Abstract
Disorientation may present as a warning sign of developing delirium. The most commonly used delirium assessment tool in Germany, the Confusion Assessment Method for Intensive Care Unit (CAM-ICU), does not rate "disorientation", since intubated patients cannot communicate verbally. However, the majority of German ICU patients are not orally intubated, so they could be examined for their orientation. This study was carried out to investigate whether the delirium feature "disorientation" in extubated patients yields diverging findings in comparison to the CAM-ICU and whether the sensitivity of the CAM-ICU may be improved when combined with the feature "disorientation" (CAM-IMC). A total of 86 paired assessments were completed in 50 extubated patients. Delirium was found in 19.8% (N = 17). The CAM-ICU had a sensitivity of 71% (95% confidence interval [CI] 44-90%) and a specificity of 100% (95-100%). Disorientation, if taken as the only delirium feature, had a sensitivity of 77% (50-93%) and a specificity of 93% (89-100%). The CAM-IMC reached a sensitivity of 88% (64-99%) and a specificity of 100% (95-100%). The receiver operating characteristics (ROC) analyses found an area under the curve (AUC) of 0.941 (95%CI 0.851-1.000) for the CAM-IMC, which was the highest compared to the other delirium tests (CAM-ICU, AUC 0.853 [0.720-0.986]; disorientation, AUC 0.868 [0.745-0.991]). This research emphasizes the importance of the feature "disorientation" for delirium assessments in patients able to verbally communicate and explains some controversial delirium ratings in daily practice. The CAM-IMC appears to be an attractive tool for delirium assessment in nonintubated patients and deserves further research.