necessary. Simple checklists and summary documents incorporating essential elements—such as central venous catheter care, infection prevention, and adverse-event monitoring—can support clinician decision-making and promote consistency in care delivery. Stepwise administration protocols outlining indications, contraindications, and monitoring plans may serve as particularly useful tools for less experienced clinicians.

In patients with advanced HF who experience symptoms of congestion, several therapeutic options may be considered alongside inotropic therapy. First, optimal GDMT and careful adjustment of diuretics should be maintained as foundational treatment. When necessary, intensified decongestion strategies—such as high-dose or intravenous diuretics and vasodilators (e.g., venous or arterial vasodilators)—may be considered to improve volume control. Other palliative interventions, including positioning, low-dose opioids, and anxiolytics, may help alleviate symptom burden in patients with advanced HF. These approaches should be integrated into a holistic management plan that considers the patient’s prognosis, comorbidities, caregiving environment, and overall goals of care (life prolongation versus QoL), together with consideration of whether to initiate or continue inotropic therapy.

Basis for recommendation

1) Level of evidence

Randomized trials evaluating inotropic therapy in advanced HF are limited by risk of bias, small sample size, and imprecision. Overall certainty of evidence is rated as low (Supplementary Tables 7 and 8, Supplementary Figures 13, 14, 15, 16).

2) Benefits and harms

Inotropic therapy provides clear short-term benefits in dyspnea relief and functional improvement, but long-term survival benefits are uncertain. Central line–related infection and arrhythmia represents meaningful risks. The overall net benefit is considered small to moderate when symptom relief is prioritized.

3) Values and preferences

Patients vary in the relative importance they assign to symptom relief, life prolongation, and avoidance of complications. Therefore, decisions regarding inotropic therapy should be made through shared decision-making aligned with individual treatment goals.

4) Resources and equity

Continuous inotropic therapy requires substantial infrastructure and healthcare resources. Limited access in certain regions or institutions may restrict feasibility and raise equity concerns. Strengthening regional support systems may improve equitable access.

In conclusion, inotropic therapy in advanced HF provides short-term symptomatic benefits but carries meaningful risks and uncertain long-term survival impact. Therefore, inotropic therapy should not be routinely recommended for all patients, but may be conditionally considered in those whose primary goal is symptom relief. Decisions should be individualized, incorporating patient values, caregiving environment, and available healthcare resources.

KQ 5. Is opioid administration beneficial for patients with end-stage HF who complain of dyspnea or pain?

Patients with advanced HF experience complex symptoms, including dyspnea and pain. Pain is common yet frequently underrecognized in HF, and its prevalence increases with disease severity; more than 80% of HF patients with NYHA class IV report significant pain.⁶⁵⁾ Persistent and inadequately managed pain imposes substantial physical burden and negatively affects QoL.⁶⁶⁾ Accordingly, pain management represents an important component of palliative care in HF.⁶⁷⁾

Opioids are sometimes selectively used in clinical practice to relieve refractory symptoms. However, unlike in patients with malignancy, evidence supporting the efficacy and safety of opioids for dyspnea relief in HF is limited and inconclusive.⁶⁸,⁶⁹,⁷⁰,⁷¹⁾ In the literature review conducted for this guideline, only one randomized controlled trial and one retrospective observational study specifically evaluated opioid use in advanced HF.⁷⁰,⁷²,⁷³⁾ In the randomized trial conducted by Johnson et al.,⁷⁰⁾ opioid therapy in patients with advanced HF did not result in statistically or clinically meaningful improvements in either average dyspnea or worst dyspnea compared with control, and no consistent benefit was observed in pain, HF-specific QoL, functional status, cognition, or sleep outcomes. In the retrospective study by Sheriff et al.,⁷³⁾ opioid use in advanced HF was associated with worse long-term outcomes, including higher mortality at 30 days, 1 year, and 8.6 years of follow-up (hazard ratio [HR] for all-cause mortality at 8.6 years, 1.49; 95% confidence interval [CI], 1.11 to 1.99). Although causality cannot be inferred, the absence of clear symptomatic benefit and the signal toward harm make the overall benefit-harm balance unfavorable for routine use. For most patients, management of pain and dyspnea should focus first on identifying treatable causes and using non-opioid approaches. Ischemic chest pain should be treated with appropriate antianginal therapy, while neuropathic, nociceptive, or inflammatory pain may require condition-specific non-opioid strategies such as anticonvulsants, antidepressants, or non-pharmacologic interventions.⁷⁴,⁷⁵⁾ For dyspnea, clinicians should assess congestion, tachycardia, anemia, and concomitant pulmonary disease, and prioritize optimization of diuretics, hemodynamic management, rhythm control, positioning, breathing support, and psychological support. In patients with extremely limited life expectancy who have severe, refractory dyspnea or pain despite optimal management, opioid therapy may be considered only within a hospice context when treatment goals are explicitly focused on comfort rather than life prolongation. In such cases, the minimum effective dose and close monitoring are essential (Table 9).

Table 9. Opioid administration in patients with advanced HF.

Categories	Details
Recommendation	Routine use of opioids in patients with advanced HF is not recommended.
Strength of recommendation	Not recommended
Level of evidence	Low
Key rationale beyond evidence	Safety concerns in HF (respiratory depression, hemodynamic instability); insufficient HF-specific data

HF = heart failure.

Basis for recommendation

1) Level of evidence

Evidence evaluating opioid use in advanced HF is extremely limited, with small sample sizes, risk of bias, and imprecision. The overall certainty of evidence is rated as low to very low (Supplementary Tables 9 and 10, Supplementary Figures 17, 18, 19, 20, 21, 22).

2) Benefits and harms

No clear improvement in dyspnea, pain, or QoL has been demonstrated, while observational data suggest a potential increase in long-term mortality. The overall net benefit is small or unfavorable.

3) Values and preferences

Although symptom relief is highly valued in patients with advanced HF, current evidence does not support routine opioid use to achieve this goal. Limited use may be considered only in patients with severe, refractory symptoms within a hospice context and following shared decision-making.

4) Resources and equity

Opioids are inexpensive and widely available, and major equity concerns are unlikely. However, unnecessary or excessive use may increase healthcare utilization, reducing overall resource efficiency.

In conclusion, routine opioid use in patients with advanced HF is not recommended due to limited symptomatic benefits and potential signals of harm. Opioids may be considered only in exceptional circumstances—specifically in patients with very poor prognosis and refractory symptoms—within a hospice or end-of-life context, following careful shared decision-making that aligns with patient values and goals of care.

KQ 6. Is benzodiazepine administration beneficial for patients with end-stage HF who complain of dyspnea?

Patients with advanced HF frequently experience severe dyspnea accompanied by anxiety or panic symptoms. For this reason, benzodiazepines have been used empirically in clinical practice. However, potential harms—including respiratory depression, cognitive impairment, and increased fall risk—necessitate careful evaluation of their effectiveness and safety in this population. A systematic review identified no randomized controlled trials directly evaluating benzodiazepine use for dyspnea in patients with advanced HF. Available evidence was limited to indirect data from studies in patients with advanced cancer or COPD, as well as observational HF cohorts (Table 10).⁷⁶,⁷⁷,⁷⁸⁾

Table 10. Benzodiazepine administration in patients with advanced HF.

Categories	Details
Recommendation	Routine use of benzodiazepines for dyspnea relief in patients with advanced HF is not recommended.
Strength of recommendation	Not recommended
Level of evidence	Very low
Key rationale beyond evidence	No HF-specific efficacy data; HF-specific safety risks (hypotension, falls, arrhythmia)

HF = heart failure.

No direct evidence from randomized controlled trials or cohort studies in advanced HF demonstrates that benzodiazepines relieve dyspnea. A Cochrane meta-analysis including eight randomized controlled trials in advanced cancer or COPD populations found that short-term benzodiazepine use (e.g., midazolam, diazepam) did not significantly improve dyspnea scores compared with placebo or other agents and was associated with increased sedation and drowsiness, which included only 4 patients with chronic heart disease among 298 total participants.⁷⁶⁾ Although benzodiazepines may provide short-term improvement in sleep quality in general insomnia populations, long-term use is discouraged due to risks of tolerance, dependence, and cognitive impairment.⁷⁸⁾ In HF patients, systematic assessment of anxiety or insomnia outcomes is lacking. Overall, indirect evidence suggests that benzodiazepines are unlikely to provide meaningful dyspnea relief in advanced HF. A limited benefit in patients with concomitant anxiety may be plausible, but it has not been quantitatively established. In an acute HF cohort, benzodiazepine users (who tended to have more severe illness) did not demonstrate statistically significant differences in 7-day mortality (relative risk [RR], 1.49; 95% CI, 0.83 to 2.68), 30-day or 180-day mortality, or 6-month HF readmission and emergency department visits compared with non-users; however, residual confounding cannot be excluded.⁷⁷⁾ In a chronic HF cohort of patients with insomnia, benzodiazepine use was associated with a significantly increased risk of HF readmission (HR, 1.53; 95% CI, 1.03 to 2.28) and a trend toward increased cardiac mortality (HR, 1.36; 95% CI, 0.94 to 1.95).⁷⁸⁾ Additional concerns in older HF populations include increased risk of falls, fractures, and delirium, although these outcomes were not systematically reported in the included studies.

Taken together, benzodiazepines did not demonstrate a significant effect on dyspnea relief and were associated with increased sedation. Some observational studies suggested a possible association with increased risk of HF-related readmission and mortality. Given the absence of direct high-quality evidence and signals of potential harm, the overall certainty of evidence was rated as very low, and the recommendation was graded as “Do not recommend.”

Basis for recommendation

1) Level of evidence

No direct randomized trials in advanced HF exist. Evidence relies on indirect populations and observational cohorts. Due to risk of bias, indirectness, and imprecision, overall certainty of evidence is rated as very low (Supplementary Tables 11 and 12, Supplementary Figures 23, 24, 25, 26).

2) Benefits and harms

No convincing evidence supports a direct effect of benzodiazepines on dyspnea relief. Sedation and cognitive impairment are common adverse effects, and observational data suggest possible association with higher HF readmission and a potential signal toward increased cardiac mortality. Overall net benefit is minimal or unfavorable.

3) Values and preferences

While patients often prioritize dyspnea relief, many also value preservation of alertness and cognitive clarity. Concerns regarding dependence, cognitive impairment, and loss of autonomy may influence patient preferences. Therefore, if benzodiazepines are considered, decisions should follow careful discussion of potential benefits and risks.

4) Resources, equity, and acceptability

Benzodiazepines are inexpensive and widely accessible. However, indirect societal costs related to falls, delirium, and rehospitalization may increase overall healthcare burden. Older and vulnerable populations may be disproportionately affected by adverse effects. A recommendation against routine use is feasible to implement, as it does not require additional infrastructure and may be supported by simple strategies such as educational materials or prescribing protocols. Limited, case-by-case use in patients with prominent anxiety or insomnia is likely to be acceptable to clinicians when supported by multidisciplinary consensus.

The evidence bases for benzodiazepine use for dyspnea derives predominantly from oncology and COPD populations. This represents a significant limitation in extrapolating findings to HF patients, who differ in several important ways: 1) hemodynamic instability and reduced cardiac output reduce the safe dosing margin for agents with vasodilatory properties; 2) cardiorenal syndrome alters drug clearance unpredictably; and 3) polypharmacy with vasodilators, diuretics, and antiarrhythmic agents creates interaction risks absent in oncology or COPD populations. HF-specific observational data yield conflicting results, with some studies demonstrating increased rehospitalization and cardiovascular mortality with benzodiazepine use in patients with heart failure with reduced ejection fraction. Given the absence of HF-specific efficacy evidence and the disease-specific safety concerns, routine benzodiazepine use for dyspnea in advanced HF is not recommended. Short-term, low-dose use may be considered only in selected patients with prominent anxiety or panic contributing to breathlessness, after non-pharmacologic measures and optimization of HF therapy, through shared decision-making.

KQ 7. Is pharmacological treatment beneficial for relieving depression and anxiety in patients with end-stage HF?

Depression and anxiety are major contributors to reduced QoL in patients with advanced HF.⁶⁹,⁷⁹,⁸⁰⁾ However, patients may be reluctant to initiate pharmacologic treatment due to concerns regarding adverse effects and polypharmacy. To evaluate whether antidepressant therapy improves depressive symptoms in advanced HF, relevant literature was reviewed. The analysis included randomized controlled trials.⁸¹,⁸²,⁸³⁾ in patients with advanced or terminal illness and observational studies⁸⁴,⁸⁵⁾; however, study sizes were small and risk of bias was substantial (Table 11).

Table 11. Antidepressant for depression in patients with advanced HF.

Categories	Details
Recommendation	Antidepressant may be considered for the relief of depressive symptoms in patients with advanced HF.
Strength of recommendation	Conditional recommendation
Level of evidence	Very low
Key rationale beyond evidence	Psychological benefit; acceptable safety profile in selected agents (sertraline preferred); expert consensus

HF = heart failure.

The primary benefit of antidepressant therapy was defined as reduction in depressive symptoms. In two randomized controlled trials using MADRS, antidepressant-treated patients demonstrated a trend toward increased response at 14 days (RR, 2.16; 95% CI, 0.74 to 6.27), although this did not reach statistical significance.⁸⁶,⁸⁷⁾ In a single randomized trial using HADS-D, a modest reduction of 1.3 points (median, −1.30; 95% CI, −2.10 to −0.50) was reported.⁸⁷⁾ A 1.7-point reduction in ESAS-depression was also observed in one study,⁸⁸⁾ although the limited number of trials prevents firm conclusions. Non-randomized studies suggested symptomatic improvement, but absence of control groups precludes causal inference.⁸⁹⁾ Critically, two landmark randomized controlled trials conducted specifically in HF patients—SADHART-CHF (sertraline, n=469)⁸³⁾ and MOOD-HF (escitalopram, n=372)⁸¹⁾—both failed to demonstrate efficacy of selective serotonin reuptake inhibitors (SSRIs) beyond placebo for depression in HF, in contrast to findings in oncology or general adult populations. MOOD-HF was stopped early for futility. SSRIs carry HF-specific safety risks including hyponatremia (via SIADH, particularly hazardous in patients receiving diuretics), QTc prolongation (especially citalopram and escitalopram), platelet inhibition with anticoagulation interactions, and CYP2D6-mediated metabolic interactions with beta-blockers (fluoxetine, paroxetine). Venlafaxine and duloxetine (serotonin-norepinephrine reuptake inhibitors) carry additional risk of HF decompensation via norepinephrine-mediated neurohormonal activation.

Antidepressant therapy may provide modest short-term improvement in depressive symptoms, but effect magnitude and persistence remain uncertain. No clear safety signal has been identified, although the limited evidence base prevents complete exclusion of potential harm. Therefore, antidepressants should not be universally recommended for all patients but may be considered selectively in those with clinically meaningful depressive symptoms and where symptom relief is a priority.

Basis for recommendation

1) Level of evidence

Available studies are small, indirect, and at risk of bias, with imprecision in effect estimates. The overall certainty of evidence is rated as very low (Supplementary Tables 13 and 14, Supplementary Figures 27, 28, 29, 30, 31).

2) Benefits and harms

Antidepressants may provide modest short-term improvement in depression scores, but magnitude and consistency of benefit are limited. No clear increase in mortality or serious adverse events has been observed; however, evidence remains insufficient to fully exclude harm. The overall net benefit is considered small.

3) Values and preferences

Patients often value relief of depressive symptoms but may be concerned about medication burden and adverse effects. Decisions should therefore be individualized through shared decision-making.

4) Resources and equity

Antidepressants are inexpensive and widely accessible, and do not require additional infrastructure. Monitoring for drug interactions and adverse effects remains necessary in patients with polypharmacy.

In conclusion, antidepressant therapy may provide limited short-term benefit for depressive symptoms in patients with advanced HF, but the certainty of evidence is very low. Therefore, antidepressant therapy should not be routinely prescribed for all patients, but may be conditionally considered in those with clinically meaningful depression and where symptom relief is a key treatment goal, within an integrated approach that includes non-pharmacologic interventions.

KQ 8. Is non-pharmacological intervention beneficial for relieving depression and anxiety in patients with end-stage HF?

Patients with advanced HF experience not only physical symptoms but also emotional, psychosocial, and spiritual distress that pharmacological therapy alone is often insufficient to address. Non-pharmacological interventions encompass a broad range of approaches — including multidisciplinary palliative care programs, psychosocial counseling, ACP facilitation, digital and telehealth-based platforms, virtual reality–based symptom interventions, and dignity therapy—each targeting different dimensions of patient well-being. The systematic review included four randomized controlled trials conducted in patients with advanced or progressive HF (e.g., online dignity therapy,⁹⁰⁾ virtual reality–based pain interventions,⁹¹,⁹²⁾ home-based case management,⁹³⁾ and spirituality-based palliative care education), as well as one meta-analysis of palliative care interventions and four randomized trials included within that meta-analysis.¹⁴,²⁰,⁹⁴,⁹⁵,⁹⁶⁾ Although considerable heterogeneity existed across studies in terms of intervention components, delivery modes, and outcome measures, findings were directionally consistent in suggesting benefit in QoL and emotional well-being. The overall certainty of evidence was rated as low, reflecting the small sample sizes, heterogeneous study designs, and variability in outcome assessment across included studies (Table 12).

Table 12. Non-pharmacological intervention for depression in patients with advanced HF.

Categories	Details
Recommendation	Non-pharmacologic interventions may be considered to improve QoL and to address anxiety and depressive symptoms in patients with advanced HF.
Strength of recommendation	Conditional recommendation
Level of evidence	Low
Key rationale beyond evidence	Range of modalities available; patient preference; minimal harms; multidisciplinary integration

HF = heart failure.

Non-pharmacological interventions reviewed were generally multicomponent in nature and were delivered by multidisciplinary teams comprising nurses, physicians, and social workers. Core elements commonly included systematic symptom assessment, psychosocial and emotional support, communication regarding prognosis and disease trajectory, ACP, goal-of-care discussions, and caregiver support. Some programs incorporated digital components such as telehealth visits, online platforms, and virtual reality–based approaches. Across studies, interventions were associated with improvements in health-related QoL—particularly in palliative care–specific domains such as symptom control and psychosocial well-being—though effects on general functional status and on depression and anxiety scores were less consistent and did not uniformly reach statistical significance. Notably, non-pharmacological interventions were associated with a meaningful increase in ACP documentation, suggesting a clear benefit in facilitating treatment preference discussions and shared decision-making beyond direct symptom outcomes.

With respect to safety, non-pharmacological interventions were not associated with increased rates of hospitalization, mortality, or other serious medical harms in the reviewed studies. The absence of significant adverse effects, combined with the directionally positive signals in patient-centered outcomes, supports the clinical rationale for their use as adjunctive strategies in comprehensive palliative care. However, wide confidence intervals and limited study numbers preclude definitive safety conclusions.

Patient acceptance of non-pharmacological interventions is generally favorable, as these approaches directly address the psychological, social, and existential dimensions of suffering that patients with advanced HF consistently identify as unmet needs. Preferences may vary according to cultural background, religious beliefs, and digital literacy, and tailored delivery options—including in-person alternatives to digital platforms—should be ensured to avoid widening disparities in access.

Basis for recommendation

1) Level of evidence

Included studies were small and heterogeneous in terms of intervention content and outcome assessment. Although directionally consistent benefit signals were observed, uncertainty remains; thus, the overall certainty of evidence is rated as low (Supplementary Tables 15 and 16, Supplementary Figures 32, 33, 34).

2) Benefits and harms

Non-pharmacologic interventions may provide clinically meaningful improvements in symptom burden, QoL, and psychosocial well-being, with minimal medical harm reported. Although resource investment (time and personnel) is required, patient-centered benefits appear to outweigh these burdens.

3) Values and preferences

Patients with advanced HF and their families often highly value symptom relief and psychological stability. Acceptance of non-pharmacologic interventions is generally favorable, although preferences may vary depending on cultural background, religious beliefs, and digital literacy. Tailored approaches are therefore recommended.

4) Resources and equity

Provision of non-pharmacologic interventions requires trained personnel and time investment. However, these approaches may improve resource efficiency over the long-term period, potentially reducing hospital utilization. To prevent disparities related to digital access, alternative in-person support options should be ensured.

In conclusion, non-pharmacologic interventions in patients with advanced HF may improve QoL, particularly in palliative care–specific domains such as symptom control and psychosocial well-being, and facilitate ACP. Evidence regarding anxiety and depression outcomes is directionally positive but less consistent. Given the low certainty of evidence and heterogeneity of available studies, these interventions are conditionally recommended and should be implemented selectively according to individual clinical circumstances and patient preferences, ideally within an integrated multidisciplinary care model alongside pharmacological therapy.

KQ 9. Is deactivation of an implantable cardioverter defibrillator beneficial in patients with end-stage HF?

Implantable cardioverter-defibrillator (ICD) plays a critical role in preventing sudden cardiac death. However, in the terminal phase of illness, ICD shocks may become repetitive and distressing, significantly compromising quality of death. Cases have been reported in which ICDs continue to deliver shocks even in patients with DNR orders, causing psychological trauma to both patients and families during the dying process. In this key question, the impact of ICD deactivation on quality of death in advanced HF was evaluated. Available evidence was extremely limited, consisting of a single retrospective observational study. Accordingly, the certainty of evidence was rated as very low.⁹⁷⁾ Based on available evidence, the overall effect of ICD deactivation in advanced HF may be summarized as follows: deactivation is likely meaningful in reducing distress caused by shocks, although quantitative evidence is limited and its impact on long-term prognosis remains uncertain (Table 13).

Table 13. Deactivation of an implantable cardioverter-defibrillator in patients with advanced HF.

Categories	Details
Recommendation	Deactivation of implantable cardioverter-defibrillator may be considered to improve quality of death in patients with advanced HF.
Strength of recommendation	Conditional recommendation
Level of evidence	Very low
Key rationale beyond evidence	Ethical imperative; patient values and shared decision-making central; HF specialist + electrophysiolist coordination required

HF = heart failure.

In the included retrospective study,⁹⁷⁾ comparison between patients who underwent ICD deactivation and those who did not showed no statistically significant differences in total shock frequency within 30 days (RR, 0.89; 95% CI, 0.29 to 2.66) or 90 days (RR, 1.02; 95% CI, 0.47 to 2.19) before death. However, wide confidence intervals and a small sample size (n=51) limit interpretability. Importantly, detailed findings revealed that shocks frequently occurred within one month and even within 24 hours prior to death. Notably, some patients with DNR status continued to receive ICD shocks until death, highlighting a clinically significant issue of potentially avoidable suffering. The primary goal of ICD deactivation in advanced HF is not survival modification but prevention of repetitive, painful shocks in the final stage of life. Although quality of death was not directly measured in the included study, reducing the likelihood of terminal shocks is consistent with clinical and ethical principles aimed at minimizing pain and anxiety. International observational studies and consensus statements have consistently reported that ICD shocks near death can cause significant physical pain and emotional distress.⁹⁸⁾ Therefore, ICD deactivation is widely supported in principle, in patients with advanced disease, recurrent shocks, or transition to palliative care. Thus, even in the absence of robust quantitative data, prevention of terminal shocks represents a clinically meaningful potential benefit.

A commonly cited concern is that ICD deactivation may permit untreated malignant arrhythmias, potentially shortening survival. However, in patients whose goals of care have shifted from life prolongation to comfort-focused management, this possibility may be considered ethically acceptable when discussed and agreed upon with patients and families. In the included study, the impact of ICD deactivation on timing of death was not directly assessed, and no quantitative evidence demonstrated shortened survival. In the context of advanced HF and end-of-life care, choosing to avoid painful shocks may align more closely with patient-centered goals than continuing potentially distressing life-prolonging interventions.

From a technical standpoint, ICD deactivation is relatively straightforward. A trained professional can reprogram the device to disable shock therapy without requiring invasive procedures or hospitalization. However, practical barriers exist, including misunderstanding among patients and families (e.g., belief that deactivation causes immediate death, deactivation means cessation of all treatment), emotional burden for clinicians, lack of structured documentation and communication processes, and limited access to programming equipment and trained personnel in home or primary care clinic-based settings. Although the intervention itself is low-cost and minimally resource-intensive, equitable access may vary depending on geographic region and availability of specialized personnel. Establishing standardized protocols and personnel training programs may improve implementation.

Despite limited quantitative evidence, the well-recognized clinical reality that ICD shocks near the end of life may cause severe pain and distress provides a strong ethical rationale for discussing and considering ICD deactivation in appropriate patients. Therefore, a conditional recommendation is made to proactively consider ICD deactivation to prevent unnecessary suffering and facilitate a peaceful death.

Basis for recommendation

1) Level of evidence

Evidence evaluating ICD deactivation in advanced HF is extremely limited and based primarily on small observational data with risk of bias. The overall certainty of evidence is rated as very low (Supplementary Tables 17 and 18, Supplementary Figure 35).

2) Benefits and harms

In patients at the end-of-life, prevention of painful terminal shocks and improvement of quality of death are considered clinically more meaningful benefits than theoretical harm of losing a potential opportunity for resuscitation. The risk ratio for shock occurrence within 30 days before death did not demonstrate a statistically significant reduction by deactivating ICDs; however, this paradoxically suggests that the risk of shocks persists until the very end of life. Therefore, despite uncertainty regarding quantifiable benefit, proactive discussion of ICD deactivation to prevent patient suffering and promote comfort is clinically justified.

3) Values and preferences

Patients and families often prioritize a peaceful and dignified death, although some may strongly prefer life prolongation. Therefore, ICD deactivation decisions must be made through thorough shared decision-making, including clear discussion of prognosis and shock risk.

4) Resources and equity

ICD deactivation is technically simple and low-cost, but access to trained personnel and programming equipment may vary by geographic regions. To reduce regional disparities, standardized protocols and integration with palliative care services are recommended. Multidisciplinary collaboration enhances feasibility and acceptability.

The conditional recommendation for ICD deactivation at end of life is driven primarily by ethical and values-based considerations rather than by formal trial evidence. ICD shocks at end of life are associated with significant distress for both patients and families, and the continuation of an ICD in a patient who is actively dying may prolong suffering without meaningful benefit. The ethical imperative to minimize non-beneficial interventions at end of life, alignment with patient and family values regarding quality of dying, and the principle of patient autonomy through shared decision-making are collectively determinative in this recommendation. Expert consensus within the committee, supported by existing international position statements,⁹⁸⁾ unanimously endorsed that ICD deactivation should be available as a patient-centered option in the end-of-life setting.

PRACTICAL CONSIDERATIONS FOR CLINICAL IMPLEMENTATION

In real-world clinical practice, palliative care for patients with advanced HF may be most effectively delivered through a collaborative co-management model between the HF team and palliative medicine specialists. Palliative care may begin when a patient who is hospitalized in the cardiology ward or followed in the cardiology outpatient clinic is referred to a palliative medicine specialist for consultation. After palliative care is initiated, HF-specific treatment may continue under regular follow-up by the HF specialist at the usual interval according to clinical status (for example, every 1–3 months in the outpatient setting), with ongoing optimization of GDMT and disease monitoring. At the same time, symptom-focused issues—including worsening dyspnea requiring adjustment of diuretics, constipation, insomnia, depression, anxiety, or other distressing symptoms—may be addressed initially through close communication with the palliative care team, with escalation to the HF specialist when disease-specific reassessment is needed (Figure 1).

Figure 1. Practical collaborative care pathway for palliative care integration in advanced HF.

HF = heart failure; KSHF = Korean Society of Heart Failure; GDMT = guideline-directed medical therapy.

The decision-making process for ICD deactivation may vary according to the patient’s existing pattern of cardiovascular care. When a patient is already being followed by the cardiology team at the same institution, discussion and implementation of ICD deactivation may be conducted directly in the cardiology outpatient clinic through shared decision-making involving the patient, family, and treating cardiologist. By contrast, some patients receiving palliative care may not have active cardiology follow-up at the institution providing end-of-life care—for example, a patient with terminal cancer who previously underwent ICD implantation at another hospital and is currently managed only by the palliative care team at the present institution. In such cases, the palliative care team may request consultation from an HF specialist. If, after interdisciplinary review and shared decision-making with the patient and family, ICD deactivation is considered consistent with the patient’s goals of care, the HF specialist may then refer the patient to a cardiologist with expertise in electrophysiology for the formal ICD deactivation process (Figure 2).

Figure 2. Practical decision pathway for ICD deactivation in advanced HF and palliative care settings.

ICD = implantable cardioverter-defibrillator; HF = heart failure; GDMT = guideline-directed medical therapy.

LIMITATIONS AND APPLICABILITY OF INDIRECT EVIDENCE

A limitation of this guideline is that, for several key questions, direct evidence in patients with advanced HF was sparse or unavailable, and some recommendations therefore relied partly on data derived from non-HF populations, particularly patients with advanced cancer or COPD. Such extrapolation should be interpreted cautiously. Although these populations share important palliative care needs with advanced HF—such as refractory dyspnea, anxiety, depression, pain, and high symptom burden—the underlying pathophysiology, illness trajectory, comorbidity profile, and treatment context differ substantially. Patients with advanced HF often have fluctuating hemodynamics, polypharmacy, renal dysfunction, and vulnerability to hypotension, arrhythmia, sedation, and drug–drug interactions, which may alter both the efficacy and safety of symptom-directed interventions. Accordingly, evidence from non-HF populations was used primarily to inform the direction of effect and potential harms rather than to assume direct equivalence of treatment benefit in HF. Recommendations based on indirect evidence were therefore interpreted conservatively and weighed together with available HF-specific observational data, expert consensus, and clinical applicability.

CONSENSUS ON RECOMMENDATION AND GUIDELINE UPDATE PLAN

Consensus on the recommendations was achieved using the RAND-UCLA Appropriateness Method, which incorporates a modified Delphi process to provide a comprehensive prioritization of issues and the level of agreement among participants.⁹⁹⁾ This process ensured that recommendation strength reflected not only the available evidence but also collective expert judgment in areas where empirical data were limited. All developers participated by voting on a scale of 1 (disagree) to 9 (agree) for each recommendation, classifying them as appropriate (median 7–9, no disagreement), inappropriate (median 1–3, no disagreement), or uncertain (median 4–6 or presence of disagreement). A plan was in place to share the results of the first-round voting and conduct a second-round vote with discussion for any recommendations where appropriateness or agreement remained uncertain; however, all recommendations reached an appropriate level of consensus after the first round (Table 14).

Table 14. Summary of key recommendation with the guideline development committee's grading results.

Recommendation	Strength	Evidence	Voting median	Appro-priateness	Agreement level
1. Palliative care is recommended for patients with advanced HF.	Strong	Low	9	Appropriate	Agree
2. Routine oxygen therapy for dyspnea in non-hypoxemic advanced HF is not recommended.	Not recommended	Low	8.5	Appropriate	Agree
3. Diuretic therapy for congestion in advanced HF can be recommended.	Conditional	Very low	9	Appropriate	Agree
4. Inotropic therapy for symptom relief in selected advanced HF patients can be recommended.	Conditional	Low	9	Appropriate	Agree
5. Routine opioid use in advanced HF is not recommended.	Not recommended	Low	8	Appropriate	Agree
6. Routine benzodiazepine use for dyspnea relief in advanced HF is not recommended.	Not recommended	Very low	8	Appropriate	Agree
7. Antidepressants may be considered for clinically meaningful depressive symptoms.	Conditional	Very low	8.5	Appropriate	Agree
8. Non-pharmacologic interventions may be considered to improve QoL and psychological symptoms.	Conditional	Low	8.5	Appropriate	Agree
9. ICD deactivation may be considered to improve quality of death in advanced HF.	Conditional	Very low	9	Appropriate	Agree

HF = heart failure.

During the guideline development process, the literature review revealed a significant lack of high-quality evidence. Therefore, updates are planned approximately every 5 years, incorporating new recommendations or revisions/supplements to existing ones as additional evidence accumulates and high-quality data on new medications and therapies emerge. Key questions for updates will be selected by surveying guideline users for additional needs and prioritizing through expert discussions on newly introduced drugs and treatments. The update process will follow the existing methodology, and evidence for previously developed recommendations will include newly added content published after December 19, 2025.

SUPPLEMENTARY MATERIALS

Supplementary Table 1

Summary of findings table of key question 1

Supplementary Table 2

Evidence to decision framework of key question 1

Supplementary Table 3

Summary of findings table of key question 2

Supplementary Table 4

Evidence to decision framework of key question 2

Supplementary Table 5

Summary of findings table of key question 3

Supplementary Table 6

Evidence to decision framework of key question 3

Supplementary Table 7

Summary of findings table of key question 4

Supplementary Table 8

Evidence to decision framework of key question 4

Supplementary Table 9

Summary of findings table of key question 5

Supplementary Table 10

Evidence to decision framework of key question 5

Supplementary Table 11

Summary of findings table of key question 6

Supplementary Table 12

Evidence to decision framework of key question 6

Supplementary Table 13

Summary of findings table of key question 7

Supplementary Table 14

Evidence to decision framework of key question 7

Supplementary Table 15

Summary of findings table of key question 8

Supplementary Table 16

Evidence to decision framework of key question 8

Supplementary Table 17

Summary of findings table of key question 9

Supplementary Table 18

Evidence to decision framework of key question 9

Supplementary Figure 1

PRISMA 2020 flow diagram of the study selection process across all key questions.

Supplementary Figure 2

Meta-analysis about impact of palliative care on quality of life.

Supplementary Figure 3

Meta-analysis about impact of palliative care on hospitalization.

Supplementary Figure 4

Meta-analysis about impact of palliative care on DNR (do not resuscitate) documentation.

Supplementary Figure 5

Meta-analysis about impact of palliative care on mortality.

Supplementary Figure 6

Meta-analysis about impact of oxygen therapy on exercise duration.

Supplementary Figure 7

Meta-analysis about impact of oxygen therapy on dyspnea based on borg score.

Supplementary Figure 8

Meta-analysis about impact of oxygen therapy on dyspnea based on NRS (numeric rating scale).

Supplementary Figure 9

Meta-analysis about impact of oxygen therapy on quality of life based on Minnesota living with heart failure questionnaire.

Supplementary Figure 10

Meta-analysis comparing impact of continuous versus intermittent diuretics on weight reduction, freedom from congestion and patient’s symptom.

Supplementary Figure 11

Meta-analysis comparing impact of high-dose versus low-dos diuretics on weight reduction and patient’s symptom.

Supplementary Figure 12

Meta-analysis comparing impact of combination of diuretics versus loop diuretics only on decongestion and dyspnea.

Supplementary Figure 13

Meta-analysis about impact of inotropics on mortality and HF hospitalization.

Supplementary Figure 14

Meta-analysis about impact of inotropics on cardiac function and patient’s symptom.

Supplementary Figure 15

Meta-analysis about impact of inotropics on home death rate.

Supplementary Figure 16

Meta-analysis about adverse effect due to inotropics usage.

Supplementary Figure 17

Meta-analysis about impact of opioids on patient’s symptom, performance status and quality of life.

Supplementary Figure 18

Meta-analysis about impact of opioids on quality of sleep, cognitive function and oxygen saturation.

Supplementary Figure 19

Meta-analysis about impact of opioids on physical activity level.

Supplementary Figure 20

Meta-analysis about impact of opioids on readmission.

Supplementary Figure 21

Meta-analysis about impact of opioids on mortality.

Supplementary Figure 22

Meta-analysis about impact of opioids on HF readmission or mortality.

Supplementary Figure 23

Meta-analysis about impact of benzodiazepine on all-cause mortality.

Supplementary Figure 24

Meta-analysis about impact of benzodiazepine on cardiac death.

Supplementary Figure 25

Meta-analysis about impact of benzodiazepine on HF readmission.

Supplementary Figure 26

Meta-analysis about impact of benzodiazepine on emergency department visits.

Supplementary Figure 27

Meta-analysis about Impact of Antidepressant on Depression based on MADRS (Montgomery–Åsberg Depression Rating Scale), HADS (Hospital Anxiety and Depression Scale)-D, ESAS (Edmonton Symptom Assessment System)-depression and CGI (Clinical Global Impression).

Supplementary Figure 28

Meta-analysis about impact of antidepressant on all-cause mortality.

Supplementary Figure 29

Meta-analysis about impact of antidepressant on cardiovascular mortality.

Supplementary Figure 30

Meta-analysis about adverse events due to antidepressant usage.

Supplementary Figure 31

Meta-analysis about impact of antidepressant on hospitalization-related results.

Supplementary Figure 32

Meta-analysis about impact of non-pharmacological intervention for depression on quality of life based on Minnesota living with heart failure questionnaire and FACIT-PAL (functional assessment of chronic illness therapy-palliative care).

Supplementary Figure 33

Meta-analysis about impact of non-pharmacological intervention for depression on anxiety and depression based on HADS (hospital anxiety and depression scale).

Supplementary Figure 34

Meta-analysis about impact of non-pharmacological intervention for depression on readmission, mortality, ACP (advance care planning) and hospice use.

Supplementary Figure 35

Meta-analysis about impact of ICD (implantable cardioverter-defibrillator) deactivation on ICD shocks.