Palliative Care in Heart Failure: An Updated Systematic Review and Meta-Analysis of Randomised Controlled Trials and Associations With Intervention and Patient Characteristics

Background Systematic reviews show palliative care improves outcomes in patients with heart failure (HF), but prior meta-analyses were not HF-specific, or explored how intervention and patient characteristics influence effectiveness, whilst new randomized controlled trials(RCTs) have been published. Methods We conducted a systematic review and meta-analysis of RCTs exploring PC in HF. PubMed, EMBASE, CENTRAL and CINAHL were searched until January 24 th 2025(PROSPERO ID:CRD42024607104). The primary outcome was hospitalisations. Secondary outcomes included change in quality of life(QoL), assessed by Kansas-City-Cardiomyopathy Questionnaire (KCCQ) and Functional Assessment of Chronic Illness Therapy-Palliative Care(FACIT-Pal) and mental health, assessed by Hospital-Anxiety-Depression-Scale(HADS). Subgroup analyses were conducted based on intervention characteristics (mode and duration), and patient characteristics (gender distribution and HF symptom severity, measured by NYHA class). Results Twenty-one RCTs were identified; thirteen were included in meta-analyses( n = 1919). PC was associated with reduced hospitalisation [log OR -0.6 (95% CI -1.14, −0.07); I 2 = 69%], and improvement in QoL [mean difference(MD) in KCCQ: 3.09 [95% CI 1.43, 4.75], I 2 = 35%] and depression [MD HADS-D: −0.44 (95% CI -0.75 to −0.13)]. No clear differences were observed between intervention modes. Longer interventions (>12 weeks) and interventions targeting patients with advanced HF symptoms (NYHA III/IV ≥70%) were associated with greater improvement in KCCQ and reduction in hospitalisation respectively. No differences in outcomes were observed by gender distribution. Conclusions Palliative care reduces hospitalisations, improves QoL and mental health in HF. Longer interventions and those targeting patients with advanced symptoms were linked to differences in outcomes, whereas mode of intervention and gender did not appear to impact outcomes.