Characteristics and outcomes of patients who self-extubate from ventilatory support: a case-control study.

OBJECTIVE To identify factors associated with the occurrence of deliberate self-extubation and to describe associated patient outcomes. DESIGN Case-control study. SETTING ICUs of a national referral, tertiary medical center. PARTICIPANTS Fifty adult, intubated patients who had self-extubated from mechanical ventilatory support. Two control subjects who had not self-extubated were matched to each case based on age, gender, primary discharge diagnosis, and time hospitalized (within same quarter). MEASUREMENTS Standardized coding of medical record information, including demographic characteristics, clinical information, intubation and mechanical ventilation characteristics, medications, and selected laboratory indexes. RESULTS As compared to the control subjects, patients who self-extubated were more likely to be medical than surgical patients (p<0.001) and have a current history of smoking (p<0.05). Prior to the self-extubation, patients had a greater likelihood of hospital-acquired infections (p<0.001) or other hospital-acquired adverse events (p<0.001), abnormal (<10, >50 mg/dL) BUN (p<0.05), and abnormal (<20, >50 mm Hg) PaCO2 (p<0.05); they also were more likely to be restless or agitated (p<0.001), and more likely to be physically restrained (p<0.001). A logistic regression model demonstrated that presence of restlessness or agitation and presence of a hospital-acquired adverse event were independently associated with self-extubation from mechanical ventilatory support. In examining outcomes, as compared to the control subjects, those who self-extubated had longer lengths of stay in ICU and hospital, were more likely to need reintubation, and were more likely to suffer complications from intubation. However, none of the cases died within 48 h of self-extubation. CONCLUSION The results underscore the need for clinical guidelines for weaning and for monitoring patients at risk of self-extubation.