Bispectral analysis in trauma patients: An observational study

Study objectives: Bispectral analysis uses processed electroencephalograph information to measure level of consciousness and depth of sedation. Trauma patients often experience painful procedures performed in the emergency department. We determined correlations between bispectral analysis, Glasgow Coma Scale (GCS) score, and trauma patient recall, which included recall after rapid sequence intubation (RSI). Methods: We prospectively studied a convenience sample of trauma patients older than 17 years and meeting trauma team activation guidelines at an urban Level I trauma center. Providers were blinded to bispectral analysis scores, which were measured every 5 minutes for 60 minutes. Initial GCS score, interventions (painful procedures), and RSI were recorded. A follow-up survey conducted within 7 days assessed (1 to 5 Likert scale) the level of patient recall. Responses were dichotomized (1, 2 as negative and 3, 4, 5 as positive). Spearmanʹs correlations between GCS score (<8 versus ≥8) and bispectral analysis were calculated using 75, 80, and 85 as bispectral analysis breakpoints. Fischerʹs exact tests compared recall rates of patients grouped using these points. Recall rates of patients with or without RSI and with GCS score less than 8 versus 8 or greater were also compared. Results: Data were obtained for 200 subjects and follow-up for 55 (nonsedated with GCS score=15 at survey). Initial bispectral analysis median was 95. GCS score less than 8 occurred in 37 patients. RSI was performed on 67 patients. Significantly reduced recall rates were found for patients with GCS score less than 8 (P=.03) and bispectral analysis less than 75 (P<.01) or bispectral analysis less than 80 (P=.01) but not for bispectral analysis less than 85 (P=.09). Patients with RSI were less likely to experience recall (P<.01). GCS and bispectral analysis were significantly correlated at all breakpoints (ρ=.746, .754, .689). Conclusion: Correlation between GCS and bispectral analysis is highest when GCS score less than 8 and bispectral analysis less than 80 are used as breakpoints, and the likelihood of recall of initial traumatic resuscitation within 7 days is similar when patients are grouped by GCS score less than 8 or bispectral analysis less than 80. As assessed by patient recall, adequate sedation appears to occur after RSI. A bispectral analysis threshold less than 80 may be useful to monitor the adequacy of sedation after RSI. The number of patients lost to follow-up limits results.