Introduction

The first practice guideline on awake tracheal intubation (ATI) was published by the Difficult Airway Society (DAS) in 2020 [1]. Despite the high success rate and a favorable safety profile associated of ATI; this essential technique may be underused for anticipated difficult airway management [1]. Other national data on the use of ATI is largely lacking [2], with available data reported only for single tertiary institutions in Canada, the USA and the UK [3–6].

ATI is most-commonly conducted using flexible bronchoscopy (ATI:FB) in accordance with current recommendations. Our recent study on the frequency of ATI in Germany revealed a discrepancy between the availability of ATI:FB equipment for German anesthesiologists (> 90 %) and their proficiency in the technique (60 %) [7]. Only half of German anesthesiologists stated that they had performed more than 25 ATI:FBs throughout their career. This reluctance in performing ATI may be attributed to a fear of discomfort to the awake patient and poor patient cooperation, a lack of ATI practice training, and the availability of alternative techniques used after the induction of general anesthesia [6]. ATI may also be perceived as time-consuming and complicated, and hence, anesthesiologists may particularly favor video laryngoscopy after the induction of anesthesia. It is noteworthy to mention at this point that video laryngoscopy performed as an awake tracheal intubation (ATI:VL) is an alternative method to ATI:FB when the presenting anatomy permits, and with sufficient clinician experience. ATI:VL is advantageous for preserving airways in severe physiologic derangements within critical care or emergency medicine, but contraindications include certain anatomical presentations, such as limited mouth opening. However, these concerns are offset by the major potential safety benefits of ATI. During a correctly performed ATI the airway remains continuously open, and the patient breathes by himself, making hypoxia much less likely [8]. Therefore, ATI is also recommended in the German S1 guideline for airway management for the anticipated difficult airway. The recommendation entails the use of a flexible intubation endoscope (ATI:FB) and stresses the importance of training and practice in this technique [9]. In our lately published large scale analysis over a 16-year period, 5.5 % of all airway procedures in our high volume tertiary centre were ATI:FBs [10]. This proportion is comparatively high in light of previously reported data [3, 5, 6] Considering ATI as a routine component in our airway management, we conducted now a retrospective study with the aim of contributing more detailed data on ATI from Germany to the international airway community for the first time.

The main objective of our retrospective cohort study was to evaluate the overall success of ATI at our tertiary center over a 2-year period. The secondary outcome focused on assessing the safety profile of ATI. Safety was defined through an analysis of complications and the time required for the procedure. Complications and prolonged procedure duration present potential risks that could compromise patient safety. Therefore, we examined complications, the time needed for ATI, and factors such as indication, patient characteristics, and experience of the anesthesiologist to identify and understand influencing factors on safety.

Our hypothesis posits that ATI, when integrated into routine airway management, represents a successful and safe method for managing the expected difficult airway.

Materials and methods

The study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations [11] and approved in September 2020 by the Ethics Committee of the Medical Association of the State Rhineland-Palatinate (Approval number: 2020-15342-retrospectiv). The requirement for individual patient consent was deemed not required due to the absence of identifiable data.

This historical cohort study was conducted based on records from the University Medical Centre Mainz internal data acquisition and accounting system (DAQ). The DAQ is based on the data set for external quality control in anesthesia on behalf of the German Society of Anesthesiology and Intensive Care Medicine (Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin, DGAI) and the Association of German Anesthesiologists (Bund Deutscher Anästhesisten, BDA) [12]. Mandatory data input into the DAQ is carried out directly by the responsible anaesthesiologist. Datasets can be subdivided by different medical/operative specialities and airway management techniques. The DAQ was electronically searched for all ATI:FB between January 1, 2019, and December 31, 2020, to retrieve the relating case numbers which allow a manual claim search of associated scanned archived anesthesia records in the electronic hospital information system SAP (SAP Deutschland SE & Co KG, Walldorf, Deutschland). The study only includes data on ATI:FB due to the lack of coding for ATI:VL in the DAQ and its infrequent application in our institution. The data were accessed for research purposes starting from October 2020 after receiving the ethics approval.

Extracted data included demographic patient characteristics: age (years), sex (female/male), BMI (kg/m²), ASA status (I,II,III,IV), medical/operative speciality, priority of procedure (elective, urgent, emergency) and median time required for intubation using ATI:FB in minutes. The median time required for intubation was defined as the time elapsed from the patient entering the operation room until release for surgery.

Data extracted from scanned archived anesthesia records included detailed procedural information of the ATI:FB (overall success, route (nasal or oral), experience of anesthesiologist, sedation technique and indication). Complications documented in free-text form on the anesthesia records were retrospectively categorized during analysis into the following groups: technical difficulties (Tube or endoscope replacement with a smaller size due to obstacles in the glottis or tumor area), agitated patient, obstruction of visibility due to mucous secretion, and unspecified complications. Failed ATI:FB cases were excluded from complications and subjected to separate analyses. Authors had access to participant information that could potentially identify individuals, but complete anonymization was conducted prior to analysis to ensure privacy and confidentiality.

Results

A total of 44,971 anesthesia procedures in 2019 and 2020 were electronically searched in the DAQ for ATI:FB, and 2.114 cases of ATI:FB cases were identified. Of these, 94 cases without an associated anesthesia record (separately archived in the electronic hospital information system SAP) were excluded, just like 109 cases in which the manual review of the anesthesia record indicated that it did not involve an ATI:FB procedure (Fig 1). Therefore, 1,911 ATI:FB performed from January 1, 2019, to December 31, 2020, were evaluated, accounting for 4.3 % of all airway management procedures (oral intubation, nasal intubation, mask ventilation [without additional airway intervention], supraglottic airway device, tracheal cannula, double lumen intubation, jet ventilation). 1861 cases with complete data were included in the regression analyses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Search strategy flow chart: Electric DAQ (internal data acquisition and accounting system) search for ATI:FB (awake tracheal intubation using flexible bronchoscopy) and manual review of the associated anesthesia records.

https://doi.org/10.1371/journal.pone.0299071.g001

Characteristics of study population

As shown in Table 1, the median (IQR, [range]) age of the patients was 62 (50–72 [<1–98]) years and BMI was 25.5 (22.3–29.5 [13.1–68.9]) kg/m². 22.7 % of the patients undergoing ATI:FB were obese (433/1,911), with obesity classes 1, 2, 3 occurring in 247 (12.9 %), 96 (5.0 %) and 90 (4.7 %) patients, respectively. 34.7 % (664) of the ATI:FB were performed in female patients and 65.3 % (1,247) in male patients. Most of the patients undergoing ATI:FB were classified as ASA III (51 %), followed by ASA II (39 %), ASA I (5 %), and ASA IV (5 %).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of study population and procedural characteristics of awake tracheal intubation using flexible bronchoscopy ATI:FB.

Absolute numbers and percentages. Age (years) and body mass index (BMI) are reported as medians and (IQR, [range]).

https://doi.org/10.1371/journal.pone.0299071.t001

Patients in dental, oral and maxillofacial surgery accounted for 888 (46.5 %) ATI:FB and 733 (38.4 %) in otorhinolaryngology. The remaining ATI:FB were performed in other specialties (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Frequency of awake tracheal intubation using flexible bronchoscopy (ATI:FB) by medical/operative specialty, indication, and complication.

https://doi.org/10.1371/journal.pone.0299071.t002

Overall success rate and analysis of failed cases

The overall success rate of ATI:FB was 99.6 %. Table 3 lists the eight failed cases in which airway management could not be achieved by ATI:FB alone. No cases were cancelled or deferred after a failed ATI:FB. In two of the eight cases, open surgical tracheotomy was successfully performed by an ENT (ear, nose and throat) surgeon.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Indication, type of failure, reason for failure and management of eight cases of failed awake tracheal intubation using flexible bronchoscopy (ATI:FB).

https://doi.org/10.1371/journal.pone.0299071.t003

Safety of ATI:FB

Complications of ATI:FB and influencing factors.

Overall, there were 104 (5.4 %) ATI:FB with complications, of which 36 (34.6 %) were not further specified, 25 (24.0 %) were technical difficulties (Tube or endoscope replacement with a smaller size due to obstacles in the glottis or tumor area), 26 (25.0 %) occurred in agitated patients, 17 (16,4 %) were caused by obstruction of visibility due to mucous secretion secretions.

In a multivariable logistic regression model, the type of medical/operative speciality (p = 0.0158 for the overall test) and experience of anesthesiologist were the only factors (p = 0.0279 for the overall test) with a significant impact on the risk of complications. As shown in Fig 2, compared to dental, oral and maxillofacial surgery, complications were more likely to occur in all other types of medical/operative speciality. The strongest effect was observed in the group of “other” type of medical/operative specialties with OR = 5.10 (95 % CI [1.51; 18.31] and in otorhinolaryngology (OR = 4.54, 95 % CI [1.64; 14.06]).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Odds ratios with 95 % confidence intervals when comparing a specific outcome to a reference category and corresponding p-values for the comparison to the reference category.

Overall p-values for each category are displayed in the category headings. Reference categories have OR = 1 and no confidence interval.

https://doi.org/10.1371/journal.pone.0299071.g002

Required time to perform ATI:FB and influencing factors.

The median time required for ATI:FB was 16 min [IQR: 11 to 23]. ATI:FB with the indication of a cervical spine fracture took the longest with a median of 24 min [IQR: 17.5 to 35], whereas median time required for ATI:FB with the indication of an inflammation had a median duration of 15 min [IQR: 11 to 20].

In a multiple linear regression model, significant effects on median time required for intubation using ATI:FB were found for indication (p < 0.0001), level of experience (p < 0.0001), sedation technique (p = 0.0408), priority of procedure (p = 0.0134) and medical/operative specialty (p < 0.0001), while BMI (p = 0.0663) and route (p = 0.0730) did not show significant effects (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Marginal mean estimates of time to perform an awake tracheal intubation using flexible bronchoscopy (ATI:FB) adjusted for all covariates in the model.

Explanatory variables were included in multivariable linear regression model for log(time). Results were back transformed to linear time scale. Overall p-values for each category are displayed in the category headings.

https://doi.org/10.1371/journal.pone.0299071.g003

Discussion

This historical study is the only retrospective analysis of contemporary awake intubation practices in Germany. The results indicate that when incorporated into routine airway management, ATI:FB proves to be both effective and safe. In our tertiary institution, we demonstrated that ATI:FB constituted a proportion of 4.3 % of all airway management procedures. Our center achieved a remarkably high success rate for ATI:FB, with only 0.4 % of cases resulting in failure and no procedure-related deaths observed. As anticipated, the majority of cases involved dental, oral, maxillofacial surgery, and otorhinolaryngological procedures, with the most frequent indication being oropharyngeal neoplasia. These elective procedures predominantly opted for an oral route, with one-third performed by an anesthesiologist with less than one year of experience.

The incidence of ATI:FB in our center is notably higher than in previous studies, despite the widespread availability of modern equipment, such as videolaryngoscopes. Joseph et al. from the USA report 0.8 % of all tracheal intubations [5] and Law et al. from Canada 1.06 % [3]. El-Boghdadly et al. from the UK reported a proportion of 1.71 % ATI from all general anaesthetics administered [6]. In this context, it is pertinent to highlight that our ATI:FB ratio of 4.3 % in contrast to data from Joseph and Law is derived from all airway management procedures rather than being limited to tracheal intubations alone. We consistently opt for ATI in patients with changes in the larynx, pharynx, and trachea, considering it a vital aspect of comprehensive diagnostics. This approach is particularly emphasized during panendoscopies in otorhinolaryngology and dental, oral, and maxillofacial surgery procedures. Furthermore, we apply ATI to patients exhibiting multiple predictors for challenging mask ventilation or laryngoscopy, such as a higher Mallampati score, specific dental considerations, or those with a beard. We have been placing a high value on ATI:FB training for decades. As for any manual activity, a solid education and regular training are indispensable prerequisites. In our centre, our residents learn this technique in the first year. To guarantee a sufficiently high number of cases, we set the indication for ATI:FB very generously, because this allows us to train all residents, ensure the required number of 25 ATI to become a specialist, and also ensure that all assisting staff maintain the technical skills of the procedure. Additional solutions involving simulators or rotations in otorhinolaryngology specialties may contribute to training the intubation skills necessary for ATI. Nevertheless, ‘hands-on’ training under experienced guidance with real patients is essential for mastering aspects such as peri-procedure oxygenation, topical anesthesia, and sedation techniques (4). ATI:FB requires a considerable level of expertise and training to avoid complications and reduce failure rates. Demographic changes and economic pressures in German hospitals contribute to reduced education. Additionally, German anesthesiologists often bear external training costs. Quality management efforts, like certificate workshops, aim to ensure practice quality and guideline-appropriate approaches, emphasizing the need for comprehensive initiatives to address training gaps and resource challenges in ATI:FB [7].

Based on our previous findings, the use of ATI:FB in our institution has remained stable over decades, signifying its integration into routine airway management [10]. In contrast, a recent study, encompassing 692 cases of ATI between 2014 and 2020, revealed a significant decrease of approximately 50 % (4). Originally not focused on reasons for a decrease in ATI, the study provides a potential explanation for the widespread belief that ATI is declining. Increasing comfort and trust among clinicians in using videolaryngoscopy to manage anticipated difficult airways after the induction of general anesthesia might be one factor. We share these insights with our Canadian colleagues. Additionally, we align with the thoughts of our American counterparts, including Aibek E. Mirrakhimov et al. [13] who emphasize the importance of high-fidelity simulation and, in our view, particularly stress the significance of deliberate practice, which we actively incorporate into our approach.

The high success rate of ATI.FB at 99,6 % aligns with previously reported success rates [3, 5, 6]. In a recent Norwegian retrospective study involving 833 ATI:FB (0.92 % of all intubations) 3.5  % resulted in failure [14], attributed to low procedural volume per anesthesiologist and exclusive performance of ATI:FB in difficult airway scenarios. Approximately two-thirds of the failed ATI:FB were salvaged with direct and videolaryngoscopy after induction of anesthesia, while one-third required surgical access. These findings are consistent with our data and underscore the significance of these techniques in managing the difficult airway.

The complication rate in our study was 5.4 % of all ATI:FB. Due to varying definitions of complications in other studies, which, in part, included minor complications such as multiple intubation attempts or cough during intubation–incidents not routinely reported by us–we opted not to engage in a comparison of the percentage distributions. The nature of documented complications in our study primarily revolves around technical difficulties (24 %), mirroring the observations in Joseph’s study, where ATI failures were most frequently attributed to the inability to pass the endotracheal tube over the fiberscope (5). This was followed by the challenge of managing agitated patients (25 %), a complication related to cough or gag reflex during intubation also mentioned by Law (3).

This encompasses the two aspects of sedation technique and topical anesthesia within the overall concept of ATI. Notably in our study, the complication arising from the challenge of visibility obstruction due to mucous secretions is observed to be 16.4 % and therefore higher than in other studies. This increased mucosal secretion may be attributed to the fact that the application of anti-salivation, which could potentially prevent this complication, is not routinely practiced in our center. This breakdown of complication analysis provides valuable insights into the challenges encountered during ATI, highlighting areas for potential improvement and targeted interventions to enhance the safety of ATI.

Multivariable logistic regression identified the type of medical/operative specialty and anesthesiologist experience as significant factors influencing complication risk. However, it should be noted that the correlation between increasing experience and decreasing risk was not included in the logistic regression used to assess the joint impact of the influencing factors on complications. Further, when considering the odds ratios in Fig 2, it’s important to note that the individual influencing factors are not independent. The elevated risk of complications (OR = 5.10) in the ‘others’ category may be attributed, for instance, to the low incidence in these unfamiliar settings such as medical diagnostics indicating a less routine environment (Table 2). In contrast, the increased rate of complications (OR = 4.54) in otolaryngology may be linked to the fact that our least experienced colleagues are often deployed in this specialty. Additionally, the presence of large tumors that obstruct airways and may bleed, characteristic of this field, poses a particular challenge during procedures. This emphasizes the importance of tailoring training programs and support measures to address specific challenges associated with different specialties.

The median time required for intubation using ATI:FB was 16 min in our study. Joseph et al. reported a median time of 24 minutes for successful ATI:FB in awake patients. The time is defined as the duration from the patient entering the operating room to the time of intubation. In our study, the endpoint for measuring time is defined as the release into the operating room after completing intubation and all other necessary procedures. It is assumed, however, that these additional procedures, such as a second IV or arterial line, are only necessary in complex, individual cases. We assume that, due to the high case volume, these isolated instances do not significantly impact the median time required for ATI:FB. The median time for ATI:VL (n = 22) in Joseph’s study was also 24 minutes, consistent with data from Denmark, which similarly showed no time difference between ATI:FB and ATI:VL [15] However, having an estimation of the additional time required for ATI:FB is helpful for managing expectations. In elective surgeries, decisions about airway management, particularly for difficult airways, should not be influenced by perceived time pressure or airway manager discomfort with performing ATI. According to updated recommendations from the Canadian Airway Focus Group, seeking assistance from a more experienced colleague in performing ATI is recommended [16].

The limitations include the retrospective nature of the data analysis, which is associated with a risk of bias and a potentially diminished transferability of the data into practice. Underreporting of complications caused by subjective assessment may have affected the reported incidence rates. Moreover, there was no direct control group with alternative intubation techniques to conclude on the safety and complication rate in comparison to other anesthesia procedures. As ATI:FB is often replaced by video laryngoscopy used after the induction of general anesthesia, a comparison between the frequency of ATI:FB and video laryngoscopy after induction of anesthesia in our centre would be of value, yet the DAQ does not map data on video laryngoscopy after induction of anesthesia nor on ATI:VL. Our results show that in 5 out of 8 failures related to ATI:FB, airway management was eventually accomplished using video laryngoscopy after induction of anesthesia. Moreover, the fact that our data reflects the situation in a single institution must be acknowledged as a limitation. In terms of the linear regression model, the explanatory covariates are not independent from each other and therefore the estimated marginal means sometimes differ from the naïve descriptive measures. In other words, the figures estimated as part of the linear regression may not be identical with the calculated descriptive figures because the latter do not take into account other variables that could affect the figure.

Conclusion

ATI:FB is a successful and safe technique, characterized by a low complication rate and a manageable time to perform. A modifiable factor contributing to enhanced safety is the experience of the anesthesiologist. Consequently, ATI:FB should be an integral component of routine airway management.

Furthermore, early incorporation of ATI:FB into anesthesiology training is crucial for the safe handling of difficult airways. Anesthesiologists should be actively encouraged to regularly perform ATI to acquire and sustain valuable experience in this technique.

Supporting information

Acknowledgments

Data shown in this manuscript are part of the professorial dissertation (Habilitation) of N Pirlich and part of the doctoral thesis (Doktorarbeit) of A Berk presented to the Faculty of Medicine of the Johannes Gutenberg-University Mainz. We thank H-J Baldering for his support with data acquisition. Finally, we would like to thank all anesthesiologists in the Department of Anesthesiology, University Medical Centre Mainz, performing ATI:FB that frequently for their support and careful completion of our internal data acquisition system.