Forecasting Caseload of Critically Ill Patients Who Are Alert and Without Delirium for at Least Two Consecutive Days for the Assessment of Their Psychological Distress.

Introduction: One-quarter of alert, non-delirious patients in critical care units report significant psychological distress. Treatment of this distress depends upon identifying these high-risk patients. Our aim was to characterize how many critical care patients remain alert and without delirium for at least two consecutive days and could thus predictably undergo evaluation for distress.

Methods: This retrospective cohort study used data from a large teaching hospital in the United States of America, from October 2014 to March 2022. Patients were included if they were admitted to one of three intensive care units, and for >48 hours all delirium and sedation screenings were negative (Riker sedation-agitation scale four, calm and cooperative, and no delirium based on all Confusion Assessment Method for the Intensive Care Unit scores negative and all Delirium Observation Screening Scale less than three). Means and standard deviations of means for counts and percentages are reported among the most recent six quarters. Means and standard deviations of means for lengths of stay were calculated among all N=30 quarters. The Clopper-Pearson method was used to calculate the lower 99% confidence limit for the percentages of patients who would have had at most one assessment of dignity-related distress before intensive care unit discharge or change in mental status.

Results: An average of 3.6 (standard deviation 0.2) new patients met the criteria daily. The percentages of all critical care patients (20%, standard deviation 2%) and hours (18%, standard deviation 2%) meeting criteria decreased slightly over the 7.5 years. Patients spent a mean of 3.8 (standard deviation 0.1) days awake in critical care before their condition or site changed. In the context of assessing distress and potentially treating it before the date of change of condition (e.g., transfer), 66% (6818/10314) of patients would have zero or one assessment, lower 99% confidence limit of 65%.

Conclusions: Approximately one-fifth of critically ill patients are alert and without delirium and thus could be evaluated for distress during their intensive care unit stay, mostly during a single visit. These estimates can be used to guide workforce planning.