Factors predicting reintubation after unplanned extubation

Background and Purpose: Factors predicting reintubation after unplanned extubation (UE) are not well established. We prospectively studied the incidence and clinical features of UE and predictive factors for reintubation in medical intensive care unit (ICU) patients. We also validated the scoring system of Listello and Sessler to predict the outcome after UE. Methods: Over a 14-month period, patients who had a first UE while being treated in our medical ICU were identified. Data including clinical features, laboratory data and ventilator parameters were collected and compared between the reintubation and non-reintubation groups. Results: Fifty episodes of UE occurred during the study period, 8.5% of all endotracheal intubations. More than half of the episodes of UE (26/50, 52%) occurred within 48 hours after intubation; reintubation was required in 33 episodes (33/50, 66%). In univariate analysis the only significant difference between the reintubation and non-reintubation groups in pre-extubation parameters was pneumonia as the cause of respiratory failure, which was significantly associated with the need for reintubation (52% vs 18%, respectively; odds ratio 4.96; confidence interval 1.24-19.91; p = 0.02). All patients with pneumonia who had UE within 5 days after intubation required reintubation. In contrast, there was a low rate of reintubation in patients with heart disease (3/9, 33%). Validation of the scoring system of Listello and Sessler revealed that outcome of UE was correctly predicted in only 62% (31/50) of episodes. Conclusions: Pneumonia as the cause of respiratory failure was the most important predicting factor for reintubation after UE. Reintubation after UE may not be necessary, especially in patients with heart disease as the cause of respiratory failure. The scoring system of Listello and Sessler did not accurately predict reintubation after UE in this study.