Preferences for Life-Sustaining Treatments Examined by Hidden Markov Modeling Are Mostly Stable in Terminally Ill Cancer Patients' Last Six Months of Life

Siew Tzuh Tang*, Fur Hsing Wen, Wen Cheng Chang, Chia Hsun Hsieh, Wen Chi Chou, Jen Shi Chen, Ming Mo Hou

*Corresponding author for this work

Research output: Contribution to journalJournal Article peer-review

8 Scopus citations

Abstract

Context Stability of life-sustaining treatment (LST) preferences at end of life (EOL) has not been well established for terminally ill cancer patients nor have transition probabilities been explored between different types of preferences. Objective We assessed the stability of cancer patients' LST preferences at EOL by identifying distinct LST preference states and examining the probability of each state transitioning to other states between consecutive time points. Methods Stability of LST preferences (cardiopulmonary resuscitation, intensive care unit [ICU] care, cardiac massage, intubation with mechanical ventilation, intravenous nutrition support, and nasogastric tube feeding) was examined among 303 cancer patients in their last six months by hidden Markov modeling. Results Six distinct LST preference states (initial size) were identified: uniformly preferring (8.3%), uniformly rejecting (33.8%), and uniformly uncertain about (20.5%) LST, favoring intravenous nutrition support but rejecting other treatments (19.9%), and favoring (3.6%) or uncertain about (14.0%) nutrition support and ICU care while rejecting other treatments. Shifts between LST preference states were relatively small between any two time points (transition probability of staying at the same state was 92.1% to 97.5%), except for the state characterized by uncertainty about nutrition support and ICU care while rejecting other treatments, in which 8.3% of patients shifted LST preferences toward uniform uncertainty at a subsequent assessment. Conclusions Our patients' LST preferences remained stable without prominent shifts toward preferring less aggressive LSTs even when death approached. Clarifying patients' understanding and expectations about LST efficacy and tailoring interventions to the unique needs of patients in each state may provide personalized EOL care.

Original languageEnglish
Pages (from-to)628-636.e2
JournalJournal of Pain and Symptom Management
Volume54
Issue number5
DOIs
StatePublished - 11 2017

Bibliographical note

Publisher Copyright:
© 2017 American Academy of Hospice and Palliative Medicine

Keywords

  • Preference stability
  • cancer
  • end-of-life care
  • life-sustaining treatments
  • oncology

Fingerprint

Dive into the research topics of 'Preferences for Life-Sustaining Treatments Examined by Hidden Markov Modeling Are Mostly Stable in Terminally Ill Cancer Patients' Last Six Months of Life'. Together they form a unique fingerprint.

Cite this