Is the bispectral index appropriate for monitoring the sedation level of mechanically ventilated surgical ICU patients?

Objective: To determine the value of the bispectral index (BIS) in assessing the depth of sedation in sedated and mechanically ventilated ICU patients, compared with clinical sedation scores. Design and setting: Prospective convenience sample in a 12-bed anesthesiological-surgical ICU of a university hospital. Patients: 19 consecutive patients without any central neurological diseases requiring mechanical ventilation for more than 24 h. Measurements: BIS version 3.12 and clinical depth of sedation assessed by the modified Observers's Assessment of Alertness/Sedation Scale, modified Glasgow Coma Scale, modified Ramsay Scale, Cook Scale, and Sedation-Agitation Scale were measured twice daily while patients were intubated and once daily after extubation until discharged from ICU. Results: there was a moderate correlation between BIS and each sedation score in 11 patients (58%, "BIS patients") and no correlation in 8 patients (42%, "non-BIS patients"). We found no parameters distinguishing between these two groups. On average eight measurements were necessary to establish a statistical correlation. In the BIS patients the slopes of the linear regression curves showed significant differences for all BIS score combinations with increasing scattering at deeper sedation levels. Conclusions: BIS is correlated only in some ICU patients with the clinical assessment of their sedation level as based on various scores. At deeper sedation levels the interindividual differences increase. There were no criteria found to distinguish patients with and without correlation. This suggests that the BIS is not suitable for monitoring the sedation in a heterogeneous Group of surgical ICU patients.