Very Elderly Patients' Treatment Preferences and Participation in Shared Decision-making in the Event of Acute Life-threatening Illness

Do very elderly adults wish intensive care in the event of acute life-threatening illness and are their next of kin able to predict these preferences?

Very elderly patients are a steeply increasing patient population in intensive care units (ICUs), but the overall benefit of intensive care for these patients remains controversial. Will ICU admission improve survival and quality of life, or will it prolong suffering and delay natural death? Little is known about very elderly Norwegians life sustaining treatment (LST) preferences in these situations where treatment benefit is uncertain.

This project aims to improve critically ill very elderly patients' ICU trajectories by bringing forth knowledge about their treatment preferences, their family members' ability to predict these preferences, and by directing attention to the challenges of consent to critical care in cases of medical uncertainty.

A selv administered, mailed survey will be distributed among 400 outpatients aged 80 years or older and their next of kin. Respondents will be recruited at the ophthalmologic, ear-nose-and-throat and orthopaedic outpatient clinics at Haukeland University Hospital Bergen, Norway.

The investigators developed and validated a survey tool for this purpose, containing 3 hypothetical scenarios of acute life-threatening illness. The scenarios are randomly chosen from 20 hypothetical patient histories and are representative for ICU admission diagnoses in Norway and Europe. The participants will be asked for treatment choices, i.e. wishing admission to intensive care or not. A response option 'not wishing to engage in the treatment decision' is also provided.

Furthermore, the questionnaire includes factors that may influence elderlies' treatment preferences and proxies' ability to predict these preferences including: demographics, religion, previous experience with and / or communication about critical illness, comorbidity, frailty, quality of life, and projections (i.e. the proxy's own treatment preferences).

The respondents are requested to explain their choices by free-text comments after each scenario. They are also asked to elaborate how they wish next-of-kin should contribute to decision making in these cases. Additional space for free-text comments is provided in the end of the questionnaire.

The study design is exploratory. Responses will be analysed with both quantitative statistics and qualitative methods.