Prognostic Models to Trigger Palliative Care Referrals in Heart Failure: A Scoping Review

Accurate prognostication in heart failure (HF) is challenging, due to its nonlinear course. This course impacts the timing of goals of care conversations and referrals to palliative care. Consequently, prognostication may contribute to misunderstandings that can lead to poor end-of-life experiences for patients. As many hospice and palliative care programs have prognostic referral criteria of 6 months or less, we aimed to identify, compare, and synthesize prognostication models that would inform decision-making for palliative consultation and referrals. A scoping review was undertaken to map and characterize prognostic models that predict 6-month mortality in HF patients. An iterative literature search was conducted using PubMed, Embase, Emcare, Google Scholar, and Ovid MEDLINE, supplemented by hand searching, to identify applicable articles. Twenty studies from cardiology, palliative care, and geriatric medicine disciplines were given priority for inclusion in data charting and synthesis. Most models demonstrated moderate-to-high risk of bias and applicability concerns. The models were developed in diverse clinical settings with mortality timeframes ranging from in-hospital death to 6 months. Four common themes for prognostic indicators emerged across the disciplines: age, functional status, comorbidities/organ involvement, and prognostic awareness. Over time, discipline-specific scores have evolved to become more alike, to better address the unique needs of the advanced HF population. Papers differed widely in patient population, goals of prognostication, and applicability for informing palliative or hospice referrals. Developing more unified, multidisciplinary prognostic approaches could have value in supporting timely goals of care conversations, and palliative care involvement, for HF patients.