Prevention of unplanned endotracheal extubation in intensive care unit: An overview of systematic reviews

Abstract Aims This study was performed to identify and summarize systematic reviews focusing on the prevention of unplanned endotracheal extubation in the intensive care unit. Design Overview of systematic reviews. Methods This overview was conducted according to the Preferred Reporting Items for Overviews of Systematic Reviews, including the harms checklist. A literature search of PubMed, the Cochrane Library, CINAH, Embase, Web of Science, SINOMED and PROSPERO was performed from January 1, 2005–June 1, 2021. A systematic review focusing on unplanned extubation was included, resulting in an evidence summary. Results Thirteen systematic reviews were included. A summary of evidence on unplanned endotracheal extubation was developed, and the main contents were risk factors, preventive measures and prognosis. The most important nursing measures were restraint, fixation of the tracheal tube, continuous quality improvement, psychological care and use of a root cause analysis for the occurrence of unplanned endotracheal extubation. Conclusions This overview re‐evaluated risk factors and preventive measures for unplanned endotracheal extubation in the intensive care unit, resulting in a summary of evidence for preventing unplanned endotracheal extubation and providing direction for future research. Trial registration details The study was registered on the PROSPERO website.


| INTRODUC TI ON
Unplanned endotracheal extubation (UEE) is defined as any dislodging of an endotracheal tube from the trachea that is not intentional or ordered by a healthcare professional (Klugman et al., 2020). UEE can lead to serious complications such as asphyxia, malignant arrhythmia and aspiration pneumonia (Chao, Sung, et al., 2017) and serves as an important and sensitive indicator of quality of care (Chuang et al., 2015). A meta-analysis of 34 studies shows that the pooled prevalence of UEE is 6.69% (Li et al., 2022). Notably, 70% of UEE events are preventable (da Silva & Fonseca, 2020).

| BACKG ROU N D
Care bundles are considered as valuable and proven to affect care quality (da Silva et al., 2022), but finding enough systematic and valid evidence to achieve a high bundle fidelity is challenging. In recent years, growing numbers of original studies and systematic reviews are focusing on UEE; however, the quality of available nursing studies related to UEE prophylaxis is mixed, and single studies or systematic reviews provide only scattered evidence on individual issues with no currently available relevant clinical practice guidelines or accepted best practices (Crezeé et al., 2017). A systematic review synthesizes evidence at a higher level and contains a richer and more comprehensive body of information with greater clinical utility (Ji et al., 2020). In the current situation of the growing number of systematic reviews related to UEE, an overview of systematic reviews is recommended as a suitable and logical step for comparison and contrast of individual reviews (Connolly et al., 2015) and can improve the scientific accuracy and validity of evidence-based practice for preventing UEE (Li et al., 2019).
This overview of systematic reviews related to prevention of UEE was conducted according to the Preferred Reporting Items for Overviews of Systematic Reviews, including the harms (PRIOharms) checklist (Bougioukas et al., 2018).
Research questions: • What are the current systematic reviews on UEE in the intensive care unit?
• What are the best measures to prevent UEE in the intensive care unit?

| ME THODS
The protocol of this overview has been registered on the PROSPERO website, and the registration number is CRD42021253097. This overview of systematic review was reported according to the PRIO-harms.

| Literature inclusion and exclusion criteria
3.1.1 | Literature inclusion criteria • The study subjects were intensive care unit patients.
• The study topics were related to unplanned extubation of tracheal intubation in the intensive care unit, including risk factors and preventive measures for unplanned extubation of tracheal intubation, etc.
• The type of literature was systematic review or meta-analysis of all study designs.
• Both English and non-English language articles were considered.
• Literature for which the full text was not available.

| Search methods
The search was conducted in June-July 2021 by the first author and the second author, and we searched the literature from the time the database was built to June 1, 2021, using a subject term plus free word search combined with a literature traceability method. A presearch was conducted in the PubMed database prior to the formal search, and the research team discussed and analysed the search results and revised the search strategy before conducting a full formal search.

| Selection of reviews
Two researchers independently selected the studies identified by the search strategies. The researchers eliminated a study based on its title and abstract and then independently assessed the eligibility of the remaining studies based on their full text. Disagreement was resolved by consensus.

| Data extraction
Two researchers trained in Evidence-Based Nursing independently screened the literature, imported their respective searches into EndNote X9 software (Clarivate, London, UK) for initial screening to exclude duplicates, manually ranked the literature to remove duplicates and independently extracted information based on the inclusion criteria. The extracted data were then compared, and a third party was asked to arbitrate when disagreement was encountered to ensure the accuracy of the information. In the case of non-English articles, two native speakers translated them independently and the translation articles were verified by a third party. Information was extracted by author, year, country/region, institution, content/subject, type of study included, number of studies included, quality assessment tool and primary outcome by creating a uniform form. The form was validated on a trial basis before the reviewers began to independently collect data. Two reviewers extracted data from both articles, compared their answers and discussed their findings to ensure that they both interpreted the questions in the same way.

| Quality assessments of included reviews
Systematic reviews considered eligible for inclusion were appraised for their methodological quality using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Systematic Reviews and Research Synthesis (Aromataris et al., 2015). The JBI Critical Appraisal Checklist is a tool that evaluates both quantitative and qualitative reviews based on principles common across accepted quality assessment tools. It contains 11 questions related to the systematic review method, and each question is answered as "yes," "no," "unclear" or "not applicable." Two reviewers worked in pairs again and finished the assessment independently in duplicate.
Before beginning an independent assessment, the two reviewers piloted the form, testing it on two studies; these reviewers then compared their results and discussed the acceptable level of information required to decide whether a review fits the criteria and what is considered "unclear." The systematic review was considered of insufficient quality for inclusion and thus excluded at this step if at least one of the following was rated "unclear" or "not applicable": "clear review questions," "appropriate inclusion criteria," "appropriate search strategy" or "appropriate critical appraisal criteria."

| Data analysis and synthesis
The included literature was read and analysed iteratively by two researchers, and the extracted data were presented in tabular form to consolidate and descriptively analyse the findings and form the results. Because of the heterogeneity of the included systematic reviews/meta-analyses and the fact that data of results from some studies were not merged, only a descriptive analysis was performed. In addition, evidence related to prophylaxis for UEE was graded for quality using the JBI Levels of Evidence and Grades of Recommendation (Munn et al., 2014). The recommended strengths were determined through group discussion based on the "feasibility, suitability, significance and effectiveness" of the evidence.

| Ethical Statement
The Research Ethics Committee approval was not required.

| Characteristics of included studies
Meta-analyses were performed on 6 of the 13 included systematic reviews (Ai et al., 2018;da Silva et al., 2017;Gardner et al., 2005;Liu et al., 2019;Yan et al., 2016;Zhu et al., 2014); the others underwent descriptive analyses. The methodological aspects of systematic review production mainly involved the following four methodological quality assessment tools: The Newcastle-Ottawa  Table 1, which shows the characteristics of the included systematic reviews.

| Methodological quality of the included systematic reviews
Methodological quality was determined using a critical evaluation checklist tool by the JBI (see Table 2, which shows the results of the critical appraisal as a whole). The critical appraisal of the methodological quality of the included reviews varied across the studies, with three studies meeting 100% of the criteria, six studies meeting >90% of the criteria and four studies meeting >81% of the criteria. The most commonly inapplicable item was related to the assessment of publication bias (Q9), which was assessed in 7 (54%) of the 13 studies. The most commonly unmet item was recommendations for policy and/or practice supported by the reported data (Q10), which was assessed in 6 (46%) of the 13 studies. Other items (Q1-Q8 and Q11) were fulfilled in all or nearly all the studies.

| Risk factors for UEE
Eight studies involved risk factors for UEE (Ai et al., 2018;da Silva et al., 2017;da Silva & Fonseca, 2012;Kiekkas et al., 2013;Lucas da Silva & de Carvalho, 2010;Silva et al., 2013), which were male sex (Ai et al., 2018;, delirium (Ai et al., 2018;, restraint (Ai et al., 2018;Kiekkas et al., 2013), a higher Glasgow Coma Scale (GCS) score (Ai et al., 2018;  • Psychological care and root cause analysis. Psychological care and root cause analysis can reduce the incidence of UEE. A meta-analysis (Yan et al., 2016) that included eight randomized controlled trials showed that psychological care reduced the incidence of UEE. Another meta-analysis (Liu et al., 2019) that included 12 cohort studies showed that root cause analysis reduced the incidence of UEE.
The use of restraints to prevent UEE has been controversial.
Restraint in patients with a GCS score of >9 increased the risk of UEE by a further 6.16 times . Restraint on its own was not effective in preventing UEE, and relevant nursing interventions in conjunction with restraint were more effective than restraint alone . Restraint was one effective way to reduce UEE (Kiekkas et al., 2013).
• Securing the tracheal tube with suitable fixation methods.
An optimal method of tracheal intubation fixation has not yet been established. Two studies included tracheal intubation fixation; one was for adults published in 2005 (Gardner et al., 2005) and the other for newborns in 2014 (Lai et al., 2014). Many methods for fixation of tracheal intubation are both local and specific, and there is not enough information to prove which is the best method for tracheal intubation fixation for adults. And more randomized controlled trials related to tracheal intubation fixation need to be carried out to evaluate the cost and duration of treatment (Gardner et al., 2005). There is a lack of evidence to determine the most effective and safe method of stabilizing the tracheal tube in ventilated newborns (Lai et al., 2014).
Continuous quality improvement has been proved an effective method to reduce the UEE. Two studies show that continuous quality TA B L E 2 The methodological quality of systematic reviews assessed by the JBI critical appraisal checklist for systematic reviews and research syntheses • Management of sedation and risk factors.
Sedation management and enhanced monitoring of high-risk groups and high-risk time periods can reduce the incidence of UEE . One study revealed the need to enhance monitoring during high-risk times to prevent UEE .
Although many studies have focused on UEE, few have examined UEE prevention strategies and related clinical trials (da Silva & Fonseca, 2012). In the present overview, there was a lack of UEE risk assessment tools and studies related to the timing and frequency of assessment among the included studies.

| Summary of evidence and recommended levels
Based on the JBI Levels of Evidence and Grades of Recommendation, we compiled a summary of evidence about the results and levels of evidence of the studies in this overview (see Table 3). The evidence  Silva et al., 2017) and irritability (Silva et al., 2013) Level3 A 3 Poor fixation of the tracheal tube (Silva et al., 2013) Level3 A 4

TA B L E 3 Summary of evidence and recommended levels
The high-risk occurrence times for UE are night, 1 h before and after shift changeover and ventilator withdrawal stage .
The patient was confused, male, physically limited, and with a high GCS (Ai et al., 2018) Level3 A 6 Monitoring, nurse-patient ratio and catheter fixation, physical restraint remains controversial (Kiekkas et al., 2013)  Reintubation is an important outcome indicator (Kiekkas et al., 2013) Level3 A 2 Outcome indicators include: mortality, incidence of perioral skin trauma (Lai et al., 2014) Level1 A Note: The methods of grading and recommendations of the evidence: Evidence was graded for quality using the JBI Levels of Evidence and Grades of Recommendation (Munn et al., 2014). The recommended strengths were determined through group discussion based on the "feasibility, suitability, significance and effectiveness" of the evidence. Evidence level: The evidence level is divided into five levels: level 1 indicates the highest evidence studies, level 3 moderate evidence studies, and level 5 the lowest evidence studies. Recommended strengths: A represents the strong recommendation, and B represents the weak recommendation.
was graded into five levels based on the JBI evidence levels and recommendation levels. The strengths of evidence decrease from level 1-level 5, with level 1 indicating the highest evidence studies and level 5 being the lowest level of evidence. Based on the JBI FAME structure, the strength of recommendations was independently rated by two researchers who had systematically studied evidencedbased nursing according to the validity, feasibility, appropriateness and clinical significance of the evidence, and they performed this rating in accordance with the JBI recommendation strength grading principle of the evidence. In cases of disagreement, a third person arbitrated the dispute. "A" represents a strong recommendation, and

| DISCUSS ION
To our knowledge, this is the first overview of systematic reviews for UEE in the intensive care unit. We conducted a descriptive analysis of 13 pieces of literature related to UEE. The systematic review related to UEE was mainly concerned with two aspects: risk factors and preventive measures. After quality evaluation of the literature and evidence, 20 relevant pieces of evidence were compiled into evidence summaries. This evidence can provide a useful reference for clinical care and clinical nursing practice.

| Risk factors
With regard to risk factors for UEE, there was general agreement that male sex, delirium, higher GCS scores (Ai et al., 2018;, restraint (Ai et al., 2018;Kiekkas et al., 2013), patient agitation (da Silva et al., 2017;Kiekkas et al., 2013), monitoring, the nurse-patient ratio and catheter fixation (Kiekkas et al., 2013;Silva et al., 2013), and bedside handling (da Silva et al., 2017;Silva et al., 2013) were important risk factors. However, the effects of sex and restraint on the tracheal tube have long been controversial.
One study has suggested that the impacts of age and sex on tracheal intubation are unclear (Zhang & Liu, 2021), whereas the other has shown that UEE is more common in younger people (Kwon & Choi, 2017). In this study, two systematic reviews (Ai et al., 2018; show that men are an important factor, probably because men are often physically stronger than women, and the tracheal tube can be more easily removed from young patients than from elderly patients and children because the young usually have more strength than the old and children. Whether restraint can prevent UEE is debated worldwide. Although restraint is still the preferred option (Zhang & Liu, 2021), UEE still occurs when restraint is used. The existence of this controversy may be due to inappropriate restraint; thus, the method of restraint and the pointer to constraint should be strictly controlled.

| Prevention
5.2.1 | Further research is needed on the specific implementation of psychological care to prevent UEE As modern medicine has changed from a "disease-centred biomedical model" to a "human-centred biopsychosocial medical model" (Miao et al., 2020), psychological care has gradually been emphasized (Chambers et al., 2018;Thomas et al., 2018) and has become an essential part of daily care. Psychological care can reduce the incidence of UEE and is often included in our nursing routines as a necessary care measure (Hay et al., 2019), and our study also con-

| Methods and pointers for restraints should be strictly mastered
The effect of restraints on UEE has been considered controversial (Li et al., 2017;Luk et al., 2014). Despite that, clinical staff has included restraints as a preventive measure against UEE because they give medical staff a sense of security.  (Silva et al., 2013;Wang et al., 2019).
In recent years, however, few high-quality studies have focused on fixation of the tracheal tube. Systematic reviews of fixation methods for tracheal intubation in adults and infants were conducted by Gardner et al. (Gardner et al., 2005) and Lai et al. (Lai et al., 2014) (Klugman et al., 2020). A similar study shows that after 15 years of continuous quality improvements, the unplanned extubation rate decreased from 9.0/1000 ventilator days-1.36/1000 ventilator days .

Summary of the evidence
Studies related to UEE are numerous and fragmented, but there are few such forms of evidence aggregation for systematic review. In our research, the evidence about UEE was summarized and marked with the level of evidence and recommendation. Specific evidence was extracted from three aspects of UEE: risk factors, prevention and outcome indicators, and the evidence extracted was all moderate (Level 3) or above, providing medical and nursing staff with a clear and definite basis of information for clinical application, and indicating methods for the prevention of UEE, which is of great clinical significance.

| Implications for research
Enhancing psychological care for endotracheally intubated patients (Yan et al., 2016)

| Limitations
This overview of systematic reviews had two main limitations. First, we retrieved data from reviews instead of primary studies. The reviews could have had several weaknesses in methodological quality, which would have affected the findings of this overview. Thus, we assessed the methodological quality of the included reviews to identify weaknesses. When interpreting the evidence, the methodological quality was considered. Second, because of the heterogeneity of the literature included in the study, the data were not combined and the effect values could not be analysed.

| CON CLUS IONS
This systematic review reassessed the risk factors and preventive measures for preventing UEE in the intensive care unit and developed an evidence summary with 20 pieces of evidence and recommendations, thus serving as a reference and providing evidence to support the construction of clinical practice guidelines. Moreover, it is recommended that future studies related to specific psychological care for patients with tracheal intubation, continuous quality improvement, different fixation methods for specific conditions of patients and strictly mastering restraint guidelines and methods can be carried out to prevent the occurrence of UEE.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.