Assessing sedation during intensive care unit mechanical ventilation with the Bispectral Index and the Sedation-Agitation Scale

Abstract

To describe the level of sedation for a cohort of mechanically ventilated adult intensive care unit (ICU) patients using validated subjective and objective tools. Prospective convenience sample. Multidisciplinary 34-bed ICU at Maine Medical Center, a 599-bed nonuniversity, academic medical center. Sixty-three adult ICU patients were monitored during 64 episodes of ventilatory support. Patients were prospectively evaluated by one trained investigator using the revised Sedation-Agitation Scale (SAS) and were simultaneously monitored for 1 to 5 hrs using the Bispectral Index (BIS), a numeric scale from 0 to 100 derived from the electroencephalogram. BIS values were assigned to baseline, stimulated, and average conditions for each patient by a separate investigator blinded to SAS scores. Ventilator settings, medications, and the lung injury severity (LIS) score were also recorded. Sedation levels varied from very deep sedation (SAS score = 1, BIS score = 43) to mild agitation (SAS score = 5, BIS score = 100). Heavily sedated patients (SAS score = 1-2, n = 20) had higher FIO2 (0.52 vs. 0.42, p = .008), oxygenation index (9.4 vs. 5.4, p = .03), and LIS scores (1.3 vs. 0.7, p = .004) and lower baseline (66 vs. 78, p = .01), average (66 vs. 81, p < .001), and stimulated (89 vs. 96, p = .016) BIS scores compared with more awake patients. Patients with intermittent neuromuscular blockade use (n = 4) had higher FIO2 (0.65 vs. 0.44, p = .006), minute ventilation (14.6 vs. 9.9 L/min, p = .005), positive end-expiratory pressure (7.5 vs. 4.8 cm H2 O, p = .05), oxygenation index (15.7 vs. 6.0, p < .001), and LIS scores (3.3 vs. 1.0, p = .036) and were more sedated, with higher suppression ratios (3.5 vs. 0.6, p = .05) and lower SAS scores (1.5 vs. 4, p = .035). The average BIS values correlated well with SAS (r2 = .21, p < .001). SAS and BIS work well to describe the depth of sedation for ventilated ICU patients. Deeper sedation and intermittent neuromuscular blockade were used for patients with greater ventilatory requirements and more severe lung disease. The correlation between subjective and objective scales varied in medical, surgical, and trauma patients. Further research with SAS and BIS may facilitate the development of quantitative sedation guidelines for the ICU. (Crit Care Med 1999; 27:1499-1504)