Objective: Bispectral index (BIS) is being evaluated as a monitor of consciousness, yet it is unclear what components of consciousness (i.e., arousal vs. content of consciousness) the BIS measures. This study compared BIS levels to well-validated clinical measures of arousal and the presence or absence of delirium.
Design: A prospective, blinded, observational cohort study.
Patients: 124 mechanically ventilated, adult, medical ICU patients.
Measurements and results: Using BIS 3.4 and BIS-XP 4.0 algorithms, BIS values were calculated immediately prior to clinical assessments. The clinical assessments included the Richmond Agitation-Sedation Scale (RASS) and presence or absence of delirium using the Confusion Assessment Method for the ICU. A total of 484 assessments were collected among 124 patients. BIS-XP values demonstrated greater correlation with RASS than BIS 3.4 ( R(2)=0.36 vs. 0.20), although considerable overlap of BIS-XP scores remained across RASS levels. Median BIS-XP values for delirious and nondelirious observations were 74 and 96, respectively, while BIS 3.4 values were 91 and 96, respectively. However, neither BIS 3.4 nor BIS-XP were significantly associated with delirium after controlling for RASS value.
Conclusions: In comparison with clinical measures of arousal in mechanically ventilated patients, BIS-XP algorithm demonstrated stronger correlation with RASS levels than did BIS 3.4, yet marked overlap across different levels of arousal persist using both algorithms. After controlling for level of arousal, neither BIS-XP nor BIS 3.4 algorithms distinguished between the presence and absence of delirium.