Differential Effects of Gamma-Aminobutyric Acidergic Sedatives on Risk of Post-Extubation Delirium in the ICU: A Retrospective Cohort Study from a New England Health Care Network

Abstract

OBJECTIVES: To evaluate whether different gamma-aminobutyric acidergic (GABAergic) sedatives such as propofol and benzodiazepines carry differential risks of post-extubation delirium in the ICU. DESIGN: Retrospective cohort study. SETTING: Seven ICUs in an academic hospital network, Beth Israel Deaconess Medical Center (Boston, MA). PATIENTS: Ten thousand five hundred and one adult patients mechanically ventilated for over 24 hours. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We tested the hypothesis that benzodiazepine versus propofol-based sedation is associated with fewer delirium-free days within 14 days after extubation. Further, we hypothesized that the measured sedation level evoked by GABAergic drugs is a better predictor of delirium than the drug dose administered. The proportion of GABAergic drug-induced deep sedation was defined as the ratio of days with a mean Richmond Agitation-Sedation Scale of less than or equal to -3 during mechanical ventilation. Multivariable regression and effect modification analyses were used. Delirium-free days were lower in patients who received a high proportion of deep sedation using benzodiazepine compared with propofol-based sedation (adjusted absolute difference, -1.17 d; 95% CI, -0.64 to -1.69; p < 0.001). This differential effect was magnified in elderly patients (age > 65) and in patients with liver or kidney failure (p-for-interaction < 0.001) but not observed in patients who received a low proportion of deep sedation (p = 0.95). GABAergic-induced deep sedation days during mechanical ventilation was a better predictor of post-extubation delirium than the GABAergic daily average effective dose (area under the curve 0.76 vs 0.69; p < 0.001). CONCLUSIONS: Deep sedation during mechanical ventilation with benzodiazepines compared with propofol is associated with increased risk of post-extubation delirium. Our data do not support the view that benzodiazepine-based compared with propofol-based sedation in the ICU is an independent risk factor of delirium, as long as deep sedation can be avoided in these patients.