Objective: To investigate staff attitudes toward assisted suicide in the hospital and factors influencing its practices in Switzerland.

Design: Cross-sectional study

Setting: Two University Hospitals in French speaking regions of Switzerland.

Participants: 13’834 health care professionals including all personal caring for patients were invited to participate.

Main outcome measures and other variables: Attitudes towards the participation of hospital health care professionals in assisted suicide were investigated with an online questionnaire.

Results: Among all invited professionals, 5’127 responded by filling in the survey at least partially (response rate 37.0%) and 3’683 completed the entire survey (26.6%). 72.6% of participants approved that this practice should be authorized in their hospital and saw more positive than negative effects. 57.6% would consider assisted suicide for themselves. Non-medical professionals were 1.23 to 5.23 times more likely to approve assisted suicide than physicians (p<0.001). 68.8% of respondents indicated that each professional should have the choice whether or not to assist in suicide.

Conclusions: This multi-professional survey sheds light on hospital staff perceptions of assisted suicide happening within hospital walls, which may inform the development of rules considering their wishes but also their reluctance. Further research using a mixed-methods approach could help reaching an in-depth understanding of staff’s attitudes and considerations towards assisted suicide practices.

A cross-sectional study was conducted using a self-administered online questionnaire. The instrument was derived from a questionnaire employed in a nationwide survey of general practitioners by the Swiss Academy of Medical Sciences. It was adapted for use by all hospital health care professionals. It consisted of 15 main questions with a total of 80 sub-questions, addressing experiences and practices as well as personal attitudes concerning assisted suicide. Items on demographic characteristics were also included. The questionnaire was converted to an online form using the software REDCap (Vanderbilt University) and a link was sent to targeted population in a two-step procedures: at first participating center first (University hospitals of Geneva), then at second participating center (Centre Hospitalier Universitaire Vaudois). Finally both databases were exported in STATA format and combined in a single database.

For confidentiality matters, we deleted some variables that could help identifying some respondents potentially (service and department affiliations, religion). The dataset includes all observations to reproduce potentially our analyses. There are some missing data in almost every variables.