Unplanned Extubation: Insight from Three New Studies

Abstracts & Commentary

Three recent prospective studies from europe shed light on the problem of unplanned extubation in patients in the ICU. In addition to summarizing their findings, considering the three studies together provides further insight into this frequent and important clinical issue.

The frequency of unplanned extubations (UEX) among intubated ICU patients in the study of Boulain et al was 46 of 426 or 10.8%. At the time of UEX, 61% of the patients were agitated. Of the 46 instances of UEX, 28 (61%) required reintubation. Factors contributing to UEX were chronic respiratory failure, poor endotracheal tube fixation, orotracheal intubation, and lack of intravenous sedation.

This study was a prospective, multicenter, observational study performed in 11 ICUs in 11 French hospitals. Data including clinical status, sedation, approach to mechanical ventilation, nurse-to-patient ratio, the need for reintubation after UEX, and complications of extubation were recorded on each mechanically ventilated patient in every participating ICU during a two-month period. A total of 426 patients were ventilated, and of these, 46 (10.8%) had at least one UEX. Five patients had two UEXs and two patients had four UEXs, for a total of 57 episodes of UEX, at a rate of 1.59 UEX per 100 intubated days.

Fifty-five of the UEXs were considered self-extubations. Overall, 10 occurred during nursing procedures and one patient died as a direct result of the UEX. Multivariate analysis identified four factors contributing to UEX: chronic respiratory failure, poor endotracheal tube fixation, orotracheal intubation, and lack of intravenous sedation. There were no differences in nosocomial pneumonia rate or mortality between patients with UEX and those who did not self-extubate. Of the patients with UEX, 39% did not require reintubation.

In the study of Betbesé et al, the frequency of UEX in 750 mechanically ventilated patients in two Spanish hospitals was 7.3%. Deliberate patient extubation occurred 78% of the time; 56% of the time patients were in the process of weaning, and 37% of the patients who deliberately self-extubated did not require reintubation.

This prospective evaluation of the frequency of UEX was conducted in two ICUs in a single hospital in Barcelona, Spain. Data on clinical presentation, approach to ventilation, size of endotracheal tube, level of sedation, and complications were gathered on 750 ventilated patients over a 32-month period. During this time 55 patients (7.3%) self-extubated, for a total of 59 UEX. Of the 59 UEX, 78% were judged to have been deliberate on the part of the patient and 22% were considered accidental. A total of 32 patients (54%) did not require reintubation after UEX. All patients in both ICUs were orally intubated; no patient was nasally intubated during the study period.

Twenty-seven (46%) UEXs occurred during full ventilatory support. In the remaining UEXs, patients were in the process of weaning. Of the 46 patients who deliberately self-extubated, 37% required reintubation, while 77% of the accidental UEXs (10/13) required reintubation. In addition, only 16% (5/32) of patients who were weaning at the time of the UEX required reintubation, while 82% of the patients who had been receiving full ventilatory support required reintubation. No patient died as a result of an UEX. Multivariate analysis indicated that reintubation was more likely to be required after UEX the longer the duration of mechanical ventilation at the time of the UEX; weaning was associated with a lack of need for reintubation.

In the third study, by Chevron et al, the frequency of UEX in 414 mechanically ventilated patients in a single French hospital was 15.9% (66 instances). Deliberate self-extubation occurred in 87% of UEX.

This study was performed in a single French ICU over a 15-month period. During this time 414 patients were intubated and ventilated, of whom 66 (15.9%) extubated themselves. Of the total UEXs, 87% were considered deliberate patient self-extubations, with 13% considered accidental. UEX occurred more frequently in patients who were orally intubated, who were agitated, and had their hands restrained. At the time of the UEX 17 of the patients were in the process of weaning.

Patients who required reintubation (23) following a UEX had a Glasgow Coma Scale score of less than 11 and a PaO2/FIO2 ratio of less than 200 mm Hg, and were most commonly extubated accidentally. In contrast, patients who deliberately removed their endotracheal tubes had Glasgow Coma Scale scores of more than 11, PaO2/FIO2 ratios greater than 200 mm Hg, and only one of them was reintubated. Only one of the patients who self-extubated died as a direct result of the UEX. (Boulain T, et al. Am J Respir Crit Care Med 1998;157:1131-1137; Betbesé A-J, et al. Crit Care Med 1998;26:1180-1186; Chevron V, et al. Crit Care Med 1998;26:1049-1053.)


These three studies, when taken in total, identify a number of important issues regarding patients who are most likely to extubate themselves in the ICU. Considered in total, of 1590 ventilated patients in these studies, 167 (10.5%) had an unexpected extubation. What is more interesting is that 53% (89 patients) did not require reintubation. In the one study that provided data on the number of patients participating in weaning (the study of Betbesé and colleagues), the reintubation rate following an UEX was only 15.6%. This emphasizes problems in identifying readiness for ventilator discontinuation and extubation. It seems reasonable to state that those patients who did not require reintubation should already have been weaned and extubated.

Patients in all three studies who were agitated, regardless of the reason, had the greatest likelihood for self-extubation. In two of these studies the authors identify oral endotracheal intubation as a risk factor for UEX. However, the Betbesé study, which reported on the largest number of ventilated patients, had the lowest rate of UEX (7.3% vs 10.8% and 15.9%), and only used oral intubation. More important, as identified by Boulain and associates, was the method used to secure the endotracheal tube. Boulain et al indicated that the least likely securing method to be associated with UEX was the use of a cloth tie that went around the back of the patient’s neck. This was the same method that was used by Betbesé et al in all patients. However, Chevron and colleagues, who reported the highest UEX rate, used this same method in orally intubated patients.

The likelihood of UEX in all these studies was associated with agitation and insufficient sedation, as well as with inappropriate prolongation of the process of mechanical ventilation. Both of these factors indicate the need for more rapid identification of patient readiness to wean and be extubated following ventilatory support. As shown by the experience of Chevron et al and Betbesé et al, patients who are in the process of weaning, especially those with Glasgow Coma Scale scores higher than 11 and PaO2/FIO2 ratios greater than 200 mm Hg who deliberately extubate themselves, most likely will not require reintubation. In patients who fit this description, a trial of spontaneous breathing should occur before reintubation is considered. More important, institutions with a high percentage of patients who extubate themselves during weaning and who do not require reintubation should review their approach to both weaning and extubation.