Hospice Use by Cause of Death: A Cohort Study Using Utah Population Database

Abstract

Purpose

Illnesses such as cancer often follow a predictable trajectory of decline, while others, such as Alzheimer’s Disease and Related Dementias (ADRD) and Chronic Obstructive Pulmonary Disease (COPD), follow a more dwindling and protracted decline. Hospice, a form of EOL palliative care for individuals with less than 6 months of remaining life, was originally created for patients with life-limiting cancer but is increasingly used by all patients with all types of life-limiting diagnoses. This study assesses whether hospice use differs by the underlying cause of death and whether current hospice eligibility and practices provide optimal EOL to all causes of death.

Method

Based on data from 17,321 hospice users drawn from the Caregiver Population Science (C-PopS) decedent cohort of the Utah Population Database (UPDB), we used multinomial regression analyses and Kaplan-Meier Survival Curves to model hospice length-of-stay by cause of death, in order to identify patterns of hospice use by patients who have different underlying causes of death.

Major Findings

Non-cancer decedents were more likely to have sub-optimal patterns of hospice care, including minimal use, lasting less than a week, and extended use, lasting more than 6 months. Stroke decedents were the most likely to have minimal-use patterns of hospice, whereas dementia and COPD decedents were most likely to have extended use.

Conclusion

New models of hospice-like EOL care that can accommodate both short-term and long-term palliative care needs may help meet the diverse needs of patients and families facing different EOL trajectories associated with common causes of death.