Objectives: To evaluate the impact of individual manifestations of delirium on outcome, describe them in critically ill adults, and validate nurses' bedside item assessments from the Intensive Care Delirium Screening Checklist (ICDSC).
Design: Prospective study.
Setting: Single 16-bed medical/surgical university hospital intensive care unit.
Patients: Six hundred consecutive patients admitted to the intensive care unit for >24 hrs.
Interventions: All patients were evaluated with the eight-item ICDSC throughout their intensive care unit stay. In all patients scoring positive on any ICDSC item, individual checklist items were tallied throughout the intensive care unit stay and assessed for impact on mortality. In addition, when the ICDSC score indicated delirium (> or = 4 of 8), the subsequent overall frequency of each item was also independently documented to describe delirious patient symptoms. ICDSC items were tested for discrimination between delirious and nondelirious patients. Throughout the study, the validity of bedside delirium assessments was assessed in 30 nurses.
Measurements and main results: We were able to assess 537 patients. In nondelirious patients, psychomotor agitation by ICDSC assessment was associated with a higher risk of mortality after adjustment for Acute Physiology and Chronic Health Evaluation, age, and the presence of coma. One hundred eight-nine patients (35.1%) developed delirium (i.e., ICDSC score > or = 4). On presentation (and throughout the intensive care unit stay), the most frequent features of delirium were inattention, disorientation, and psychomotor agitation. Each ICDSC item was highly discriminating between delirious vs. nondelirious patients. Correlation between gold standard adjudicators and nurses for the overall bedside evaluations of delirium were excellent (Pearson's correlation R = 0.924, p < .0005). Individual symptom evaluation by nurses varied: Alteration in level of consciousness was poorest (R = 0.681, p < .0005), and both disorientation and hallucinations evaluated best (R = 1.000).
Conclusions: In nondelirious patients, agitation was associated with a higher risk of mortality. Each of the eight ICDSC items is highly discriminating for the diagnosis of delirium, suggesting that any screening or diagnostic scales should incorporate them. Quality assurance and educational efforts should, therefore, emphasize independent assessment of the individual features of delirium.