Approach to the Airway in the Emergency Department (ED) and Intensive Care Unit (ICU) in a Portuguese University Hospital Center: a prospective observational study

Background Airway management is a commonly performed procedure in the Emergency Department (ED) and Intensive Care Unit (ICU), being tracheal intubation the gold standard with relatively high rates of complications in these settings. The purpose of our study was to analyse the airway approach in our institution at ED and ICU identifying the main complications associated and taking notice of possible factors related to them. Methods Prospective observational study conducted between May and September 2014 in the ICU and ED of Santo Antonio Hospital, with the primary aim of identifying the main complications of the airway approach in the critical patient and secondary aim to take notice of possible factors related with complications of the airway approach related to technical skills of the operator, patient´s specicities and airway approach techniques. The statistical analysis was done in cooperation with the Medical Informatics and Biostatistics Department of the Faculty of Medicine of the University of Porto. P values <0.05 were considered signicant for all hypothesis testing. The analysis was done using the statistical analysis program SPSS® v.21.0 Results A total of 182 patients were included, corresponding to 257 attempted tracheal intubations (ATI). The rate of successful rst-pass orotracheal intubation was 67%. Complications related to ATI were reported in 30% of all attempts, being most frequently among residents (81%, p=0.001) and those with little experience in airway approach (46%, p<0.001). The majority of complications was failed intubation (18%). In the group with complications, relation with an identiable obvious cause was seen in 87% of cases (p<0.001), the major being inexperience in 42% of operators with complications, instead of 1% in the group without complications (p<0.001). Conclusions Failed ATI and its complications are largely dependent on operator ´s expertise. Factors related to patient, lack of appropriate equipment and inappropriate strategy delineation also play a role. Recognition of patients at particular risk of dicult airway management is crucial.


Background
Airway management is a commonly performed procedure in the ED and ICU, being tracheal intubation the gold standard. Several studies identify relatively high rates of complications in these settings, ranging from 20 to 31% in ICUs [1], which can be life-threatening, including severe hypoxia, hemodynamic collapse, cardiac arrest and death [2]. This pool of patients differs from those in elective surgery intubated in the operating room taking into account that they often present with severe respiratory failure, hemodynamic instability, absence of a fasting state and other comorbilities [2]. In general, hypoxemia is common in patients requiring intubation and care should be given to its quick performance to avoid aspiration [3] and further dessaturation, given that previous studies reported a signi cant increase in hypoxemia with the increase in laryngoscopic attempts [3]. In order to minimize the prevalence of severe hypoxemia and severe collapse following intubation and associated complications, speci c risk factors for complications related to intubation and di cult intubation in the ICU have been identi ed [4].
Further circumstances contribute to high incidence of di cult intubation when compared to elective intubation in the operating room [2]. Intubation of criticall ill patients is impaired by lack of training and assistance, unrecognition of patients at risk, failure to plan and to lead an alternative approach, if necessary, and few resources concerning available equipment and the limited space around ICU beds [2].
Differences in outcomes, with higher rate of adverse outcomes and important de ciencies of airway management in ICU and ED, compared with anaesthetic practice, were already emphasized by the 4th National Audit Project of the Royal College of Anaesthetists and Di cult Airway Society (NAP4), leading to a direct response with the recent publication of the guidelines for the management of tracheal intubation in critically ill adults by Di cult Airway Society, Intensive Care Society, Faculty of Intensive Care Medicine and Royal College of Anaesthetists [5]. The particularities and unique challenges of intubation of the critical ill patients were, therefore, recognized. The nature of the urgent procedure, as well as due to patient factors, hinder a complete or proper airway assessment [5], increasing related complications. Moreover, delayed tracheal intubation or multiple attempts at laryngoscopy also potentiate these complications [5].
Unsucessful rst attempt intubation occurs in up to 30% of ICU intubations, again a higher percentage than that seen in the operating room, and around 6% of ICU patients have a predicted di cult airway [5]. Critical illness and the measures taken for its healing can increase airway edema and distortion, increasing the challenge of intubating those patients [5].
It is important to be the most effective possible at rst attempt concerning intubation to avoid complications [6] and improve patient outcome who is, by itself, already debilitated. This goal must be achieved, in addition to other measures, using a detailed plan to airway management in the setting of ED and ICU. There are new videolaryngoscopes devices in order to improve airway management, however, its role, particularly in ICU, is still discussed due to lack of scienti c evidence in this context [5,7], existing only observational studies and small randomized studies supporting the use of video laryngoscopy for orotracheal intubation in ICU [8].
This prospective observational study aims to analyse how we are dealing with the airway in the ED and ICU in our institution as a way to audit the practice of the airway approach in the referred settings.

Methods
A prospective observational study was conducted between May and September 2014 in the ICU and ED of Santo Antonio Hospital. The study was approved by the Ethics Committee of our institution. It was requested and granted by the Ethics Committee exemption from patients´s informed consent because there was no interference in their approach (observational study), no sensitive data was used and an anonymous database has been built ensuring the con dentiality of the data.
An anonymous registry of the airway approach after its performance in critical ill patients was done by the emergencist/intensivist doctor. The paper was identi ed with a code number separately associated with patient identi cation, with the purpose of retrospectively completing eventual missing informations.
The main purposes of the study were the follow: -primary aim: identify the main complications of the airway approach in the critical patient; -secondary aim: take notice of possible factors related with complications of the airway approach related to: 1) technical skills of the operator 2) patient´s speci cities 3) airway approach techniques All patients submitted to tracheal intubation attempts were included. After building an electronic anonymised database, the surveys were destroyed to preserve patients identity.
In the registry paper, the following information was recorded: In the descriptive analysis of the sample, appropriate summary statistics were applied. Categorical variables were described through absolute (n) and relative (%) frequencies. Continuous variables were described as mean plus standard deviation or median, 5th percentile and 95th percentil, according to their symmetrical or assymmetrical distribution, respectively.
Chi-square test of independence was used to analyze categorical variables associations. When more than 20% of contingency table cells regarding the association analysis of two categories were inferior to 5, Fisher exact test (when the two variables had only two categories) or Chi-square exact test (when at least one of the variables had more than two categories) were used. P values <0.05 were considered signi cant for all hypothesis testing. The analysis was done using the statistical analysis program SPSS® v.21.0.

Results
During the period of the study 182 patients were included corresponding to a total of 257 attempted tracheal intubations (ATI). The leading cause for tracheal intubation (TI) was primary respiratory failure (26%) and the rate of successful rst-pass orotracheal intubation was 67% (table 1). The majority of the operators had basic training in Internal Medicine (55%), followed by Anesthesiology (39%), being mostly residents (64%). Of all, only 39% reported to have good experience in approaching the airway, contrasting to 5% with none and 27% with little experience. Fifteen percent of the operators with no experience in airway approach and 50% of those with little experience didn´t manage to sucessfully intubate at rst attempt, while this number decreased substantially with doctors having good (20%) and excellent (3%) airway expertise (p<0.001). Eighteen percent of anesthetists who ATI didn´t succeed to intubate at rst attempt, whilst that number was 82% for other specialities (p=0.013). When rst ATI failed, residents were the operators in 88% of cases, but when this attempt was done by senior doctors, they just missed in 12% (p=0.033) (table 2). The intubation success rate after a rst miss was 83%. Sixty patients (33%) needed 2 or more attempts (maximum 4). No statistical difference was found in intubation success rate at rst ATI during day or night (p=0.503). Reported causes to explain the failure of TI were seen in 28% of cases, being operator´s inexperience the most frequent (13%). Other causes included patient related factors (11%), lack of suitable equipment (2%) and lack of appropriate strategy delineation (2%). In 1% there was an inadequate airway evaluation.
There were 48% of patients with predictive signs of di cult intubation, being obesity the most frequent (25%), followed by neck deformity (21%). 70% of patients with failed rst ATI had predictive signs of di cult airway (p<0.001).
In what concerns the drugs used to airway approach, the majority used opioid (68%) and hypnotic (76%). However, muscle relaxant was used only in 25% of patients in all attempts of TI.
With regard to material, McCoy blade was used in 12%, videolaryngoscopy in 2% and no use of broscopy was registered neither laryngeal mask. In 3% of all attempts, the use of bougie® was necessary. No surgical airway rescue technique was needed.
Complications related to ATI were reported in 30% of all attempts, being most frequently among residents (81%, p=0.001) and those with little experience in airway approach (46%, p<0.001) (table 3). The majority of complications were failed intubation (18%), followed by esophageal intubation (9%), aspiration of gastric contents (2%) and cardiac arrest (1 case). Complications were more often in patients submitted to 2 ATI (66%), followed by 4 ATI (20%), being only 4% in those who had merely 1 ATI (4%, p<0.001). Besides, they were seen mostly in patients with predictive signs of di cult intubation (74%, p<0.001). No relation was found between complications and the diagnostic of the intensive care patient or reason to TI, neither with drugs used in the airway approach.
In the group with complications, relation with an identi able obvious cause was seen in 87% of cases (p<0.001). Inexperience was seen in 42% of operators with complications, instead of 1% in the group without complications (p<0.001). Lack of appropriate equipment (5%, p=0,007), lack of appropriate strategy delineation (8%, p=0.001) and factors related to patient (32%, p<0.001) were present in the group with complications with higher percentages than those seen in the group without complications (table 4). With regard to esophageal intubation as a complication, 54% of operators in this group had little experience in approaching the airway, as opposed to 24% in the group without this complication, where the majority (41%) had good experience (p=0.039). Esophageal intubation was also most frequently as the number of ATI increased (p<0.001, table 5).

Discussion
Failure of tracheal intubation at rst attempt occurs in up to 30% of ICU intubations [5]. In our institution, that number was 33%, although it included intubations in ED. Trainee doctors, mostly internal medicine residents without major experience in airway approach, are primarily responsible for intubation of critically ill patients in our hospital, with implications on successful rate of rst ATI. Regarding the reported reasons to explain failure of TI, this particular aspect was highlighted in our study, with operatoŕ s inexperience accounting for 13%. Patient related factors also played a role, particularly anatomic changes/deformations, as well as obesity. Robust evidence exists that this feature is an important risk factor for intercurrences in airway approach of critically ill patients. For example, in NAP4, a patient with a BMI >30 kg/m2 was twice as likely as a slim patient to have a complication of airway management [5].
Obesity di cults facemask ventilation, supraglottic airway placement, tracheal intubation and FONA (front of neck airway), also conditioning quick and severe desaturation, mainly in the presence of airway obstruction [5]. Severe hypoxemia is a concern when intubating critically ill patients and several preoxygenation techniques like noninvasive ventilation or high-ow nasal cannula oxygen therapy have been developed with the aim of preventing and limiting its incidence and complications following the intubation procedure [9,10,11].
The presence of predictive signs of di cult intubation is related to an increase in failure of TI. Lack of suitable equipment is another issue as towards the critically ill patient there isn´t an optimized and controlled environment as seen in the operating room. The urgency of the procedure in that setting may preclude a standard airway assessment [5] and an appropriate strategy delineation, as was present in our study.
Regardless of the recommendation of using neuromuscular blocking agents, as it reduces intubation complications in the critically ill [5], as well as it improves overall intubating conditions [12], in our study they were used only in 25% of cases, however, no relation was found between drugs used in the airway approach and complications rate. It is crucial for critical care physicians to have skills in advanced airway management as the adoption of algorithmic approaches and rapid-sequence intubation by anesthesiologists and emergency medicine physicians improved the successful rates for the emergency intubations in the critically ill and reduced complications [13].
At the time of our study there were not yet in our institution a wide use of videolaryngoscopy dedicated to ED and ICU and its use was limited to 2% of cases. Videolaryngoscopy in critically ill patient is still requiring further high-quality research. A systematic review and meta-analysis provided evidence that videolaryngoscopy for orotracheal intubation in ICU helps to reduce di cult orotracheal intubation, esophageal intubation, Cormack 3/4 grade and increases rst-attempt success, although without reducing other complications such as severe hypoxemia, severe cardiovascular collapse or airway injury [14]. Although in elective surgery procedures, the use of glidescope® for endotracheal tube placement also resulted in reduction in the incidence and severity of postoperative sore throat compared to Macintosh laryngoscope [15]. In a randomized clinical trial among ICU patients requiring intubation, videolaryngoscopy compared with direct laryngoscopy did not improve rst-pass orotracheal intubation rates and was associated with higher rates of severe life-threatening complications [3]. In another randomized controlled trial comparing direct versus videolaryngoscopy using the C-MAC® for TI in the ED, there was no difference between videolaryngoscopy or direct laryngoscopy using the C-MAC® device in rst-pass success and duration of intubation attempt [16]. However, the 2018 guidelines for the management of tracheal intubation in critically ill adults recommend early use of a videolaryngoscope, with a screen visible to all, and second generation supraglottic airways for airway rescue [5]. The role of laryngeal mask airway in critical care medicine has been addressed long time ago, particularly its use in the di cult airway and its role as an aid to intubation [17].
No use of broscopy was registered in our study seen that awake intubation is often inappropriate in the critically ill [5]. No surgical airway rescue technique was necessary, although 3% of patients had 4 ATI.
Complications related to ATI were reported in 30% of cases, the most frequently being failed intubation (18%). Also esophageal intubation, aspiration of gastric contents and cardiac arrest occurred. A multicenter cohort study reported an ICU intubation-related cardiac arrest of 1 in 40 procedures with high immediate and 28-day mortality, identifying ve independent risk factors for cardiac arrest (overweight or obesity, age more than 75 years old, low SBP [systolic blood pressure] prior to intubation, hypoxemia prior to intubation and absence of preoxygenation before intubation procedure) [4]. In our study, only one case of cardiac arrest was reported, corresponding to less than 1%. Complications increased with the increase in TI attempts, as well as in patients with predicitive signs of di cult intubation. This is a well known relationship. When they occur it is extremely important to identify the cause(s) of those complications and what should have been done differently. In our study, that cause was reported in 87% of cases. Again, inexperience accounted mostly as it attained 42% of operators who had complications, instead of 1% in those without, having a statistical signi cance. Lack of appropriate equipment, inappropriate strategy delineation and factors related to patient were also higher in the group with complications.
Recently published guidelines for the management of tracheal intubation in critically ill adults highlighted the importance of human factors rather than only valuing technical skills or introducing new devices. Communication between team elements, sharing a mental model and planning is of crucial importance.
Task xation and delays in tracheal intubation should be avoided limiting the number of attempts, as well as encouraging the transitioning to the next algorithm step as early as possible once failure has occurred [5]. In fact, it was already recognized in a prospective, multiple-center study that the implementation of an intubation management protocol can reduce immediate severe life-threatening complications related to intubation of ICU patients [18].
As far as we know there are very few studies analysing so many variables and relationships concerning the operator´s expertise performing intubation in critically ill patients. One of them, a prospective cohort study, reported an overall risk of complications of 39% [19], higher than in our study. Nevertheless, our study has several limitations. First, it is a single center prospective observational study, with a small sample. Second, in our institution, rst airway approach in the ED and ICU is largely done by residents, which may not represent the reality in other centers and, thus, affecting the correlations found. Third, at the time of the study, the 2018 guidelines for the management of tracheal intubation in critically ill adults haven´t been published, so our study does not re ect the impact of applying these guidelines.

Conclusion
As seen in our study, failed ATI and its complications are largely dependent on operator´s expertise which should be improved once TI in our ED and ICU is performed mostly at rst by trainee doctors. Factors related to patient, lack of appropriate equipment and inappropriate strategy delineation also play a role.
Recognition of patients at particular risk of di cult airway management is crucial and is recommended even in the most urgent situations [5], being a central element in planning the approach of the di cult airway in these settings. Currently videolaryngoscopy is widely available in our institution in ED and ICU, being its use of great value. Declarations · Ethics approval and consent to participate: The study was approved by the Ethics Committee of Porto University Hospital Centre. It was requested and granted by the Ethics Committee exemption from patients´s informed consent because there was no interference in their approach (observational study), no sensitive data was used and an anonymous database has been built ensuring the con dentiality of the data.

List Of Abbreviations
· Consent for publication: Not applicable.
· Availability of data and material: Authors can con rm that all relevant data are included in the article and/or its supplementary information les. Additional datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
· Competing interests: The authors declare that they have no competing interests.
· Authors' contributions: HF and CMT contributed to study concept and design. CMT, TC, HF, RO, AM and IA contributed to study execution and in acquisition of the data. TC and CMT contributed to statistical data analysis, interpretation, article preparation and drafting. CMT and IA contributed in article critique and review. All authors approved the article submitted.
· Acknowledgements: We want to give a special thanks to Dr. Orquídea Ribeiro from Faculty of Medicine of University of Porto for its contribution in investigation methodology and biostatistics in this work.