Timing matters: sex differences in treatment limitation decisions in intensive care

Le résumé fourni par la source

BACKGROUND: Sex differences in intensive care treatment and mortality are well documented, but the timing of decisions to limit treatment remains unclear. We investigated whether sex differences in decisions to limit treatment arise at ICU admission or during the ICU stay. METHODS: Nationwide cohort study using the Swiss Minimal Dataset for Intensive Care Units, including adult (≥ 18 years) ICU admissions between 2016 and 2024. Two adjusted logistic regression models assessed treatment limitations documented at ICU admission and those occurring later among patients admitted without limitations. RESULTS: Among 654,660 ICU stays, treatment limitations at admission were more common in women than men (12.0% vs. 8.6%), whereas rates during the ICU stay were similar (5.5% vs. 5.5%). Female sex was independently associated with limitations at admission (aOR 1.26, 95% CI 1.24-1.28) but only weakly associated with later limitations (aOR 1.10, 95% CI 1.08-1.13). Differences at admission varied by diagnosis and were most pronounced in trauma and cardiovascular conditions. Women more often had ceiling-of-care decisions and documented patient wishes, whereas men more frequently underwent withdrawal of life-sustaining therapies and physician-driven decisions. Mortality was highest with limitations at ICU admission and lowest without limitations, with minimal sex differences within categories. CONCLUSIONS: In Switzerland, sex differences in treatment limitations occur mainly at ICU admission and vary across diagnoses. These findings suggest that differences may reflect early triage heuristics, societal norms influencing advance care planning, and potential implicit biases under prognostic uncertainty. Structured goal-of-care discussions at ICU admission may help promote consistent and equitable decision-making.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.