The relationship of sedation to deliberate self-extubation

Study Objectives: To evaluate the relationship between sedative therapy and self-extubation in a large medical-surgical intensive care unit (ICU). Design: Retrospective, case-controlled study. Setting: Large teaching hospital. Patients: All adult patients who underwent unplanned self-extubation during a 12-month period (n = 50). Each patient was matched to two control patien ts who did not self-extubate based on age, gender, dates in hospital and di agnosis. Interventions: none. Measurements: Data collected included time to self extubation, dosages and types of benzodiazepines, opioid analgesics, antipsychotics, and hypnotics. Data on the degree of agitation as assessed by nursing staff also were obt ained. Main results: When compared to controls, patients in the self-extubation gr oup were more likely to have received benzodiazepines (59% vs. 35%; p < 0.0 5), but equally likely to have received opioids and/or paralytic drugs. Pat ients who self-extubated were twice as likely as controls to be agitated (5 4% vs. 22%; p < 0.05). Use of benzodiazepines was more common in agitated p atients than in nonagitated patients (62% vs. 35%; p < 0.02). Among nonagit ated patients who self-extubated, increased use of benzodiazepines (57% vs. 29%; p < 0.05) was noted when compared to nonagitated controls. Conclusions: In intubated ICU patients, benzodiazepines may not consistentl y treat agitation effectively or prevent self-extubation. Such an effect ma y be due to paradoxical eh citation, disorientation during long-term admini stration, or differences in drug administration between ICU and operating r oom (OR) environments. (C) 2001 by Elsevier Science Inc.