Annals of Thoracic Medicine Official publication of the Saudi Thoracic Society, affiliated to King Saud University
Search Ahead of print Current Issue Archives Instructions Subscribe e-Alerts Login 
Home Email this article link Print this article Bookmark this page Decrease font size Default font size Increase font size

Year : 2006  |  Volume : 1  |  Issue : 2  |  Page : 71-75
Unplanned extubation in the ICU: Impact on outcome and nursing workload

1 King Khalid National Guard Hospital, King Abdulaziz, Medical City - Jeddah, Saudi Arabia
2 Program in Critical Care, London Health Sciences Centre, London, Ontario, Canada

Correspondence Address:
Ayman Krayem
King Khalid National Guard Hospital, Intensive Care Unit, P.O. Box 9515, Jeddah 21423
Saudi Arabia
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/1817-1737.27105

Rights and Permissions


PURPOSE: To determine the incidence and factors associated with unplanned extubation (UE) in the intensive care unit (ICU) and its relationship with nursing workload. MATERIALS AND METHODS: A retrospective case-control study was carried out within a cohort of ventilated patients in two teaching hospital medical-surgical ICUs. A total of 50 adult patients with UE were studied. Controls were subjects without UE who were matched to the cases on the following Five factors: age, gender, admission diagnostic category, admission date (within 3 months) and duration of mechanical ventilation. Other data including patient demographics, comorbid conditions, APACHE III score, ventilation parameters, use of sedation, re-intubation, mortality rate and ICU/hospital length of stay were collected. Nine equivalents of nursing manpower use score (NEMS) and multiple organ dysfunction score (MODS) were calculated in both, cases and controls, 24 h before and after the event. RESULTS: Sixty-eight episodes of UE occurred in 66 patients during the 24-month study period (1.1%). Patients with UE were more agitated ( P <0.001) and required higher doses of benzodiazepines ( P =0.023) than their controls. UE was associated with a higher rate of re-intubation compared to the control group ( P <0.001) but was not associated with a longer length of stay in ICU or hospital or excess mortality ( P >0.05). The mean NEMS were not significantly different between the two groups 24 h before ( P =0.69) and after ( P =0.99) the extubation event. Also, the mean MODS were similar between both groups 24 h before ( P =0.69) and after ( P =0.74) extubation. CONCLUSION: In this study, agitation and greater use of benzodiazepines were frequently associated with UE and potentially can be used as risk factors for UE. We have found no significant impact of UE on increasing mortality and, in a manner not shown before, nursing workload.

Keywords: Acute physiology and chronic health evaluation score, multiple organ dysfunction score, noninvasive ventilation, positive end expiratory pressure, unplanned extubation

How to cite this article:
Krayem A, Butler R, Martin C. Unplanned extubation in the ICU: Impact on outcome and nursing workload. Ann Thorac Med 2006;1:71-5

How to cite this URL:
Krayem A, Butler R, Martin C. Unplanned extubation in the ICU: Impact on outcome and nursing workload. Ann Thorac Med [serial online] 2006 [cited 2023 Feb 9];1:71-5. Available from:

Unplanned extubation (UE) is a major complication of translaryngeal intubation, with a reported incidence ranging from 3 to 16% of mechanically ventilated patients.[1],[2],[3],[4] A UE has potentially serious consequences, since in 31 to 78% of cases, unplanned extubation requires re-intubation or is complicated by arrhythmias, bronchial aspiration, difficulty in re-intubation or even death.[2],[3],[5],[6]

Some studies have shown higher mortality for patients with failed UE as compared with those who successfully tolerate UE.[3],[7] Also, re-intubation was reported as a risk factor for ventilator-associated pneumonia (VAP).[7] In one study,[8] re-intubation was required for a high proportion (74%) of the patients who self-extubated with most re-intubations occurring within the first 24 h after UE. Patients who self-extubated were twice as likely as controls to be agitated and the use of benzodiazepines was more common in agitated than in non-agitated patients.[9] Published studies have shown variable results concerning the impact of UE on intensive care unit (ICU) mortality. Although in most studies, mortality was similar to that of controls, there is some evidence that patients with failed UE had a significantly longer duration of ventilation, longer ICU and hospital stay and survivors were more likely to require chronic care.[10] The prevention of self or accidental extubation is of significance to health care professionals.

There is evidence to suggest that with the implementation of a concerted continuous quality improvement program, the incidence of UE can be reduced significantly.[11],[12]

Studies to date have not examined the association between UE and nursing workload. We hypothesized that a higher nursing workload is associated with the occurrence of UE and that the UE event is likely to be associated with an increase in nursing workload on the individual patient level. In order to examine these associations, we conducted a retrospective case-control study to identify the factors associated with the event, its relationship to the nursing staff workload and the outcome of UE.

   Materials and Methods Top

Following institutional review board approval, we retrospectively identified all patients with UE in two medical-surgical ICUs in London, Ontario, Canada, From all the patients admitted to both ICUs over a 24-month period. The two ICUs have a total of 56 beds with more than 2,500 patients per year receiving mechanical ventilation. The study population consisted of all adult patients with any episode of UE, whether self-inflicted or accidental. This information is collected prospectively on all ICU patients and maintained in our critical-care database. Many of the patient data variables were extracted from the critical-care database. The remaining data were abstracted from the medical charts through individual chart review.

Each patient with an UE episode was matched to a single control patient from the same ICU who underwent mechanical ventilation during his/her ICU admission using the following criteria: same gender, age ( 2.5 years), admission date ( 3 months) and same diagnostic category. Finally, controls had to have a total duration of ventilation that was not less than 48 hours than the duration of ventilation of the cases at the point where the cases experienced their UE episode. The matching of the ventilation period was done to try to ensure that each case had a similar opportunity to be exposed to a UE episode as the control without producing a biased control group that was either too healthy or too sick in comparison to the case. Explicit matching on the basis of severity of illness was not done. For patients with more than one UE episode, only the first episode of UE was included in the analysis.

The variables collected on each patient included-patient demographics, ICU admission diagnosis, APACHE III score (acute physiology and chronic health evaluation score III) on admission to the ICU, major comorbidities, ventilation parameters (ventilatory mode, PaO 2 /FiO 2 ratio and PEEP) at the time of extubation, evidence of agitation as indicated in the nurses' notes, the use of sedatives and narcotics in the 24-hour period before extubation, whether re-intubation was required (date and time), use and duration of noninvasive ventilation (NIV) post-extubation, tracheostomy rate, ICU and hospital mortality and length of stay. In addition, the Nine Equivalents of Nursing Manpower Use Score (NEMS) and Multiple Organ Dysfunction Score (MODS) variables were collected 24 h before and after UE. For the control patients with planned elective extubation, these variables were collected and summed 24 h before and after the same point in time of mechanical ventilation as the UE event to control for the difference in the duration of mechanical ventilation and to allow comparison between both groups. If the control patient was not ventilated as long as the case, the NEMS and MODS were calculated 24 h before and following the control patients' planned extubation.

The NEMS was developed based on the Simplified Therapeutic Intervention Scoring System (TISS-28) and validated as a suitable and simpler therapeutic index to measure nursing workload in the ICU.[13] The 28 items in TISS-28 were reduced to 9 in NEMS with a maximum score of 66. The NEMS is utilized for the prediction of workload at the individual patient level. The NEMS represents the nursing activities associated with the monitoring and control of vital organ/system functions and covers a wide range of diagnostic and therapeutic activities in and outside the ICU usually associated with the severity of illness of the patient under treatment. Since UE carries the potential of worsening patient condition, the MODS was evaluated and compared between patients and controls. The MODS was used as a marker for the change in clinical and organ function status 24 h before and after the UE that might be attributed to the effect of UE. (For control patients, the time was matched to that of the cases.) The MODS was validated using simple physiologic measures of dysfunction in six organ systems and correlates strongly with the ultimate risk of ICU and hospital mortality.[14]

Data are presented as frequency for dichotomous data and mean and standard deviation for continuous data. Paired dichotomous data were analyzed using the McNemar test. Paired continuous data were analyzed using the Wilcoxon sign-rank test. Nonpaired dichotomous data were analyzed using Fisher's exact test. Nonpaired continuous data was analyzed using the Wilcoxon rank-sum test. All analyses were performed using SPSS 8.0 (Chicago IL). A P -value less than 0.05 was considered significant.

   Results Top

Sixty-eight UE occurred in 66 (1.1%) out of 6,223 mechanically ventilated patients during the study period. Fifty patients with UE were successfully matched to controls for all criteria. Patient demographics including age and admission diagnostic category were well matched [Table - 1]. Most of the patients were males (74%) and the majority (44%) was admitted with cardiovascular diseases as their admitting diagnostic category. The mean duration of ventilation was significantly shorter in the UE group compared to the controls (78.2 vs 186.8 h, P = 0.007) when examining the duration of ventilation up to the UE event in the cases. However, the total ventilation time, including the time after re-intubation for the UE group that got re-intubated, was not statistically different between both groups (179.4 h for UE patients vs. 186.8 h for the controls, P =0.52). Prior to extubation, there was no difference between the two groups in mode of ventilation, mean PaO 2 / FiO 2 ratio (219.5 for UE vs 209.2 for controls, P =0.98) and PEEP (6.1 cm H2O for UE vs 6.7 cm H2O for controls, P =0.43) [Table - 2]. Also, the UE patients and control patients had similar APACHE III scores on admission (81.6 vs 81.8 respectively, P =0.76).

Patients who self-extubated were more agitated and restless compared to control patients ( P <0.001) and they required the use of significantly more sedation in the form of benzodiazepine boluses prior to the episode of UE ( P =0.023) [Table - 1]. However, the use of 'as needed' narcotics ( P =0.82) or continuous infusions of sedatives or narcotics ( P =0.79) was not different between the two groups. Twenty-eight (56%) patients with UE required re-intubation, the majority within 24 h (75% of patients) and they were re-intubated more often than their matched control (28 vs 4, P =<0.001). Thirteen patients received noninvasive ventilation (NIV) post-UE compared to six controls following planned extubation ( P = 0.118). The tracheostomy rate and length of stay in ICU and hospital were also not significantly different between the two groups ( P >0.05) [Table - 1]. Hospital mortality rate was comparable between the two groups, with 10 deaths (20%) in the UE group compared to 15 (30%) in the control group ( P = 0.30). When we compared the subset of patients who failed UE with those who successfully tolerated these episodes, we found no excess mortality among patients with failed UE ( P =0.154). Similarly, there was no significant difference in the mode of ventilation (full vs weaning) prior to the UE event between the two subgroups [Table - 2]. However, among patients with UE, those who required re-intubation had significantly higher mean admission APACHE III score (88.6 vs 71.8, P =0.01) and longer ICU length of stay (13.5 vs 7.5 days, P =0.001) when compared to patients who tolerated UE [Table - 2].

Examining nursing workload using the NEMS, we found no significant difference in the mean NEMS between UE patients and their controls 24 h before (31.9 vs 32.0 respectively P = 0.77) and after the UE episode in the cases and at the same point of the ventilation time for the control patients (25.9 vs 25.9 respectively P =0.94) [Figure - 1]. In both groups, the NEM score decreased similarly post-extubation (31.9 to 25.9 for UE and 32.0 to 25.9 for the control group, P =0.99). Also, the MODS was not significantly different between groups in the 24-hour period before and after the extubation episode ( P =0.69 and 0.74 respectively). The MODS did not change 24 h after the UE episode or during the same period of time for the patients in the control group [Figure - 2]. Among the UE patients, both those who tolerated and failed UE had similar mean NEMS 24 h prior to the episode ( P =0.234), but the NEMS were significantly lower 24 h after the event among those who successfully tolerated UE compared to those who failed UE (14.1 vs 33.9, P <0.001). Patients with failed UE scored significantly higher points in MODS at both time points when compared to those who tolerated this episode (6.2 vs 4.1 pre, P =0.003 and 6.0 vs 4.3 post, P =0.017) [Table - 2].

   Discussion Top

We observed a 1.1% incidence of UE in our ICUs, which is less than the reported rates of 2 to 3.7% for surgical ICUs or 7 to 11% for medical ICUs.[2],[3],[4],[15],[16] Our medical-surgical ICUs have a significant number of patients admitted following cardiac surgery and most of these patients have relatively short durations of ventilation. These patients contribute to the denominator of total patients but may not be at the same risk of self-extubation as other patients with longer durations of ventilation. This might have accounted for the low incidence of UE. It is also possible that some of the UE episodes were not captured in our database although it is unlikely that this would change the incidence significantly.

In this study, the APACHE III score was not associated with the increase risk of UE. Although we did not match the two groups for APACHE III score, the comparable score decreases the potential impact of severity of illness on UE. UE was associated with significantly higher incidence of agitation and the cumulative dose of benzodiazepines was much higher 24 h before the UE episodes-findings similar to previous reports.[9],[17] This particular association was not shown with the use of narcotics. Short-acting benzodiazepine boluses are the preferred sedatives in the ICU due to the reliable short-term effect and the lack of significant respiratory depression compared to narcotics. However, their use may potentially be associated with paradoxical reactions resulting in agitation and restlessness in some patients. Whether or not the dosing of benzodiazepines plays a causal role in the incidence of unplanned extubation cannot be determined from a retrospective study. Although the use of hand restraints was not higher among UE patients in our study, Tominaga and colleagues[18] found increased UE events among patients with restricted use of physical restraints. In another report,[8] patients who were self-extubated had greater likelihood of being physically restrained. Although this may just be a marker of agitation, it is possible that the use of restraints might add to the stress and frustration and hence agitation of the critically ill intubated patients. It is also interesting to note that some UE patients, despite being restrained, can still self-extubate by coughing or using their tongue. Although we found no significant difference between the groups regarding the use of restraints, it is possible that this was under-reported in the medical chart.

The UE events in the study group did not result in less ventilation time when compared to patients without these episodes since many patients with UE were re-intubated. This rate is comparable to the rate of re-intubation following UE events reported by many others.[8],[17],[19] Contrary to Jiang et al,[20] who reported a significantly higher re-intubation rate among patients with assist/control ventilatory mode and lower PaO[2]/FiO[2], this relationship was not found between our UE patients compared to their controls. In fact, more patients with UE who required re-intubation were being weaned from mechanical ventilation at the time of UE, but the P -value was not significant. Consistent with a retrospective case-control study by Atkins et al[8] and a more recent prospective study by Epstein and colleagues,[19] we found no increase in mortality in UE patients compared to their controls. However, in contrast to their findings, we have not demonstrated an increase in average ICU or hospital length of stay in patients with UE. It is possible that early re-intubation in many of our self-extubated patients (64% within 12 h and 75% within 24 h) resulted in a lower complication rate from UE. Re-intubation for failed UE was not associated with increasing risk of death but was associated with an increased length of ICU stay. We also found a significant association between the risk of re-intubation following UE and higher APACHE III score, reflecting higher incidence of failed UE among patients who were sicker at the time of ICU admission. Epstein et al[19] similarly reported a higher APACHE II score for patients who failed UE compared with those who successfully tolerated this episode. Thus, patients with greater illness severity on admission or 24 h prior to the UE are more likely to fail UE and require re-intubation.

The similarity in MODS between the two groups, before and after the extubation events, suggests that morbidity is not increased by UE. Worsening MODS is a reflection of worsening clinical condition that might have been affected by an ICU event. Since both groups were comparable in many variables at the time of the UE episodes, it appears that UE extubation does not have an impact on severity of illness. It is possible that we have chosen a narrow range of time to calculate the MODS (within 24 h pre-and post-event) and potentially missed a significant change in MODS. However, the fact that mortality rate, which correlates well with MODS, was not different between the two groups supports our findings. In the largest prospective multicenter study by Boulain and co-workers,[17] together with other reports,[8],[10] UE was also not associated with increased mortality when compared with that of matched controls.

In our study, we could not demonstrate a relationship between the nursing workload, as estimated by NEMS, and the risk of UE. The UE did not appear to increase nursing workload as the mean NEMS did not increase 24 h after the UE episode. In fact, when tolerated, UE usually results in reduced nursing workload owing to the discontinuation of mechanical ventilation, which scores higher points in the NEMS score. Chevron et al[20] reported a similar result using a score that measures the overall nursing workload for the day rather than for the individual patient. Some authors suggested that the frequency of UE was increased when nurse staffing was reduced,[2],[3] but we did not measure the overall staffing in our ICU at the time of UE. Our study has the limitation of the case-control study, which suggests rather than confirms any of the associations between the exposure and the outcome.

   Conclusion Top

Despite the disturbing event of UE in the ICU, we could not demonstrate its significant association with patient mortality or morbidity. The lack of such association raises questions about the usefulness of UE as a quality of care indicator. Recognizing agitation as a risk factor for UE may be a key component in preventing UE. Since many patients with UE are agitated, proper sedation, with the recognition of the potential paradoxical effect related to benzodiazepines, may reduce the incidence of UE. Many patients do successfully tolerate the UE episode and when they do not, it does seem to prolong ICU stay, but in this study, it had no impact on patients' mortality. As well, we have demonstrated that UE has no significant association with nursing workload as measured by NEM score.

   References Top

1.Stauffer J, Olson D, Petty T. Complications and consequences of endotracheal intubation and tracheotomy: A prospective study of 150 critically ill adult patients. Am J Med 1981;70:65-76.   Back to cited text no. 1    
2.Coppolo D, May J. Self-extubations: A 12-month experience. Chest 1990;98:165-9.   Back to cited text no. 2    
3.Listello D, Sessler C. Unplanned extubation: Clinical predictors for reintubation. Chest 1994;105:1496-503.   Back to cited text no. 3    
4.Tindol G, DiBenedetto R, Kosciuk L. Unplanned extubations. Chest 1994;105:1804-7.  Back to cited text no. 4    
5.Whelan J, Simpson S, Levy H. Unplanned extubation. Predictors of successful termination of mechanical ventilatory support. Chest 1994;105:1808-12.  Back to cited text no. 5    
6.Vassal T, Anh NG, Gabillet JM, Guidet B, Staikowsky F, Offenstadt G. Prospective evaluation of self-extubations in a medical intensive care unit. Intens Care Med 1993;19:340-2.  Back to cited text no. 6  [PUBMED]  [FULLTEXT]
7.Torres A, Gatell JM, Aznar E, el-Ebiary M, Puig de la Bellacasa J, Gonzalez J, et al . Re-intubation increases the risk of nosocomial pneumonia in patients needing mechanical ventilation. Am J Respir Crit Care Med 1995;152:137-41.  Back to cited text no. 7    
8.Atkins PM, Mion LC, Mendelson W, Palmer RM, Slomka J, Franko T. Characteristics and outcomes of patients who self-extubate from ventilatory support: A case-control study . Chest 1997;112:1317-23.  Back to cited text no. 8  [PUBMED]  [FULLTEXT]
9.Tung A, Tadimeti L, Caruana-Montaldo B, Atkins PM, Mion LC, Palmer RM, et al . The relationship of sedation to deliberate self-extubation. J Clin Anesth 2001;13:24-9.  Back to cited text no. 9    
10.Epstein SK, Nevins ML, Chung J. Effect of unplanned extubation on outcome of mechanical ventilation. Am J Respir Crit Care Med 2000;161:1912-6.  Back to cited text no. 10    
11.Chiang AA, Lee KC, Lee JC, Wei CH. Effectiveness of a continuous quality improvement program aiming to reduce unplanned extubation: A prospective study. Intens Care Med 1996;22:1269-71.  Back to cited text no. 11  [PUBMED]  [FULLTEXT]
12.Maguire GP, DeLorenzo LJ, Moggio RA. Unplanned extubation in the intensive care unit: A quality-of-care concern. Crit Care Nurs Q 1994;17:40-7.  Back to cited text no. 12  [PUBMED]  [FULLTEXT]
13.Miranda D, Moreno R, Lapichino G. Nine equivalents of nursing manpower use score (NEMS). Intens Care Med 1997;23:760-5.  Back to cited text no. 13    
14.Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: A reliable descriptor of a complex clinical outcome. Crit Care Med 1995;23:1638-52.  Back to cited text no. 14  [PUBMED]  [FULLTEXT]
15.Taggart JA, Linf MA. Evaluating unplanned endotracheal extubation. Dimens Crit Care Nurs 1994;13:114-20.  Back to cited text no. 15    
16.Lamb B, Vogelson M, Trak K. Incidence of unplanned extubation. Crit Care Med 1989;17:S96.  Back to cited text no. 16    
17.Boulain T. Unplanned extubation in adult intensive care unit. Am J Respir Crit Care Med 1998;157:1131-7.  Back to cited text no. 17  [PUBMED]  [FULLTEXT]
18.Tominaga GT, Rudzwick H, Scannell G, Waxman K. Decreasing unplanned extubation in the surgical intensive care unit. Am J Surg 1995;170:586-9.  Back to cited text no. 18  [PUBMED]  [FULLTEXT]
19.Epstein SK, Ciubotaru RL, Wong JB. Effect of failed extubation on the outcome of mechanical ventilation. Chest 1997;112:186-92.  Back to cited text no. 19  [PUBMED]  [FULLTEXT]
20.Chevron V, Menard JF, Richard JC, Girault C, Leroy J, Bonmarchand G. Unplanned extubation: Risk factors of development and predictive criteria for reintubation. Crit Care Med 1998;26:1049-53.  Back to cited text no. 20  [PUBMED]  [FULLTEXT]


  [Figure - 1], [Figure - 2]

  [Table - 1], [Table - 2]

This article has been cited by
1 Unplanned versus planned extubation in respiratory intensive care unit, predictors of outcome
Mohammed F. Ismaeil,Hanan M. El-Shahat,Mohamed S. El-Gammal,Ahmed M. Abbas
Egyptian Journal of Chest Diseases and Tuberculosis. 2013;
[Pubmed] | [DOI]
2 Pressure-regulated volume controlled ventilation in acute respiratory failure of pulmonary diseases
M.E. Abou Shehata,Amina M. Abd El-Maksoud,R.A. Elmetwally
Egyptian Journal of Chest Diseases and Tuberculosis. 2012; 61(3): 151
[Pubmed] | [DOI]
3 A human factors framework and study of the effect of nursing workload on patient safety and employee quality of working life
Holden, R.J., Scanlon, M.C., Patel, N.R., Kaushal, R., Escoto, K.H., Brown, R.L., Alper, S.J., (...), Karsh, B.-T.
Quality and Safety in Health Care. 2011; 20(1): 15-24
4 Predictors and outcome of unplanned extubation in COPD patients
Elmetwally, R., Elshafey, M., Zeidan, A., Makharita, M.Y.
Egyptian Journal of Anaesthesia. 2010; 26(1): 1-11
5 Influence of Physical Restraint on Unplanned Extubation of Adult Intensive Care Patients: A Case-Control Study
Chang, L.Y. and Wang, K.W.K. and Chao, Y.F.
American Journal of Critical Care. 2008; 17(5): 408
6 Influence of Physical Restraint on Unplanned Extubation of Adult Intensive Care Patients: A Case-Control Study
Li-Yin Chang, Kai-Wei Katherine Wang, Yann-Fen Chao
American Journal of Critical Care. 2008; 17(5): 408
[Pubmed] | [DOI]


Print this article  Email this article
   Next article
   Previous article 
   Table of Contents
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Article in PDF (150 KB)
    Citation Manager
    Access Statistics
    Reader Comments
    Email Alert *
    Add to My List *
* Registration required (free)  

    Materials and Me...
    Article Figures
    Article Tables

 Article Access Statistics
    PDF Downloaded1316    
    Comments [Add]    
    Cited by others 6    

Recommend this journal