Palliative care in heart failure: extent and predictors of use, and association with cause-specific hospitalisation and mortality- a Swedish retrospective cohort analysis

Abstract Introduction Heart failure (HF) is common and associated with high mortality and high healthcare utilisation, particularly at the end of life. Palliative care (PC) may improve quality of care and other outcomes, but is under-used and under-researched in the context of HF. Aim To investigate the scale of use of PC, the factors predicting its use and associations with cause-specific hospitalisation and mortality. Methods In the Swedish Heart Failure Registry (SwedeHF) linked with national Swedish registries (SHFDB version 4.2.1), we studied individuals ≥18 years of age with HF from 1 January 2000 to 31st December 2023. Association between PC and selected baseline characteristics was evaluated using univariable and multivariable logistic regression models. Death (Overall and CV), hospitalisation (total, first, CV and HF) and composite of CV death or First HF hospitalization (CV death/1st HFH) were investigated using Kaplan-Meier and Cox regression models. Results Of 55876 individuals with HF, 336 (0.6%) received PC. Individuals receiving PC were more likely to be female (43% vs 34%; p=0.001), age80 years (62% vs 30%; p0.001), inpatient (40% vs 11%; p0.001), followed up in primary care and have HF for 6 months (73% vs 65%; p=0.003) than individuals not receiving PC, and were more likely to have ischaemic heart disease, stroke, atrial fibrillation, chronic obstructive pulmonary disease, anaemia, valvular heart disease, liver disease and comorbidities (Charlson comorbidity index 43-6 vs 21-4; p0.001). 88% of individuals receiving PC had NYHA ≥3 symptoms and 40% had HF with reduced left ventricular ejection fraction. Factors associated with PC included history of ischaemic heart disease, anaemia and malignant cancer in the last 3 years, whereas more recent HF diagnosis (e.g. 2021-2023 vs 2017-2020) and being followed up in hospital vs primary care were associated with reduced PC. Individuals receiving PC had high mortality (124108-142 per 100 person-years), total HF hospitalisation (6957-83) and total hospitalisation (192171-214), especially when compared with those without PC (e.g. HR 8.997.83-10.32 for mortality, HR 4.112.85-5.94 for total hospitalisation (Figure 1). Conclusions PC is rarely utilised in current routine HF care in Sweden. High rates of comorbidity, mortality and hospitalisation in individuals currently receiving PC show gaps in research and practice for increased and earlier use of PC in individuals with HF.Figure 1.Outcomes by palliative care