Classification of daily mental status in critically ill patients for research purposes

Abstract

Purpose: The purpose of this study is to develop a reliable tool for daily mental status classification in intensive care unit (ICU) patients for research purposes. Secondly, to identify patients with single, 1-day episodes of delirium and to compare them with patients having more delirium days or episodes. Patients and methods: A 5-step algorithm was designed, which includes Richmond Agitation Sedation Scale and Confusion Assessment Method for the ICU scores from bedside nurses, initiation of delirium treatment, chart review, and the Confusion Assessment Method for the ICU administered by researchers. This algorithm was validated against a reference standard of delirium experts. Subsequently, a cohort study was performed in patients admitted to a mixed ICU. Results: In 65 paired observations, the algorithm had 0.75 sensitivity and 0.85 specificity. Applying the algorithm, interobserver agreement was high with mean Fleiss κ of 0.94 (5 raters) and 0.97 (4 raters). In the cohort study, 1112 patients were included of whom 535 (48%) became delirious. Single, 1-day episodes occurred in 43% of the delirious patients, whom were characterized by lower age compared with those with more delirium days. Conclusions: The algorithm for daily mental status classification seems to be a valid tool. In a substantial proportion of patients, delirium occurs only once during ICU admission lasting only 1 day.