Objectives: To revalidate a means of assessing delirium in intensive care unit patients and to investigate the independent effect of delirium on the mortality of mechanically ventilated patients.
Design: A prospective cohort study.
Setting: A 37-bed medical intensive care unit of a tertiary care hospital.
Patients: Subjects were 102 of 131 consecutive mechanically ventilated patients.
Measurements: All the enrolled patients were assessed for delirium using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). Mortality rate were compared between patients with or without delirium, and the predictors of death were investigated.
Results: The two CAM-ICU assessors' sensitivities in diagnosing delirium compared with reference standard were 91% and 95%, whereas their specificities were both 98%. They also demonstrated high interrater reliability with kappa statistics of 0.91. Delirium was present in 22 of 102 (22%) patients in the first 5 days. The delirious patients had higher intensive care unit mortality rate than nondelirious patients (63.6% vs. 32.5%, respectively), with a hazard ratio of 2.57 (95% confidence interval, 1.56-8.15). In multivariate analysis, delirium (odds ratio, 13.0; 95% confidence interval, 2.69-62.91), shock (odds ratio, 12.91; 95% confidence interval, 2.93-56.92), and illness severity (odds ratio, 9.61; 95% confidence interval, 2.24-41.18) were independent predictors of mortality.
Conclusions: This study confirms previous work showing that delirium is an independent predictor for increased mortality among mechanically ventilated patients.