The Challenges of Prognosticating Persistent Critical Illness: A Qualitative Study

Medical decision-making in the ICU is influenced by a patient's anticipated clinical course. However, intensivists struggle to identify accurately which patients will experience within-ICU events-such as new late-onset shock or acute hypoxic respiratory failure (AHRF)-that contribute to the development of persistent critical illness (PerCI). Given the uniqueness of patients who experience PerCI, understanding how intensivists process and incorporate evolving clinical information into their prognostication remains unknown for this population.

What processes do intensivists use to prognosticate a patient's risk of new shock or AHRF and PerCI developing and what challenges are encountered?

We conducted semistructured interviews with 17 intensivists at a quaternary academic medical center from July 2022 through January 2023. Interviews occurred within 4 weeks of completing medical ICU service and lasted 30 to 60 minutes. Each interview focused on 2 real-world patient cases treated by the intensivist, using chart-stimulated recall to examine the information used to prognosticate each patient's risk of new shock or AHRF and PerCI developing across the ICU course and the challenges of prognostication. We performed qualitative content analysis to explore the themes discussed by the intensivists.

We identified 3 major themes about how intensivists formulate prognostic estimates about the risk of new shock or AHRF and PerCI: first, the contextualization of patient factors at ICU admission; second, the type of illness script or mental model of disease, including those within and outside of the intensivists' expertise; and third, the impact of clinical setbacks and consultant input on the patient's predicted clinical trajectory.

This study shows that intensivists rely on patient factors and the illness script to shape their prognostication. However, clinical setbacks and consultant recommendations were common, leading to changes in prognostication and subsequent medical decision-making. Future research should explore the value of time-limited trials in guiding treatment decisions in the context of prognostic uncertainty.