Palliative care in the current management of heart failure- a matched case-control, retrospective cohort study in England

Abstract Introduction Heart failure (HF) represents a major global healthcare burden. In patients with ongoing symptoms despite optimal treatment, palliative care (PC) is expected to improve coordination of care and thereby, patient orientated outcomes, but current utilisation patterns are not well-studied. Methods Using linked electronic health records from England (Clinical Practice Research Datalink), of individuals aged ≥18 years with incident HF, we assessed predictors of receiving palliative care following HF diagnosis using Kaplan-Meier estimates and Cox regression models. We matched HF patients who received PC with those who did not (on age at HF diagnosis, sex, region, year of HF diagnosis and history of cancer at HF diagnosis) to compare risks of cardiovascular and all-cause hospitalisation and mortality in follow-up. Results We included 36,358 individuals with HF who did not receive PC during follow-up and 6,991 individuals who did. Higher age at HF diagnosis, and the presence of comorbidities including type 2 diabetes, hypertension, chronic kidney disease and cancer were associated with higher likelihood of receiving PC. 5,911 PC cases and 16,893 controls were included in the matched analysis. PC utilisation varied greatly across the UK. Estimated adjusted hazard ratios for all-cause hospitalisation were higher in people not receiving PC: 1.42 (1.32, 1.53), as were HF hospitalisation: 1.27 (1.19, 1.35), all-cause mortality: 2.66 (2.53, 2.80) and cardiovascular mortality: 2.16 (2.00, 2.33). Interpretation: In individuals with HF, PC has the potential to reduce important patient-orientated outcomes and thereby improve quality of life in the later stages of the disease. Importantly, despite these positive expected outcomes, PC is currently rarely utilised. There are likely to be opportunities to use PC earlier in HF trajectories, but unmeasured confounding or confounding by indication in observational data and the lack of patient-level measures of quality of life, complicate the estimation of potential benefits of PC in HF and underscore the need for future clinical trials.Characteristics by palliative care use