Self-Extubation in Patients with Traumatic Head Injury: Determinants, Complications, and Outcomes.

BACKGROUND: Self-extubation is a common clinical problem associated with mechanical ventilation in trauma patients worldwide.
OBJECTIVES: This study aimed to evaluate the predisposing factors, complications, and outcomes of self-extubation in patients with head injury.
METHODS: This was a retrospective cohort study. SETTINGS: The study was conducted in a trauma intensive care unit (TICU). PATIENTS: All intubated patients with head injury admitted to TICU between 2013 and 2015 were included in the study.
INTERVENTIONS: Planned compared to selfextubation during weaning from sedation. MEASUREMENTS: Risk, predictors, and outcomes of self-extubation were measured. MAIN
RESULTS: A total of 321 patients with head injury required mechanical ventilation, of which 39 (12%) had self-extubation and 12 (30.7%) had reintubation. The median Glasgow Coma Scale, head abbreviated injury score, and injury severity score were 9, 3, and 27, respectively. The incidence of self-extubation was 0.92/100 ventilated days. Self-extubated patients were more likely to be older, develop agitation (P = 0.001), and require restraints (P = 0.001) than those who had planned extubation. Furthermore, self-extubation was associated with more use of propofol (P = 0.002) and tramadol (P = 0.001). Patients with self-extubation had higher Ramsay sedation score (P = 0.01), had prolonged hospital length of stay (P = 0.03), and were more likely to develop sepsis (P = 0.003) when compared to the planned extubation group. The overall in-hospital mortality was significantly higher in the planned extubation group (P = 0.001). Age-adjusted predictors of self-extubation were sedation use (adjusted odds ratio [aOR]: 0.06; P = 0.001), restraint use (aOR: 10.4; P = 0.001), and tramadol use (aOR: 7.21; P = 0.01).
CONCLUSIONS: More than one-tenth of patients with traumatic head injury develop self-extubation; this group of patients is more likely to have prescribed tramadol, develop agitation, and have longer hospital length of stay and less sedation use. Further prospective studies are needed to assess the predictors of self-extubation in TICU. Copyright:

INTRODUCTION

Self-extubation or unplanned extubation is a frequent clinical problem associated with mechanical ventilation in trauma intensive care units (TICUs) worldwide.[1] Self-extubation is self-removal of the endotracheal tube either deliberately or accidentally by a mechanically ventilated patient.[2] Notably, the incidence rate of unplanned extubation ranges from 0.5% to 14.2%.[3] Among all cases of unplanned extubation, patient-induced self-extubation accounts for most cases (60%–90%). On the other hand, the accidental extubation mainly occurs due to coughing and change of position by the patient or during nursing procedures at the bedside.[2] A prospective multicenter study of 426 ventilated patients reported unplanned extubation in 10.8% of cases (57 episodes), of which 55 episodes occurred due to self-extubation and only 2 episodes accidentally caused by health-care members including nurses.[4]

Use of sedatives and analgesics to prevent self-extubation remains challenging as overprescribed sedation increases the risk of pneumonia and light sedation is associated with higher rates of anxiety and self-extubation.[5] Agitation due to inadequate sedation is the most frequent risk factor for unplanned extubation (50%–74%) in critically ill patients.[6] Notably, sedation strategies favoring agitation may increase the risk of self-extubation by 26% than those who did not experience agitation.[7] Neurological status in terms of consciousness level measured as Glasgow Coma Scale (GCS) of 9–12 is also reported as a risk factor of unplanned extubation as more alert patients are more likely to develop agitation and self-extubation.[89] The other patient-related factors include male gender, old age (>69 years), inadequate sedation, and smoking.[10] In addition, self-extubation is related to specific conditions such as pulmonary and cardiac diseases, and also, it may occur in patients admitted to the ICU after surgery or who had burn-related injuries.[3] Furthermore, nursing care significantly affects the incidence of self-extubation as the rate of unplanned extubation occurred more frequently when inexperienced nurses were on duty.[11] Self-extubation may cause damage to the larynx, hypotension, bronchospasm aspiration, bleeding, and edema.[10] Moreover, around 20% of patients experience difficulty in reintubation following self-extubation.[3] However, the factors contributing to self-extubation and its associated complications are not well recognized and need to be explored further in our trauma center. Herein, we aim to evaluate the clinical characteristics, predisposing factors, complications, and outcomes among patients with traumatic head injury who underwent self-extubation in the ICU.

METHODS

A retrospective cohort study was conducted to include adult intubated head injury patients who required mechanical ventilatory support and were admitted to the TICU at Hamad General Hospital between January 2013 and December 2015. The data were collected from the trauma registry database at Hamad Trauma Center (HTC) for all patients who were extubated during the study period. The HTC trauma registry is a database that participates in both the National Trauma Data Bank and the Trauma Quality Improvement Program of the American College of Surgeons-Committee on Trauma.

Collected data included patients’ demographics (age, gender, and nationality), mechanism of injury, injury severity score (ISS), types of head injury lesions, tracheostomy, agitation, use of restraints, GCS, head abbreviated injury score (AIS), ISS, oxygen saturation post-self-extubation, need for reintubation, use of controlled medication including tramadol, use of analgesics (fentanyl and remifentanil), medication being used for sedation (propofol and midazolam), ICU length of stay, ventilator days, hospital length of stay, complications (pneumonia, acute respiratory distress syndrome, and sepsis), and mortality. Ventilatory mode and parameters were set according to the attending physician. Self-extubation refers to the removal of the endotracheal airway tube deliberately by the patient during the weaning process.[5] Accidental extubation at the time of repositioning or procedures at the bedside by the medical staff was not considered in the present study.

Ramsay sedation score (RSS) was used to clinically assess the depth of sedation or level of consciousness (six levels: four levels of sedation defined by responses to stimuli [levels 3–6], a level of “cooperative, oriented, and tranquil” [level 2], and a level for “anxious, agitated, or restless” [level 1]).[12] This study was granted ethical approval from the medical research center at Hamad Medical Corporation, Qatar, with a waiver of consent as data were collected anonymously and retrospectively (IRB# 16366/16). This study follows the STROBE checklist for observational study [Supplementary Table 1].

Supplementary Table 1

STROBE Statement—checklist of items that should be included in reports of observational studies

	Item no	Recommendation
Title and abstract	1	(a) Indicate the study’s design with a commonly used term in the title or the abstract Page 1,2,3
		(b) Provide in the abstract an informative and balanced summary of what was done and what was found Page 2,3
Introduction
Background/rationale	2	Explain the scientific background and rationale for the investigation being reported Pages 4 and 5
Objectives	3	State specific objectives, including any prespecified hypotheses Page 5
Methods
Study design	4	Present key elements of study design early in the paper Page 5-6
Setting	5	Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection Page 5-6
Participants	6	(a) Cohort study - Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Case-control study - Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls Cross-sectional study - Give the eligibility criteria, and the sources and methods of selection of participants Page 5-6
		(b) Cohort study - For matched studies, give matching criteria and number of exposed and unexposed Case-control study - For matched studies, give matching criteria and the number of controls per case Not applicable
Variables	7	Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if applicable Page 5-6, tables, fig
Data sources/measurement	8*	For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group Page 5-6
Bias	9	Describe any efforts to address potential sources of bias Page 6-7
Study size	10	Explain how the study size was arrived at Page 6
Quantitative variables	11	Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and why Page 6,7
Statistical methods	12	(a) Describe all statistical methods, including those used to control for confounding Page 6,7
		(b) Describe any methods used to examine subgroups and interactions Tables &figures
		(c) Explain how missing data were addressed 7
		(d) Cohort study - If applicable, explain how loss to follow-up was addressed Case-control study - If applicable, explain how matching of cases and controls was addressed Cross-sectional study - If applicable, describe analytical methods taking account of sampling strategy
		(e) Describe any sensitivity analyses Tab&fig
Results
Participants	13*	(a) Report numbers of individuals at each stage of study - e.g., numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed
		(b) Give reasons for non-participation at each stage
		(c) Consider use of a flow diagram Page 7 &8 figure
Descriptive data	14*	(a) Give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential confounders
		(b) Indicate number of participants with missing data for each variable of interest
		(c) Cohort study - Summarise follow-up time (eg, average and total amount) Page 7-8
Outcome data	15*	Cohort study - Report numbers of outcome events or summary measures over time
		Case-control study - Report numbers in each exposure category, or summary measures of exposure
		Cross-sectional study - Report numbers of outcome events or summary measures Page 7-9
Main results	16	(a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included
		(b) Report category boundaries when continuous variables were categorized
		(c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period Page 7-9
Other analyses	17	Report other analyses done - e.g., analyses of subgroups and interactions, and sensitivity analyses
Discussion
Key results	18	Summarise key results with reference to study objectives Pages 9-14
Limitations	19	Discuss limitations of the study, taking into account sources of potential bias or imprecision. Discuss both direction and magnitude of any potential bias Page 14
Interpretation	20	Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence Pages 9-14
Generalisability	21	Discuss the generalisability (external validity) of the study results Pages 9-14
Other information
Funding	22	Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based Not Applicable

*Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies. Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org.

Statistical analysis

Data were presented as proportions, medians, or mean ± standard deviation, as appropriate. The data were analyzed and compared between patients who had successfully planned extubation compared to self-extubation during weaning from sedation. Differences in categorical and continuous variables were analyzed using the Chi-square test and Student's t-test, as appropriate. Yates’ corrected Chi-square was used for categorical variables, if the expected cell frequencies were below 5. Multivariate logistic regression analysis was performed to identify the independent risk factors for self-extubation after adjustment for the potential covariates that showed significant associations in the univariate analysis. The data were presented as adjusted odds ratio (aOR) and 95% confidence interval (CI). Two-tailed P < 0.05 was considered statistically significant. All data analyses were carried out using the Statistical Package for the Social Sciences, version 22 (SPSS, Inc., Chicago, IL, USA).

RESULTS

During the study, a total of 321 head injury patients who required mechanical ventilation were admitted to the TICU, of which 282 (88%) had successfully planned extubation and 39 (12%) had self-extubation. The mean age of patients was 32 ± 13 years, and the majority of patients (93.8%) were male. The median and range of GCS, head AIS, and ISS were 9 (3–15), 3 (1–7), and 27 (5–75), respectively. The incidence of self-extubation was 0.92/100 ventilated days. Traffic accidents (67.3%) were the major cause of head trauma, followed by fall from height (24.3%) and fall of heavy objects (2.5%) [Table 1]. The commonly identified type of head injuries was brain contusion (64.8%) and skull fracture (61.7%). Figure 1 shows the proportion of self-extubation according to the type of head injury lesion.

Table 1

Comparison of demographics, mechanism of injury, and brain lesions among intubated traumatic brain injury patients admitted to trauma intensive care unit by extubation status

	Overall (n=321), n (%)	Planned extubation (n=282; 88%), n (%)	Self-extubation* (n=39; 12%), n (%)	P
Age	32.3±13.6	31.6±13.5	37.3±12.9	0.01
Males	301 (93.8)	264 (93.6)	37 (94.9)	0.76
Qatari nationals	62 (19.3)	53 (18.8)	9 (23.1)	0.52
Mechanism of injury
Traffic accidents	216 (67.3)	193 (68.4)	23 (59.0)	0.69 for all
Fall from height	78 (24.3)	66 (23.4)	12 (30.8)
Fall of heavy object	8 (2.5)	6 (2.1)	2 (5.1)
Assault	7 (2.2)	6 (2.1)	1 (2.6)
All-terrain vehicles	4 (1.2)	4 (1.4)	0
Others	8 (2.5)	7 (2.5)	1 (2.6)
TBI lesions
Brain contusion	208 (64.8)	186 (66.0)	22 (56.4)	0.24
Skull fracture	198 (61.7)	175 (62.1)	23 (59.0)	0.71
Subdural hemorrhage	127 (39.6)	108 (38.3)	19 (48.7)	0.21
Subarachnoid hemorrhage	126 (39.3)	110 (39.0)	16 (41.0)	0.80
Brain edema	119 (37.1)	106 (37.6)	13 (33.3)	0.60
Epidural hematoma	66 (20.6)	59 (20.9)	7 (17.9)	0.66
Diffuse axonal injury	67 (20.9)	55 (19.5)	12 (30.8)	0.10
Pneumocephalus	48 (15.0)	43 (15.2)	5 (12.8)	0.81
Intraventricular hemorrhage	44 (14.7)	42 (14.9)	2 (5.1)	0.13
Extra-axial hemorrhage	20 (6.2)	17 (6.0)	3 (7.7)	0.72
Intracerebral hemorrhage	5 (1.6)	5 (1.8)	0	1.00
Tracheostomy	92 (28.7)	82 (29.1)	10 (25.6)	0.65
Agitation	29 (9.0)	9 (3.2)	20 (51.3)	0.001
Physical restraints	58 (18.1)	30 (10.7)	28 (71.8 )	0.001

*12 patients had reintubation. TBI=Traumatic brain injury

Figure 1

Proportion of self-extubation according to the type of head injury lesion

Around 28.7% of patients had tracheostomy performed and 9% experienced agitation during weaning. Reintubation was performed in 31% of cases who had self-extubation.

Table 1 analyzes the clinical characteristics of traumatic brain injury (TBI) patients with successfully planned extubation compared to self-extubation. No significant differences in the baseline characteristics, mechanism of injury, and types of TBI lesions were found between the two groups, except for the fact that self-extubated patients tend to be older in age than those who had successfully planned extubation (P = 0.01). The rate of self-extubation was comparable between those who had tracheostomy compared to no tracheostomy (11% compared to 13%; P = 0.66). Only one-third of self-extubation patients required reintubation (31% compared to 0%). Moreover, self-extubated patients were more likely to be agitated (51.3% compared to 3.2%; P = 0.001) and physically restrained (71.8% compared to 10.7%; P = 0.001) as compared to the planned extubation cohort.

Table 2 shows the management and outcomes among TBI patients by extubation status. In comparison to the planned extubation group, patients with self-extubation had significantly higher GCS in ICU (P = 0.001) and RSS (P = 0.01) and lower positive end-expiratory pressure before extubation (P = 0.04). Tramadol was administered in 5% of head injury patients overall during weaning, and there was a significantly higher rate of tramadol use in the self-extubation group (30.8% compared to 1.4%; P = 0.001). Fentanyl (84.6%) and remifentanil (15.4%) were used as analgesics. The frequently used sedatives included propofol (72.1%) and midazolam (27.9%). The two groups were comparable for the use of analgesics. However, self-extubation was more frequently associated with the use of propofol (92.3% compared to 68.3%; P = 0.001). In addition, patients who had self-extubation stayed longer in the hospital (36 [15-157] compared to 30 days [5-361]; P = 0.03) and were more likely to develop sepsis (41.0% compared to 19.9%; P = 0.003) as compared to those who had a successfully planned extubation.

Table 2

Management and outcomes among traumatic brain injury patients by extubation status

	Overall (n=321)	Planned extubation (n=282)	Self-extubation (n=39)	P
GCS ICU	9.1±3.7	8.9±3.9	10.1±1.1	0.001
Head AIS	4 (1-9)	4 (1-9)	4 (2-5)	0.94
Injury severity score	27 (5-75)	27 (9-75)	29 (5-50)	0.77
Ramsay sedation score	1 (1-5)	1 (1-2)	2 (1-5)	0.01
PEEP level	5.5±0.7	5.6±0.9	5.2±0.4	0.04
Tramadol, n (%)	16 (5.0)	4 (1.4)	12 (30.8)	0.001
Analgesics (n=253), n (%)
Fentanyl	214 (84.6)	178 (83.2)	36 (92.3)	0.22 for all
Remifentanil	39 (15.4)	36 (16.8)	3 (7.7)
Sedatives (n=244), n (%)
Propofol	176 (72.1)	140 (68.3)	36 (92.3)	0.001 for all
Midazolam	68 (27.9)	65 (31.7)	3 (7.7)
ICU length of stay (days)	19 (5-163)	19 (5-163)	19 (9-79)	0.23
Ventilatory days	12 (5-47)	12.5 (5-47)	13 (5-44)	0.24
Hospital length of stay	31 (5-361)	30 (5-361)	36 (15-157)	0.03
Complications, n (%)
Pneumonia	181 (56.4)	154 (54.6)	27 (69.2)	0.08
ARDS	27 (8.4)	23 (8.2)	4 (10.3)	0.65
Sepsis	72 (22.4)	56 (19.9)	16 (41.0)	0.003
Mortality, n (%)	70 (21.8)	69 (24.5)	1 (2.6)	0.001

PEEP=Positive end-expiratory pressure, GCS=Glasgow Coma Scale, ICU=Intensive care unit, AIS=Abbreviated injury score, ARDS=Acute respiratory distress syndrome

Overall, 70 (21.8%) patients died during the hospital course, and the rate of mortality was significantly higher in the planned extubation group as compared to those self-extubated. In comparison to survivors, the deceased were more likely to have higher ISS (32 ± 12 compared to 27 ± 9; P = 0.002) and lower GCS (6 ± 3 compared to 10 ± 3; P = 0.001). Survivors were more likely to have agitation (11.6% compared to 0%), pneumonia (63% compared to 30%), self-extubation (15% compared to 1.5%), and tracheostomy (34% compared to 9%) in comparison to the deceased.

Univariate and multivariate analysis for predictors of self-extubation

Univariate analysis shows that agitation was associated with a 31-fold increase in the rate of self-extubation; however, after adjustment for confounders (i.e., age, GCS, sedation, tramadol, and restraints), this association lost its statistical significance [Table 3]. On multivariate logistic regression analysis, age-adjusted predictors of self-extubation were sedation (aOR: 0.06; 95% CI: 0.011–0.325; P = 0.001), restrain use (aOR: 10.4; 95% CI: 3.164–33.948; P = 0.001), and tramadol use (aOR: 7.21; 95% CI: 1.461–35.589; P = 0.01) [Table 3].

Table 3

Univariate and multivariate analysis for predictors of self-extubation

Variable	Univariate analysis			Multivariate analysis
	OR	95% CI	P	aOR	95% CI	P
Sedation	0.179	0.053-0.604	0.006	0.060	0.011-0.325	0.001
Agitation	31.93	12.803-79.631	0.001	3.435	0.997-11.831	0.051
Restrain use	21.29	9.631-47.094	0.001	10.363	3.164-33.948	0.001
GCS at ICU	1.08	0.987-1.200	0.080	0.94	0.766-1.168	0.60
Tramadol use	30.88	9.317-102.408	0.001	7.21	1.461-35.589	0.015
Age	1.03	1.005-1.049	0.02

OR=Odds ratio, Age - aOR=Adjusted OR, CI=Confidence interval, GCS=Glasgow Coma Scale, ICU=Intensive care unit

DISCUSSION

Several studies have reported the rates of self-extubation in critically ill patients admitted in the medical or surgical ICUs or included trauma patients as a subpopulation. To the best of our knowledge, there are only limited reports trying to identify the predictors of self-extubation in TICU patients. Most of these observational studies on unplanned extubation assessed the data primarily from medical or surgical ICU populations.[1] The rate of unplanned extubation reported in previous surgical ICU studies ranges between 1.8% and 15%.[13]

The present study focuses only on those who had traumatic head injury requiring mechanical ventilation and have several key findings. The rate of self-extubation was found to be 12% among mechanically ventilated head injury patients. Köhne and Hardcastle[1] studied 420 injured patients who require mechanical ventilation and were admitted to the TICU. The authors reported 40 events of unplanned extubation, of that self-extubation occurred in 28 (6.7%) patients, which is less than the events in our study. An earlier study by Schwartz et al.[14] investigated the rates of self-extubation in critically ill adult patients admitted to tertiary care ICUs. In that study, the rate of self-extubation was 7.3%, of which 22.1% had accidental extubation and 77.9% of the cases were reported as deliberate self-extubation. Most of the studies from medical and surgical ICUs have reported incidences of self-extubation higher than 10%.[15]

Notably, reporting factors affecting the rate of unplanned extubation are limited and need more exploration. In our cohort, self-extubated patients were more likely to be older in age, agitated, and had physical restraints. The study from South Africa[1] that was conducted as an audit of the TICU's quality of nursing and medical care revealed an incidence rate of self-extubation of 0.89/100 ventilated days, which is slightly lower than our rate (0.92). The study from South Africa included self-extubation as well as accidental extubation and did not specify the injured body region in contrast to our study which focused on head injury cases and self-extubation. Earlier studies have described advanced age as a risk of unplanned extubation in medical and adult mixed ICU patients;[1617] however, the mean age in the present study population was 32 years.

With respect to agitation, our findings are consistent with earlier reports which also showed that agitation was more frequent in patients with unplanned extubation as compared to those who had planned extubation.[91819] Therefore, agitation in the absence of appropriate sedation should be considered an important determinant of unplanned extubation. Agitated patients are unlikely to have unplanned extubation if their movement is prevented by adequate sedation. Univariate and multivariate analysis in our study showed that agitation is a frequent association, but not an independent predictor of self-extubation.

The reported rate of physical restraint varies in the literature; earlier studies have suggested various adverse outcomes associated with physical restraint, namely delirium, nerve injury, posttraumatic stress disorder, increased ICU stay, and risks of self-extubation.[2021] In line with these observations, the rate of self-extubation was significantly higher in physically restrained patients. Of note, the use of physical restraints can aggravate the occurrence of agitation or delirium and enhances the risk of self-extubation.

A previous retrospective study conducted by Cho and Yeo[22] identified advanced age (≥60 years), agitation, use physical restraints, and nurse-to-patient ratio as the major predictors for deliberate self-extubation. Similarly, another case-controlled study by Tung et al. identified agitation as a significant factor that increases the risk of self-extubation.[23] Atkin et al.[24] showed that agitation, anxiety, restlessness, and occurrence of adverse events in the hospital are independently associated with the risk of self-extubation.

It has been suggested that a higher level of consciousness is associated with a higher risk of unplanned extubation.[25] In our cohort, self-extubated patients had a higher level of consciousness in terms of GCS and RSS, which is consistent with a previous report that demonstrated higher GCS scores at hospital admission in patients with unplanned extubation.[16] Curry et al.[26] suggested that self-extubation occurs more frequently in patients with minimal sedation indicated by a lower RSS, which is consistent with our findings of median RSS score of 2. Notably, maintaining an optimal level of sedation needs regular reassessment. This is likely to be affected during the change-over procedure, which could lead to a decrease in sedation.

Sedative medication is often administered to patients who require ventilatory support in order to minimize the incidence of agitation and anxiety.[13] Suboptimal sedation can lead to restlessness and agitation, thereby increasing the risk of unplanned extubation.[23] In our cohort, approximately 92.3% of patients with self-extubation received propofol for sedation. Our results have shown an association between self-extubation and the frequent use of sedatives and controlled medication. An earlier study by Singh et al.[5] recommended that the nature of sedative medications did not affect the self-extubation rates as those patients who had self-extubation received different combinations of sedatives.

An earlier study from our center reported poor outcome in terms of agitation and extubation failure among TBI patients who received tramadol.[27] In the present study, higher proportion of patients who received tramadol were reported to have self-extubation.

There are several consequences of unplanned extubation, including severe airway complications such as bronchospasm, aspiration pneumonia, laryngeal bleeding, vocal cord damage, and respiratory distress, all of which may necessitate reintubation.[1028] In our cohort, among the self-extubated patients, 12 (30.8%) required reintubation. The reported rates of reintubation after unplanned extubation are 36%–57% for the 1st h and 37%–57% within 48 h;[111] our data corroborate with these previous findings. Moreover, self-extubated patients had prolonged hospital stay and sepsis. Consistent with our findings, previous reports have demonstrated that self-extubated patients were more likely to need reintubation, prolonged ventilatory support, and increased ICU and hospital length of stay.[724] Another prospective observational study by Hsiung Lee et al.[29] reported that unplanned extubation, noninvasive ventilation after extubation, and sepsis are the independent risk factors of failed extubation. Given these potential complications, preventive strategies are essential to decrease the incidence of self-extubation. Earlier reports have suggested that patients with unplanned extubation are at increased risk of a poor hospital course with increased length of ventilatory support, ICU, and hospital stay and higher mortality.[424] In contrast, in the present study, the rate of in-hospital mortality was significantly higher in the planned extubation group as compared to self-extubation patients.

On the multivariate logistic analysis, use of physical restraints, use of tramadol, and lack of sedation were significant predictors of self-extubation in our cohort.

In mechanically ventilated patients, physical restraints are primarily used to prevent self-extubation. Conversely, previous studies showed that the use of physical restraints is associated with an increased risk of unplanned extubation,[1121] which is in agreement with our findings. To reduce the risk of self-extubation, a patient's hands should be kept at least 20 cm away from the tubes.[11] Prior studies showed that although a 30° elevation of the head of the bed is common to avoid aspiration and promote ventilation, it increases the chance of self-extubation.[11] This study is limited by more than one factor. First, the retrospective nature of the study with possible missing information should be considered. Second, the relatively small sample size may affect the power of the study. Third, the study lacks information on the ICU nurses’ experience, supervision, and incidence of self-extubation during shift changes. Fourth, information on accidental extubation, which is greatly influenced by organizational factors and the ICU environment, was not considered. Two-thirds of self-extubation patients required reintubation, which raises questions for proper reassessment of our current extubation practice and internal policy. Finally, this study included only one ICU in a single institution which might lack generalizability. However, this study deals with subset of ICU with high-risk homogenous population (i.e., head injury) that could be deteriorated after the unplanned extubation and its consequences.

CONCLUSIONS

Self-extubation in head injury patients is significantly associated with agitation, use of restraints, type and level of sedation, prolonged inpatient care, and complications. Our study highlights the importance for the assessment of and continuous monitoring of the level of agitation and sedation to prevent the incidence of self-extubation. Future analysis should take into consideration patient safety and quality of patient care. In addition, hospital settings should include preventative measures to decrease such complications. Preventative measures should include providing staff education, increasing staff during nursing interventions, researching for risk indicators of self-extubation, using effective sedation protocols, and utilizing alternative protocols in place of physical and pharmacological restraints. Therefore, future research focusing on a prospective design involving a larger cohort is needed to validate the current findings.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.