The prognosis for patients with critical conditions remains difficult, especially for the elderly. Time-limited trials (TLT) were used to reduce predictive uncertainty in individual patients by monitoring treatment response over a predetermined period. However, determining the length of that period was extremely difficult. Based on temporal profiles of uncertainty about critical care and outcome, the study presented a probabilistic method for estimating the appropriate duration of a TLT. The VIP2 study cohort included very old patients (age≥ 80 years, n=1209) who were admitted to the ICU for between 2 and 14 days, with respiratory or circulatory support from day 1 and either no limitations of life-sustaining treatment or a decision to withdraw that treatment, as well as complete data. Multi-state critical care trajectory modeling to obtain time-dependent probabilities for transitions between different organ support levels and outcome states. Shannon’s entropy of probability distributions at discrete points in time quantifies the degree of uncertainty. Researchers discovered periods of increased prognostic uncertainty lasting up to 7 days after admission. The length of these periods was determined by the patient’s baseline characteristics (frailty, severity of critical illness) and the extent of organ support. Time-dependent patterns of uncertainty regarding critical care response could inform decisions about the duration of TLTs, which could last up to a week in very old patients.