Table 1 Characteristics of included studies (n = 18)

Agren et al. [44] Sweden

RCT Hospital and community based (outpatient clinic and in patients’ homes)

N = 71 dyads (patient & carer) Patients’ Mean Age (SD): 69 (13) Partners’ Mean Age (SD): 67 (12) Sex (patients): M 49 (69.1%), F 22 (30.9%) Sex (partners): M 22 (30.9%), F 49 (69.1%) NYHA Class II 25 (35%) NYHA Class III 39 (55%) NYHA Class IV 7 (10%) No information on LVEF

N = 84 dyads (patient & partner) Patients’ Mean Age (SD): 73 (10) Partners’ Mean Age (SD): 70 (10) Sex (patients): M 68 (80.9%), F 16 (19.1%) Sex (partners): M 16 (80.9%), F 68 (19.1%) NYHA Class II 25 (30%) NYHA Class III 43 (51%) NYHA Class IV 16 (19%) No information on LVEF

Intervention

HF nurse-led face to face counselling, a computer based CD-ROM program and written teaching materials. 3 sessions in home or clinic: each included education on CHF, development of problem-solving skills in recognising factors contributing to psychological and emotional distress and focused on changing thoughts and behaviours and implementing strategies for self-care. Opportunity for dyads to receive information, raise questions, discuss difficulties and subjects of joy, deal with emotional and practical support and self-care. Providers: HF nurses

Control

Usual care, including traditional care in hospital and outpatient education and support

1. Patient perceived control (CAS)

2. Partners’ perceived control (CAS)

3. Patients’ Physical component score (SF-36)

4. Partners’ Physical component score (SF-36)

5. Patients’ Mental component score (SF-36)

6. Partners’ Mental component score (SF-36)

7. Patients’ Depression (Beck Depression Inventory)

8. Partners’ Depression (Beck Depression Inventory)

9. Patients’ self-care behaviour (EHFscBS)

10. Partners’ caregiver burden (CBS)

1. Statistically significant improvement in perceived control at 3 months in intervention group compared to control, but not at 12 months (p = 0.03).

2. No significant difference in either group.

3. No significant difference in either group.

4. No significant difference in either group.

5. No significant difference in either group.

6. No significant difference in either group.

7. No significant difference in either group.

8. No significant difference in either group.

9. No significant difference in either group.

10. No significant difference in either group.

Aiken et al. [35] USA

Phase III RCT, mixed population study Hospice of the Valley, Phoenix, Arizona (community based)

N = 100 (patients with CHF = 67) Mean Age of total population, no CHF subset analysis (SD) = 68 (14) Sex of total population, no CHF subset analysis: M 42%, F 58% All NYHA III or IV No information on LVEF

N = 90 (patients with CHF = 62) Mean Age of total population, no CHF subset analysis (SD) = 70 (13) Sex of total population, no CHF subset analysis: M 30%, F 70% All NYHA III or IV No information on LVEF

Intervention

Intensive home-based case management. Program foci included disease and symptom management, patient self-management and knowledge of illness-related resources, preparation for end of life, physical and mental functioning and utilisation of medical services Providers: registered nurse case-managers, in coordination with primary care physician, community agencies, medical director, social worker & pastoral counsellor.

Control

Usual care provided by Managed Care Organisations (MCOs). Focus on medical and disease oriented approach, including medication and lab monitoring. Services delivered via telephone. Support services included disease and symptom education, nutrition and psychological counselling, transportation and co-ordination of medical services

1. Self-management of illness and knowledge of resources (self-developed questionnaire)

2. Preparation for end of life (self-developed questionnaire)

3. Participation in enjoyable activities (yes/no question)

4. Symptom control (Memorial symptom assessment scale)

5. Trajectories of physical and mental functioning (SF-36)

6. Medical utilisation (Emergency department visits and length of stay in hospital)

1. No CHF subset analysis. PhoenixCare participants had statistically significant sense of receiving sufficient information to handle illness emergency at 6 months (p < 0.05), sense of receiving education about community resources at 3 months (p < 0.05), information about who to talk to about medical problems at 6 months (p < 0.05) and better preparedness for daily experiences at 3 months (effect reversed by 6 months) (p < 0.05).

2. No CHF subset analysis. PhoenixCare participants had a statistically higher rate of completion of living will/advance directive after 3 months (p < 0.05). No significant difference at 6 months.

3. No significant difference in CHF participants.

4. No significant change in frequency of symptoms in CHF participants in PhoenixCare group. Significantly higher mean distress at 6 months in CHF participants in PhoenixCare compared to control (p < 0.05). 5. Stable physical scores in intervention group, while controls deteriorated at 9 months (p < 0.05). No difference in mental health scores.

6. No significant difference between groups.

Bekel-man et al. [36] USA

Multisite RCT Community based with outpatient consultations

N = 187 Mean Age (SD) = 68.3 (9.6) Sex: M 178 (95%), F 9 (5%) NYHA Class I 16 (8.9%) NYHA Class II 77 (42.8%) NYHA Class III 82 (45.6%) NYHA Class IV 5 (2.8%) LVEF: normal (78, 45.6%), 40–49% (34, 19.9%), 30–39% (46, 26.9%), < 30 (13, 7.6%)

N = 197 Mean Age (SD) = 67.9 (10.6) Sex: M 193 (98%), F 4 (2%) NYHA Class I 16 (8.5%) NYHA Class II 85 (45%) NYHA Class III 82 (43.4%) NYHA Class IV 6 (3.2%) LVEF: normal (84, 47.5%), 40–49% (34, 19.2%), 30–39% (32, 18.1%), < 30 (27, 15.3%)

Intervention

1) Multidisciplinary collaborative care: team met weekly to recommend changes based on telemonitoring data and PHQ-9 scores

2) Screening for and treatment of depression - those who screened positive received up to 11 sessions of behavioural activation and antidepressant management, a depression educational video and depression assessment & self-management education via telemonitoring

3) Daily telemonitoring of clinical observations, weight and self-reported symptoms & self-care program such as medication reminders, education about HF & dietary advice Providers: Nurse coordinator, primary care physician, cardiologist & psychiatrist

Control

Usual care provided by their regular health professionals, including telemonitoring if enrolled. Information sheets for self-care given and if screened positive for depression at baseline, primary care physicians were notified.

1. Patient-reported HF-specific Health Status (KCCQ)

2. Deaths

3. Depression (PHQ-9 score)

4. Hospitalisation rates

1. Statistically significant improvement in KCCQ overall summary scores in both groups after 1 year, with no significant difference between groups.

2. Fewer deaths in intervention group compared to control (p = 0.04).

3. Greater improvement in PHQ-9 score after 1 year in the intervention arm than usual arm if screened positive for depression (p = 0.01).

4. No significant difference in hospitalisations between intervention group and control group.

Bekel-man et al. [37] USA

Multisite RCT Outpatient consultations

N = 157 Mean Age (SD) = 64.5 (10.9) Sex: M 128 (81.5%), F 29 (18.5%) NYHA Class I 5 (3.2%) NYHA Class II 55 (35%) NYHA Class III 75 (47.8%) NYHA Class IV 20 (12.7%) LVEF: Normal (57/150, 38%), Mild (28/150, 18.7%), Moderate (30/150, 20%), Severe (35/150, 23.3%)

N = 157 Mean Age (SD) = 66.5 (11.8) Sex: M 119 (758%), F 38 (24.2%) NYHA Class I 12 (7.6%) NYHA Class II 46 (29.3%) NYHA Class III 75 (47.8%) NYHA Class IV 24 (15.3%) LVEF: Normal (64/149, 43%), Mild (18/149, 12.1%), Moderate (22/149, 14.8%), Severe (45/149, 30.2%)

Intervention

1) Registered nurse reviewed/assessed symptoms, applied motivational interviewing to promote changes in health behaviours, follow up telephone assessments 1-2x per month

2) Social worker provided structured telephone-based psychosocial intervention and address depression symptoms if present. Provided support to informal caregivers as needed

3) Weekly collaborative care team meetings Providers: Registered nurse, social worker, primary care clinician, cardiologist and palliative care (PC) physician

Control

Care at the discretion of their clinicians (could include care from cardiology, palliative care and mental health). Information sheet outlining self-care for HF. If screened positive for significant depression, patients and their clinicians were contacted.

1. Patient-reported HF-specific Health Status (KCCQ)

2. Depressive symptoms (PHQ-9)

3. Anxiety symptoms (GAD-7)

4. Overall symptom distress (GSDS score)

5. Pain (PEG score)

6. Fatigue (PROMIS score)

7. Dyspnoea (Dyspnoea score)

1. No statistically significant difference between groups.

2. Statistically significant improvement in depressive symptoms in intervention group at 6 months (p = 0.02)

3. Statistically significant difference in anxiety symptoms at 3 months in intervention group (p < 0.01), but no difference at 6 months.

4. No significant difference between groups.

5. No significant difference between groups.

6. Statistically significant improvement in fatigue scores in intervention group compared to control at 6 months (p = 0.02).

7. No significant difference between groups.

Brann-strom et al. [45] Sweden

RCT Community based with outpatient consultations

N = 36 Mean age (SD) = 81.9 (7.2) Sex: M 26 (72.2%), F 10 (27.8%) NYHA Class III 28 (77.8%) NYHA Class IV 8 (22.2%) LVEF: 40–49 13 (36.1%), 30–39 16 (44.4%), < 30 7 (19.4%)

N = 36 Mean age (SD) = 76.6 (10.2) Sex: M 25 (69.4%) F 11 (30.6%) NYHA Class III 23 (63.9%) NYHA Class IV 11 (30.6%) Unknown NYHA Class 2 (5.5%) LVEF: 40–49 12 (33.3%), 30–39 21 (58.3%), < 30 3 (8.3%)

Intervention

PREFER intervention Multidisciplinary approach involving collaboration between specialists in palliative and heart failure care. Appointment with nurses for initial assessment followed by using a model for person-centred palliative care based on patient goals. Provision of total care including assessment of symptoms, quality of life, and risk, and registration into HF and palliative care registry. Subsequent access to home care visits by MDT. Providers: Specialist HF nurses, PC nurses, cardiologist, PC physician, physiotherapist & OT

Control

Usual care provided by GPs or doctors/nurses at HF clinic.

1. Health-related quality of life (EQ-5D)

2. Symptom burden (ESAS)

3. Disease-specific quality of life (KCCQ)

4. Functional classes (NYHA score)

5. Number of hospitalisations

6. Mortality

7. Resource utilisation (utilisation of visits, phone calls and prescriptions of clinicians

1. Age-adjusted health-related QoL significantly improved in intervention group compared to control (p = 0.02).

2. No significant difference between groups in overall score. Within group analysis of intervention group found significant improvement in nausea (p = 0.02).

3. No significant difference between groups.

4. More patients in the intervention group experienced an improvement in functional class (11 of 28; (p = 0.015).

5. Mean number of hospitalisations significantly lower in intervention group compared to control (p = 0.009) & mean number of days spent in hospital significantly lower in the intervention group (p = 0.011).

6. No significant difference between groups.

7. Significantly more nurse visits in intervention group compared to usual care (1075 vs 230;, p = 0.000).

More phone calls and prescriptions by doctors in usual care group (108 vs 230; no p value given).

Ekman et al. [46] Sweden

Randomised control feasibility study Outpatient consultations

N = 79 No separate group breakdown for age or sex Mean age across groups (SD) = 80.3 (6.8) Sex (both groups): M 91 (58%), F 67 (42%) Mean NYHA class = 3.2 ± 0.5 Mean LVEF = 0.43 ± 0.18

N = 79 No separate group breakdown for age or sex Mean age across groups (SD) = 80.3 (6.8) Sex (both groups): M 91 (58%), F 67 (42%) Mean NYHA class = 3.2 ± 0.5 Mean LVEF = 0.38 ± 0.15

Intervention

Nurse-monitored outpatient clinic run in co-operation with study doctors. Focus on patient’s understanding/ knowledge of condition and management, symptom control, psychosocial and lifestyle factors. Care was individually planned and goal-centered. Practical guidelines developed for nursing staff on symptoms, nutritional status, diet and ability of the patient to participate in self-care. Subsequent regular follow up telephone call contact. Providers: HF nurses and study doctors

Control

Usual care typically from GP and attendance at the Emergency Department.

1. Follow up time (months)

2. Functional status (NYHA class)

3. Hospital re-admissions

4. Mortality

5. Survival without readmission

1. No significant difference between groups.

2. Statistically significant improvement in NYHA class seen in both groups (Control p < 0.001, Intervention p < 0.05).

3. No significant difference between groups.

4. No significant difference between groups.

5. No significant difference between groups.

Hopp et al. [38] USA

Pilot RCT Three urban hospital sites

N = 43 Mean Age (SD) = 67.0 (11.0) Sex: M 26 (60.5%); F 17 (39.5%) No information on NYHA class Mean LVEF = 36.4 (16.7)

N = 42 Mean Age (SD) = 68.0 (13.0) Sex: M 18 (42.9%); F 24 (57.1%) No information on NYHA class Mean LVEF = 38.1 (16.8)

Intervention

Palliative care consultation in hospital. Clinical interviews assessed for uncontrolled symptoms, goals of care, code status, ACP and desired post-treatment residential setting. Option to refer to other specialties (e.g. chaplaincy or social work) if indicated. Providers: Palliative care physician or advanced nurse practitioner

Control

No information

1. Election of “comfort care”:

a outpatient hospice

b inpatient hospice

c DNR order during index or subsequent hospital admission

d DNR order at home or nursing home

2. Mortality

1. 4 of 43 patients in intervention group had met primary end point, compared with 0 of 42 patients in control group; no statistically significant difference between groups.

2. No significant difference between groups.

Liljeroos et al. [47] Sweden

*Note: longer term follow up of Agren et al.* RCT Hospital and community based (outpatient clinic and in patients’ homes)

N = 71 dyads (patient & carer) Patients’ Mean Age (SD): 69 (13) Partners’ Mean Age (SD): 67 (12) Sex (patients): M 49 (69.1%), F 22 (30.9%) Sex (partners): M 22 (30.9%), F 49 (69.1%) NYHA Class II 25 (35%) NYHA Class III 39 (55%) NYHA Class IV 7 (10%) No information on LVEF

N = 84 dyads (patient & partner) Patients’ Mean Age (SD): 73 (10) Partners’ Mean Age (SD): 70 (10) Sex (patients): M 68 (80.9%), F 16 (19.1%) Sex (partners): M 16 (80.9%), F 68 (19.1%) NYHA Class II 25 (30%) NYHA Class III 43 (51%) NYHA Class IV 16 (19%) No information on LVEF

Intervention

HF nurse-led face to face counselling, a computer based CD-ROM program and written teaching materials. 3 sessions in home or clinic: each included education on CHF, development of problem-solving skills in recognising factors contributing to psychological and emotional distress and focused on changing thoughts and behaviours and implementing strategies for self-care. Opportunity for dyads to receive information, raise questions, discuss difficulties and subjects of joy, deal with emotional and practical support and self-care. Providers: HF nurses

Control

Usual care, including traditional care in hospital and outpatient education and support

1. Time to first readmission/ death (days)

2. Number of readmissions

3. Number of days in hospital

4. Patients’ and partners’ physical component scores (SF-36)

5. Patients’ and partners’ mental component scores (SF-36)

6. Patients’ and partners’ physical functioning (unclear measurement tool)

7. Role limitations due to physical health problems

8. Bodily pain (unclear tool)

9. General health (unclear measurement tool)

1. Vitality (unclear tool)

2. Social functioning

3. Role limitations due to emotional problems

4. Depression (Beck depression inventory)

5. Perceived control (CAS)

1. No significant difference between groups.

2. No significant difference between groups.

3. No significant difference between groups.

4. No significant difference in patients’ PCS between groups. Significantly greater reduction in partners’ PCS in intervention group (p < 0.05).

5. No significant difference between groups.

6. No significant difference in patient’s functioning between groups. Significantly greater reduction in partners’ functioning in intervention group (p < 0.05).

7. No significant difference between groups.

8. No significant difference between groups.

1. No significant difference between groups.

2. No significant difference between groups.

3. No significant difference between groups.

4. No significant difference between groups.

5. No significant difference between groups.

6. No significant difference between groups.

Mentz et al. [39] USA

*Note: same population/ intervention as Rogers et al.* RCT Outpatient consultations

N = 75 Mean Age (SD) = 71.9 (12.4) Sex: M 42 (56%); F 33 (44%) NYHA Class III 54 (72%) NYHA Class IV 15 (20%) LVEF: Normal > 55% 21 (28%), 40–55% 14 (18.7%), 25–40% 17 (22.7%), < 25% 23 (30.7%)

N = 75 Mean Age (SD) = 69.8 (13.4) Sex: M 37 (49.3%); F 38 (50.7%) NYHA Class III 58 (77.3%) NYHA Class IV 5 (6.7%) LVEF: Normal > 55% 14 (18.7%), 40–55% 19 (25.3%), 25–40% 14 (18.7%), < 25% 28 (37.3%)

Intervention

PAL-HF: interdisciplinary, multicomponent palliative care intervention. PC nurse practitioner coordinated patient’s care in collaboration with palliative care physician and the cardiology team. Pal-HF nurse participated in ongoing management in outpatient setting on hospital discharge. Focus on physical symptoms, psychosocial and spiritual concerns, ACP and shared goal-setting. Providers: PC nurse, PC physician, cardiology nurse

Control

Usual care. Cardiologist-directed team. Had access to inpatient PC if referred. Outpatient follow up with their general practitioners and HF cardiologist or nurse practitioner.

1. Total hospital admissions

2. Days alive and out of hospital

3. Admissions due to HF

4. Non cardiovascular-related admissions

1. No significant difference between groups.

2. No significant difference between groups.

3. No significant difference between groups.

4. No significant difference between groups.

Ng et al. [49] China

*Note: same population/ intervention as Wong et al. (2018)* RCT Community setting

N = 43 Mean Age (SD) = 78.3 (16.8) Sex: M 18 (43.9); F 25 (56.1%) NYHA class II = 6 (14.0%) NYHA class III = 31 (72.1%) NYHA class IV = 6 (14.0%) Mean LVEF = 39.0 (14.0)

N = 41 Mean Age (SD) = 78.4 (10.0) Sex: M 25 (61.0%); F 16 (39.0%) NYHA class II = 3 (7.3%) NYHA class III = 22 (53.7%) NYHA class IV = 16 (39.0%) Mean LVEF = 37.0 (17.0)

Intervention

Post-hospital discharge home visits and telephone calls by palliative care nurse case managers and supported by trained volunteers (nursing students). Key components were physical and psychological symptom assessment and management, social support, spiritual and existential aspects of care, setting goals of care and discussion of treatment preferences and end-of-life issues. PC-nurse could refer to PC physician and other services as necessary. Providers: Palliative care nurse case managers, trained volunteers (nursing students), palliative care physician

Control

Pre-discharge PC referral consultation and scheduled outpatient PC clinic. Unstructured episodic home care service arranged if needed. 2 social calls from ‘assistant’ consisting of light conversation. Unclear who service delivered by.

1. Quality of life (McGill QoL Questionnaire-Hong Kong)

2. HF-specific quality of life (CHQ-C)

3. Symptom burden (ESAS)

4. Functional status (Palliative Performance Scale)

5. Patient satisfaction

6. Caregiver burden (ZBI)

1. Statistically significant improvement in QoL in intervention group compared to control (p = 0.016), with significant improvement in specific domains: physical (p = 0.011), psychological (p = 0.04), and existential (p = 0.027).

2. Statistically significant improvement in intervention group compared to control at 4 weeks (p = 0.01), but not at 12 weeks. Subset analysis of domains showed statistically significant difference in dyspnoea (p = 0.02), emotional function (p = 0.014) and mastery (p = 0.019) at 4 weeks, but not at 12 weeks.

3. No significant difference in overall score between groups. Subset analysis of domains showed statistically significant difference in depression (p = 0.01) and shortness of breath (p = 0.018) scores at 4 weeks in intervention group compared to control, but not at 12 weeks.

4. No significant difference between groups.

5. Statistically significant higher patient satisfaction score in intervention group compared to control group at 12 weeks (p = 0.001).

6. Statistically significant lower median score after 12 weeks in intervention group (p = 0.024).

O’Don-nell et al. [40] USA

RCT Hospital and community based

N = 26 Mean Age (SD) = 74.7 (11.2) Sex: M 14 (53.9%); F 12 (46.1%) No information on NYHA class Mean LVEF (SD) = 30 (14)

N = 24 Mean Age (SD) = 69.2 (10.2) Sex: M 15 (62.5%); F 9 (37.5%) No information on NYHA class Mean LVEF (SD) = 36 (17)

Intervention

Initial palliative care SW consultation with assessment of prognostic understanding and preferences; a discussion of prognosis and its implications for decision making; an exploration of key topics including patient goals and priorities, fears and worries, sources of strength, perceived critical abilities, trade-offs between length and quality of life, and family awareness of preferences and prognosis; and summaryimpressions and recommendations. Education provided on importance of ACP, the role of health care proxy, and the need for other legal documents. Review by palliative care physician who facilitated further discussions and directed specific interventions. Telephone contact follow up from SW for next 6/12. Ongoing assessment of physical and psychological symptoms during course of follow up. Providers: Palliative care SW, palliative care physician

Control

Usual in-hospital and post-discharge treatment as directed by care team. Printed materials containing ACP information.

1. Documentation of advanced care preferences

2. Improvement in prognostic alignment (revision of patient expectations of prognosis in a direction consistent with those of the treating physician)

3. Physician-level ACP conversations (Presence of Medical Order for Life-sustaining treatment, advanced directive form or hospice referral)

4. Spiritual wellbeing (Functional Assessment of Chronic Illness Therapy-Spiritual Well-being)

5. QOL (abbreviated KCCQ)

6. Depression (Personal Health Questionnaire Depression Scale)

7. Anxiety (GAD-7)

1. Statistically significant increased proportion of patients had documented ACP in intervention group compared to control (p = 0.02).

2. Significantly more surviving patients in intervention group compared with control group had revised their baseline prognostic assessment in a direction consistent with the physician’s assessment (p < 0.001).

3. No significant difference between groups.

4. No significant difference between groups.

5. No significant difference between groups.

6. No significant difference between groups.

7. No significant difference between groups.

O’Rior-dan et al. [41] USA

Pilot RCT Hospital and outpatient based

N = 16 Mean Age (SD) = 71 (18) Sex: M 5 (31%); F 11 (69%) No separate group breakdown of NYHA class Total NYHA class: I 7 (23%), II 9 (30%), III 11 (37%), IV 3 (10%) No information on LVEF

N = 14 Mean Age (SD) = 59 (19) Sex: M 10 (62%); F 4 (28%) No separate group breakdown of NYHA class Total NYHA class: I 7 (23%), II 9 (30%), III 11 (37%), IV 3 (10%) No information on LVEF

Intervention

SMS-HF: intensive palliative care consultations provided by an interdisciplinary palliative care team. Care included prescribing medications for symptoms, discussing ACP, completing documentation and providing psychosocial and spiritual support. First consultation occurred during hospitalisation.

Providers: PC nurse practitioner, PC physician, social worker and chaplain

Control

Standard cardiology care

1. Heart-related quality of life (MLHFQ)

2. Functional status (FACIT-PAL)

3. Depression (HADS)

4. Anxiety (HADS)

5. Pain severity (BPI)

6. Dyspnoea (Borg scale)

7. Guideline-driven HF therapies

8. ACP documentation

9. Satisfaction with care

1. No significant difference between groups.

2. No significant difference between groups.

3. No significant difference between groups.

4. No significant difference between groups.

5. No significant difference between groups.

6. No significant difference between groups.

7. No significant difference between groups.

8. No significant difference between groups.

9. No significant difference between groups.

Paes [52] UK

Pilot RCT Outpatient consultations

N = 6 Mean Age (SD) = 73.2 (4.2) Sex: M 6 (100.0%); F 0 (0.0%) NYHA class III 3 (50.0%) NYHA class IV 3 (50%) No information on LVEF

N = 5 Mean Age (SD) = 78.0 (7.0) Sex: M 4 (80.0%); F 1 (20.0%) NYHA class III 3 (60.0%) NYHA class IV 2 (40%) No information on LVEF

Intervention

Initial one hour outpatient consultation. Aim was to address: background, problematic physical and psychological symptoms, patient knowledge regarding their condition and prognosis, planning for the future, drug compliance, any issues raised by patient. Followed-up by monthly 30 min consultations for 5 months. Providers: Palliative care doctor

Control

Usual care provided by their consultant physician

1. Anxiety (HADS)

2. Depression (HADS)

3. Global health status and functional scale (EORTC QLQ-C30)

4. Symptom/problem scales (EORTC QLQ-C30)

5. Disease-specific QOL (KCCQ)

6. Disease-specific physical functioning (KCCQ)

7. Disease-specific symptom stability (KCCQ)

8. Social limitation (KCCQ)

9. Disease-specific functional status (KCCQ)

1. No significant difference between groups.

2. No significant difference between groups.

3. No significant difference between groups in any domains.

4. No significant difference between groups in any domains.

5. No significant difference between groups.

6. No significant difference between groups.

7. No significant difference between groups.

8. No significant difference between groups.

9. No significant difference between groups.

Piam-jariyakul et al. [42] USA

Pilot mixed methods study with randomised assignment of dyads to intervention or standard care Outpatient consultations

N = 10 dyads Patients’ Mean Age (SD) = 60.8 (14.5) Caregivers’ Mean Age (SD) = 65.1 (8.0) Sex (patients): M 6 (60%); F 4 (40%) Sex (caregivers): M 2 (20%); F 8 (80%) No information on NYHA class or LVEF

N = 10 dyads Patients’ Mean Age (SD) = 63.7 (13.1) Caregivers’ Mean Age (SD) = 57.3 (10.9) Sex (patients): M 6 (60%); F 4 (40%) Sex (caregivers): M 1 (10%); F 9 (90%) No information on NYHA class or LVEF

Intervention

Post-hospital telephone coaching to caregivers on specific HF home care skills including symptom monitoring, medication adherence, sodium/fluid restrictions. Materials sent by mail: 1) Caregivers guide; 2) list of local support organisations; 3) national award winning book on a guide for caregivers in HF; 4) low sodium booklet; 5) pill organiser. 4 weekly coaching sessions included symptom control assessment, support for family (and SW referral if indicated), & content related to end-of-life care including advanced directive forms. Providers: HF nurse specialist

Control

Standard HF care provided by their physician and cardiologist including education materials routinely given to HF patients.

1. Patients’ HF-related hospitalisation frequency

2. Caregivers’ confidence in providing HF homecare (4-item Likert scale)

3. Caregivers’ preparedness in providing HF homecare (1-item Likert scale)

4. Perceived social support (1 item Likert scale)

5. Caregiver burden (17-item Likert scale)

6. Caregiver depression (CES-D)

1. Intervention group had a significant reduction in re-hospitalisations compared to control group (p = 0.03).

2. Significantly improved caregiver confidence in intervention group (p = 0.003).

3. No significant difference between groups.

4. Statistically significant greater perceived social support in intervention group compared to control (p = 0.01).

5. No significant difference between groups.

6. Statistically significant improved scores in intervention group compared to control (p = 0.01).

Rogers et al. [43] USA

RCT Outpatient consultations

N = 75 Mean Age (SD) = 71.9 (12.4) Sex: M 42 (56%); F 33 (44%) NYHA Class III 54 (72%) NYHA Class IV 15 (20%) LVEF: Normal > 55% 21 (28%), 40–55% 14 (18.7%), 25–40% 17 (22.7%), < 25% 23 (30.7%)

N = 75 Mean Age (SD) = 69.8 (13.4) Sex: M 37 (49.3%); F 38 (50.7%) NYHA Class III 58 (77.3%) NYHA Class IV 5 (6.7%) LVEF: Normal > 55% 14 (18.7%), 40–55% 19 (25.3%), 25–40% 14 (18.7%), < 25% 28 (37.3%)

Intervention

PAL-HF: interdisciplinary, multicomponent palliative care intervention. Palliative care nurse practitioner coordinated patient’s care in collaboration with palliative care physician and the cardiology team. Pal-HF nurse participated in ongoing management in outpatient setting on hospital discharge. Focus on physical symptoms, psychosocial and spiritual concerns, ACP and shared goal-setting.

Providers: PC nurse, PC physician, cardiology nurse

Control

Cardiologist-directed team. Had access to inpatient palliative care if referred. Outpatient follow up with their general practitioners and HF cardiologist or nurse practitioner.

1. HF-specific QOL (KCCQ)

2. General and palliative care-specific, health related QOL (FACIT-PAL)

3. Spiritual wellbeing (FACIT-Sp)

4. Depressive symptoms (HADS)

5. Anxiety symptoms (HADS)

6. Mortality

7. Hospitalisations

1. Statistically significant incremental improvement in scores in intervention group compared to control at 6 months (p = 0.030).

2. Statistically significant incremental improvement in scores in intervention group compared to control at 6 months (p = 0.035).

3. Statistically significant improvement in scores in intervention group (p = 0.027).

4. Statistically significant improvement in scores in intervention group (p = 0.020).

5. Statistically significant improvement in intervention group compared to control (p = 0.048)

6. No significant difference between groups.

7. No significant difference between groups.

Sahlen et al. [48] Sweden

*Note: substudy of Brannstrom et al.* RCT Community based with outpatient consultations

N = 36 Mean age (SD) = 81.9 (7.2) Sex: M 26 (72.2%), F 10 (27.8%) NYHA Class III 28 (77.8%) NYHA Class IV 8 (22.2%) LVEF: 40–49 13 (36.1%), 30–39 16 (44.4%), < 30 7 (19.4%)

N = 36 Mean age (SD) = 76.6 (10.2) Sex: M 25 (69.4%) F 11 (30.6%) NYHA Class III 23 (63.9%) NYHA Class IV 11 (30.6%) Unknown NYHA Class 2 (5.5%) LVEF: 40–49 12 (33.3%), 30–39 21 (58.3%), < 30 3 (8.3%)

Intervention

PREFER intervention: MDT approach involving collaboration between specialists in palliative and HF care. Appointment with nurses for initial assessment followed by using a model for person-centred palliative care based on patient goals. Provision of total care including assessment of symptoms, QoL, and risk, and registration into HF and palliative care registry. Subsequent access to home care visits.Providers: Specialist HF nurses, PC nurses, cardiologist, PC physician, physiotherapist & OT

Control

Usual care provided by GPs or doctors/nurses at HF clinic.

1. Quality Adjusted Life Years (QALYs)

2. Costs of care (Euros €)

1. Intervention group had a slight improvement in QALYs and control group had a decline. These changes were small but significant (p = 0.026). Net gain of 0.25 QALYs with intervention.

2. The intervention reduced costs by €61,000 over intervention period (due to reduced hospital admissions).

Wong et al. [50] China

RCT Community setting

N = 43 Mean Age (SD) = 78.3 (16.8) Sex: M 18 (43.9); F 25 (56.1%) NYHA class II = 6 (14.0%) NYHA class III = 31 (72.1%) NYHA class IV = 6 (14.0%) Mean LVEF = 39.0 (14.0)

N = 41 Mean Age (SD) = 78.4 (10.0) Sex: M 25 (61.0%); F 16 (39.0%) NYHA class II = 3 (7.3%) NYHA class III = 22 (53.7%) NYHA class IV = 16 (39.0%) Mean LVEF = 37.0 (17.0)

Intervention

Post-hospital discharge home visits and telephone calls by PC nurse case managers and supported by trained volunteers (nursing students). Key components were physical and psychological symptom assessment and management, social support, spiritual and existential aspects of care, setting goals of care and discussion of treatment preferences and end-of-life issues. PC-nurse could refer to PC physician and other services as necessary.

Providers: PC nurse case managers, trained volunteers (nursing students), PC physician

Control

Pre-discharge PC referral consultation and scheduled outpatient PC clinic. Unstructured episodic home care service arranged if needed. 2 social calls from ‘assistant’ consisting of light conversation. Unclear who service delivered by.

1. Hospital readmissions

2. Symptom intensity (ESAS)

3. Functional status (Palliative Performance Scale)

4. Quality of life (MQOL-HK)

5. HF-specific quality of life (CHQ-C)

6. Satisfaction with care

1. Statistically significant reduction in readmissions in intervention group at 12 weeks (p 0.001).

2. Significantly higher clinical improvement in overall symptom intensity in intervention group compared with control group (p < 0.05). Domain analysis also showed significant difference in minimally clinically important difference of depression scores (p < 0.05) and dyspnoea scores (p < 0.05).

3. No significant difference between groups.

4. Statistically significant improvement in intervention group compared to control (p < 0.05).

5. Statistically significant improvement in intervention group compared to control (p < 0.01).

6. Statistically significant higher satisfaction with care in intervention group (p < 0.001).

Wong et al. [51] China

RCT Community setting

N = 43 Mean Age (SD) = 78.3 (16.8) Sex: M 18 (43.9); F 25 (56.1%) NYHA class II = 6 (14.0%) NYHA class III = 31 (72.1%) NYHA class IV = 6 (14.0%) Mean LVEF = 39.0 (14.0)

N = 41 Mean Age (SD) = 78.4 (10.0) Sex: M 25 (61.0%); F 16 (39.0%) NYHA class II = 3 (7.3%) NYHA class III = 22 (53.7%) NYHA class IV = 16 (39.0%) Mean LVEF = 37.0 (17.0)

Intervention

Post-hospital discharge home visits and telephone calls by PC nurse case managers and supported by trained volunteers (nursing students). Key components were physical and psychological symptom assessment and management, social support, spiritual and existential aspects of care, setting goals of care and discussion of treatment preferences and end-of-life issues. PC-nurse could refer to PC physician and other services as necessary.

Providers: PC nurse case managers, trained volunteers (nursing students), PC physician

Control

Pre-discharge PC referral consultation and scheduled outpatient PC clinic. Unstructured episodic home care service arranged if needed. 2 social calls from ‘assistant’ consisting of light conversation. Unclear who service delivered by.

1. Healthcare costs (Hong Kong dollar $)

2. Quality adjusted life years (QALYs)

1. Cost per case for intervention group was HK$7848/HK$10123 (28/84 days) compared with HK$15,783/HK$36,206 (28/84 days) in control group. Net incremental costs for intervention group was -HK$7935/−HK$26,084.

2. No statistically significant difference between groups.