Self-Extubation Revisited: A Case-Control Study.

BACKGROUND: To increase the understanding of the self-extubation phenomena, we assessed its rate in our medical ICU and aimed to identify the risk factors of self-extubation and the risk factors for re-intubation. METHODS: We prospectively identified subjects who self-extubated. Their baseline characteristics, including the Richmond Agitation Severity Scale score, reason for intubation, shift, distance of the endotracheal tube tip to the carina, and outcomes were collected retrospectively. For every subject who self-extubated, a control subject was selected from the mechanical ventilation database. RESULTS: During the study period, there were 2,578 admissions with 4,072 mechanical ventilation days. Fifty-three cases of self-extubation were recorded, which resulted in a self-extubation event rate of 1.3 per 100 days of mechanical ventilation. Forty-five controls were identified. The most common reason for intubation was hypoxic respiratory failure, followed by the need for airway protection and hypercapnic respiratory failure. Sedation was administered continuously in 34% of the subjects. Thirty-seven percent received no sedation. At the time of the event, the subjects who self-extubated had a higher Richmond Agitation Severity Scale score, a longer distance from the endotracheal tip to carina on the chest radiograph preceding the event, and a shorter duration of mechanical ventilation, and were more likely to be on volume-controlled mechanical ventilation. ICU mortality was lower in the self-extubation group, despite having a trend toward a higher Simplified Acute Physiology Score II. Sixteen subjects required re-intubation. Independent predictors of re-intubation were hypoxic respiratory failure as the reason for the initial intubation and self-extubation that occurred at night. The need for re-intubation was not associated with higher mortality. CONCLUSIONS: Results of our study showed that, in the era of reduced use of sedatives in the ICU, clinicians must be vigilant of the risk of self-extubation in the first 2 d of mechanical ventilation in patients who are agitated and with a longer endotracheal tube to carina distance on chest radiograph.