Unplanned hospital admissions toward the end of life for patients who received palliative chemotherapy for breast cancer


Session type:


Chloe Bright, Casey Dunlop, Rebecca Smittenaar, Cong Chen, Sean McPhail, David Dodwell, Mick Peake, Kathy Prichard-Jones, Emma Kipps



Patients with incurable cancer are treated with chemotherapy to improve or prevent cancer-related symptoms, maintain quality of life and improve survival. In this population-based cohort study, we describe the morbidity associated with chemotherapy given towards the end of life, to support communication on prognostication and informed choice on continuing or discontinuing active treatment.


Cancer registrations for women diagnosed with breast cancer in England, during 1995-2017 who died between 2014-2017 and had a record of palliative chemotherapy (as recorded, derived using regimen if missing) within 2 years of death in the Systemic Anti-Cancer Therapy dataset were linked to Hospital Episode Statistics data. We used logistic regression (outcome=emergency hospital admission in last 90 days of life yes/no) adjusting for patient-level variables (age, income deprivation and line of palliative chemotherapy in the last 90 days of life).


10,966 patients with breast cancer were analysed. Having palliative chemotherapy (any line) prescribed in the last 90 days of life was associated with greater odds of emergency admission. Women who previously had palliative chemotherapy and had a 2nd line prescribed in the last 90 days of life had over 2-fold increased odds compared to patients not prescribed palliative chemotherapy in last 90 days of life (Adjusted Odds Ratio [OR]=2.3, p<0.001). Older age was associated with a lower odds of emergency admission (OR 80+ years vs. 50-54 years =0.5, p<0.001). Greater levels of deprivation were associated with a greater odds of emergency admission, (OR most/least deprived group=1.2, p=0.008).


In women with breast cancer, palliative chemotherapy in the 90 days before death was associated with an increased risk of emergency hospital admission. Younger age and higher deprivation were also found to be associated.

Impact statement

Patients, towards the end of their life, need clear information about the risks associated with palliative chemotherapy, to allow informed decision making and consent. Population-based cohort data can be used to better understand the morbidity associated with chemotherapy and facilitate the discussions between patients and health care professionals towards the end of life. 

Acknowledgements: This work includes patient data collated by the National Cancer Registration and Analysis