Factors associated with dying in hospital awaiting palliative care admission: a retrospective cohort study

Palliative care facilities, including palliative care units (PCUs), remain underutilized with individuals dying in hospital despite having applied for admission. Historically, lower proportions of individuals with non-cancer diagnoses versus cancer diagnoses access these facilities. The study was conducted to determine the sociodemographic, functional, and clinical factors associated with dying in hospital awaiting PCU admission.

This was a retrospective cohort study using routinely collected data from a clinical database supplemented with data from a chart review in a randomly selected subgroup. The cohort included individuals hospitalized at a single hospital in Toronto, Canada awaiting PCU admission between January and December 2023. Logistic regression was used to determine factors associated with dying in hospital awaiting PCU admission.

In the total cohort (N = 426; median age = 78 years, 52% female), a minority had a non-cancer diagnosis (32%) and an expected prognosis of hours to days (18%). Most (87%) were admitted to the PCU. The subgroup (N = 133) had similar characteristics to the total cohort. Regression analysis in the total cohort showed that dying in hospital awaiting PCU admission was associated with a prognosis of hours to days (OR = 7.42; 95% CI = 1.87-29.42) but not with having a non-cancer diagnosis, after adjusting for sociodemographic, functional, and clinical factors. Regression analysis in the subgroup showed no significant association between dying in hospital awaiting PCU admission and being followed by a palliative care specialist or having a documented goals of care discussion before the terminal admission.

In hospitalized individuals awaiting PCU admission, only a very short prognosis was associated with dying in hospital. Despite historical evidence of poor access, a non-cancer diagnosis was not associated with dying in hospital awaiting PCU admission.