Acceptance of the possibility of simulated patient death among final-year medical students: a repeated cross-sectional study across eight consecutive cohorts

Abstract

Simulation scenario design has long been guided by a recommendation that the simulated patient should not be allowed to die, on the grounds that a fatal outcome may impair learning and compromise psychological safety. This "no-death paradigm" rests largely on expert opinion and on evidence from novice learners, with little empirical data on the attitudes of final-year students, who are closest to independent practice. We examined whether final-year medical students accept the possibility of simulated patient death, and whether this acceptance varies over time or by sociodemographic characteristics. We conducted a repeated cross-sectional study of final-year medical students attending a mandatory high-fidelity simulation session at a Spanish Faculty of Medicine across eight consecutive academic years (2018–2019 to 2025–2026). Within the session, patient death was a possible but non-scripted outcome of any scenario, contingent on clinical evolution and participant decisions. The questionnaire did not record whether respondents witnessed a death during their session. The primary outcome was a single attitudinal item specified a priori (item P28), rated on a five-point Likert scale, with agreement defined as a score of 4 or 5. Differences by cohort and sex were assessed with non-parametric tests; proportions are reported with 95% Wilson confidence intervals. Of 1,280 students attending the session, 1,263 completed the questionnaire (participation rate 98.7%). The possibility of simulated death was endorsed by an overwhelming majority: the median response was 5 (interquartile range 4–5), with 92.3% agreeing and 72.0% selecting the maximum value. Agreement exceeded 90% in every academic year and remained high throughout the COVID-19 pandemic cohorts. There was no evidence of differences by sex (91.5% of female versus 93.8% of male participants) or across age groups. Final-year medical students accepted the possibility of simulated patient death as an educationally permissible scenario outcome with consistency across eight years. Because the outcome is a single attitudinal item, these findings are evidence on learner attitudes rather than on responses to an experienced death; whether acceptance translates into educational benefit warrants further study. The possibility of death within a scenario merits consideration for students approaching independent practice, supported by sound scenario design and structured debriefing.