End-of-life decision-making in very old critically ill patients

Abstract End-of-life decision-making in the intensive care unit (ICU) is ethically complex, particularly for very old patients with limited physiological reserve. Decisions to limit life-sustaining treatment are shaped by medical prognosis, patient values, cultural and socioeconomic contexts, and uncertainty regarding outcomes. Non-beneficial treatment refers to interventions unlikely to provide meaningful benefit while imposing additional burden or suffering, and avoiding such treatment is central to high-quality end-of-life care. Limitation of life-sustaining treatment may involve withholding or withdrawing therapies, which are distinct from euthanasia, and should be accompanied by a transition toward comfort-focused care. Frailty, pre-illness functional status, and illness severity are key determinants of outcomes and treatment decisions in very old ICU patients. Large international studies, including the VIP, VIP2, and COVIP cohorts, demonstrate that frailty is strongly associated with mortality, functional decline, and treatment limitation, while revealing substantial regional variation in end-of-life practices. Prognostication remains imprecise, often leading to prolonged aggressive care despite poor long-term outcomes. Shared decision-making, grounded in transparent communication and alignment with patient goals, is the ethical standard but is frequently undermined by discordance between documented preferences and delivered care. When decision-making capacity is lost, in the majority of countries shared decision-making extends to surrogate decision-makers, who often experience emotional burden and limited accuracy in predicting patient preferences. Ongoing, structured conversations, meticulous documentation, and time-limited trials of intensive therapy can support iterative reassessment of goals of care.