Longitudinal symptom profile of palliative care patients receiving a nurse-led end-of-life (PEACH) programme to support preference to die at home

Objectives Tailored models of home-based palliative care aimed to support death at home, should also ensure optimal symptom control. This study aimed to explore symptom occurrence and distress over time in Palliative Extended And Care at Home (PEACH) model of care recipients. Design This was a prospective cohort study. Setting and participants Participants were consecutive recipients of the PEACH rapid response nurse-led model of care in metropolitan Sydney (December 2013-January 2017) who were in the last weeks of life with a terminal or deteriorating phase of illness and had a preference to be cared or die at home. Outcome measures Deidentified data including sociodemographic and clinical characteristics, and symptom distress scores (Symptom Assessment Score) were collected at each clinical visit. Descriptive statistics and forward selection logistic regression analysis were used to explore influence of symptom distress levels on mode of separation ((1) died at home while still receiving a PEACH package, (2) admitted to a hospital or an inpatient palliative care unit or (3) discharged from the package (alive and no longer requiring PEACH)) across four symptom distress level categories. Results 1754 consecutive clients received a PEACH package (mean age 70 years, 55% male). 75.7% (n=1327) had a home death, 13.5% (n=237) were admitted and 10.8% (n=190) were still alive and residing at home when the package ceased. Mean symptom distress scores improved from baseline to final scores in the three groups (p<0.0001). The frequency of no symptom distress score (0) category was higher in the home death group. Higher scores for nausea, fatigue, insomnia and bowel problems were independent predictors of who was admitted. Conclusion Tailored home-based palliative care models to meet preference to die at home, achieve this while maintaining symptom control. A focus on particular symptoms may further optimise these models of care.