Article Text
PDF
Abstract
Objective To identify associations between procedural characteristics and success of neonatal tracheal intubation (NTI) using video laryngoscopy (VL).
Design Prospective single-centre observational study.
Setting Quaternary neonatal intensive care unit.
Patients Infants requiring NTI at the Children’s Hospital of Philadelphia.
Interventions VL NTI recordings were evaluated to assess 11 observable procedural characteristics hypothesised to be associated with VL NTI success. These characteristics included measures of procedural time and performance, glottic exposure and position, and laryngoscope blade tip location.
Main outcome measure VL NTI attempt success.
Results A total of 109 patients underwent 109 intubation encounters with 164 intubation attempts. The first attempt success rate was 65%, and the overall encounter success rate was 100%. Successful VL NTI attempts were associated with shorter procedural duration (36 s vs 60 s, p<0.001) and improved Cormack-Lehane grade (63% grade I vs 49% grade II, p<0.001) compared with unsuccessful NTIs. Other factors more common in successful NTI attempts than unsuccessful attempts were laryngoscope blade placement to lift the epiglottis (45% vs 29%, p=0.002), fewer tracheal tube manoeuvres (3 vs 8, p<0.001) and a left-sided or non-visualised tongue location (76% vs 56%, p=0.009).
Conclusion We identified procedural characteristics visible on the VL screen that are associated with NTI procedural success. Study results may improve how VL is used to teach and perform neonatal intubation.
- intensive care units
- intensive care units, neonatal
- neonatology
- paediatrics
- resuscitation
Data availability statement
Data are available on reasonable request. De-identified data (individual participant data, study protocol, statistical analysis plan) that underlie the results reported in this article are available immediately after publication and ending 36 months after article publication to researchers who provide a methodologically sound proposal in order to achieve aims in an approved proposal. Proposals should be directed to corderw@chop.edu. After 36 months, the data will be available in our university’s data warehouse but without investigator support other than deposited metadata.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Video laryngoscopy (VL) improves neonatal tracheal intubation (NTI) success rates and reduces adverse events.
Optimal procedural techniques using VL for NTI are not well defined.
WHAT THIS STUDY ADDS
Modifiable and observable characteristics like laryngoscope blade placement and tongue position are associated with successful NTI.
Successful intubations are associated with shorter procedural durations and improved glottic exposure.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Identifying VL screen characteristics allows for targeted training, procedural modification to enhance NTI performance, and inform guidelines and protocols for neonatal care.
Introduction
Neonatal tracheal intubation (NTI) is a life-saving procedure for critically ill patients.1 However, NTI first attempt success rates are suboptimal, ranging from 24% to 64%.2 3 Additionally, adverse events are common during NTI, occurring in 22%–39% of NTI procedures.4 5 Importantly, the number of NTI attempts is associated with increased odds of adverse events.6 7 Interventions that improve NTI first attempt success may also improve neonatal patient safety.
In randomised controlled trials, the use of video laryngoscopy (VL) resulted in improved success rates when compared with direct laryngoscopy (DL) for a range of neonatal providers.8–12 In addition, VL has been found to be independently associated with a reduction in adverse events during NTI.2 12–14 Despite the evidence for VL in improving procedural success and outcomes, a key knowledge gap relates to identifying ideal procedural technique when using VL for NTI. Little is known about how to effectively use what is visible on the VL screen to improve procedural performance and increase NTI success. We aimed to identify NTI procedural characteristics visible on the VL screen that are associated with NTI procedural success.
Methods
Settings, population and design
This was a single-centre prospective observational study in a level IV academic neonatal and infant intensive care unit (N/IICU) in the USA. As standard practice in our unit, VL is performed using a Miller blade C-MAC video laryngoscope (Karl Storz, Tuttlingen, Germany), and recordings from the video laryngoscope during NTI procedures were captured, stored and reviewed for quality improvement purposes. To ensure privacy, the video recordings did not include audio. Additionally, all recordings were trimmed and de-identified to remove images of patient or provider faces, retaining solely the video data relevant to each individual NTI attempt.
For each NTI encounter in our unit, we collect a limited dataset of patient, provider, practice characteristics and procedural outcomes; these are submitted to the National Emergency Airway Registry for Neonates (NEAR4NEOS), an international multicentre quality improvement registry. For this study, VL recordings were linked to their corresponding NEAR4NEOS data using date, location and time stamps. We included all infants with NTI procedures performed in the N/IICU using the C-MAC video laryngoscope with complete VL NTI attempt recordings. Exclusion criteria were intubations performed by anaesthesiology or surgical providers, patients with known congenital anomalies affecting the airway and inability to confirm accurate link to the NEAR4NEOS registry.
This study was reviewed by the institutional review board of the Children’s Hospital of Philadelphia and deemed exempt from institutional review board oversight.
Study exposures and outcomes
Measurable and identifiable procedural characteristics from VL recordings were defined by a team of multidisciplinary airway experts and assessed in each NTI attempt. An attempt was defined as a single advanced airway manoeuvre, beginning with the insertion of the laryngoscope blade and concluding on its removal. An encounter refers to the complete sequence of events leading to placement of an advanced airway and may include multiple NTI attempts.
A single trained observer reviewed all NTI video recordings using the following standard definitions for procedural characteristics:
Time to tracheal intubation: measured from the time the laryngoscope blade crossed the lip line on insertion to the moment when the laryngoscope blade was removed from the mouth after completion of the intubation attempt.
Time to obtain view: measured from the time of laryngoscope blade insertion to the moment when the endotracheal tube (ETT) was first visualised on the VL screen in the posterior oropharynx (indicating the proceduralist was satisfied with the view and now moving to place the ETT). In the rare event that insertion of an ETT was not attempted, the total duration of tracheal intubation was considered time to obtain a view.
Time inserting tube: measured from the moment the ETT was first visualised in the oral cavity to the moment when the blade was removed from the mouth after either success or failure with NTI attempt.
View to tube time ratio: a ratio of time to obtain view to time inserting tube was calculated for each attempt to quantify the proportion of time spent in each phase of NTI.
Tracheal tube manoeuvre: defined as each time the ETT was passed beyond the arytenoid cartilage either posteriorly oriented towards the oesophagus or anteriorly oriented towards the tracheal orifice. Many manoeuvres could occur within a given attempt.
Airway location: the VL screen was divided into three equal segments (top, middle and bottom) using an applied overlay during video editing. Identification of glottic location was performed when the ETT was first visualised on the VL screen.
Laryngoscope blade tip location: noted to be lifting the epiglottis out of view of the camera or inserted into the epiglottic vallecula, where a visibility of the epiglottis remains on camera (figure 1). If the laryngoscope blade tip was adjusted during an attempt from lifting the epiglottis to the vallecula or vice versa, the location was noted during the final intubation manoeuvre.
Gastric tube: the presence of an orogastric or nasogastric tube was noted.
Cormack-Lehane (CL) grade: noted at the time of tracheal tube insertion and compared with a standard description from the original and revised versions of the CL scoring system.15 16
Suctioning: attempts at suctioning were identified from observed passage of a suction catheter and quantified for each NTI attempt. Attempts at suctioning were deemed unnecessary if the glottic inlet was visible and not obstructed by secretions prior to suctioning. Attempts at suctioning were deemed unsuccessful if passage of the suction catheter failed to clear targeted secretions.
Tongue location: tongue position was classified as left, right or not visible in which case the tongue was assumed to be entirely lifted by the laryngoscope blade (figure 1). If multiple tongue positions were seen during an individual attempt, the location was noted during the final intubation manoeuvre.
Example tracheal blade tip lifting the epiglottis (A) in the vallecula (B), left-sided tongue (C), right-sided tongue (D).
The primary outcome was successful VL NTI attempt, defined as the successful placement of a tracheal tube.
Statistical analysis
All analyses were conducted using SPSS Statistics V.27 (IBM, Armonk, New York, USA). Patient, provider and procedural characteristics were summarised using simple frequencies and descriptive statistics. While patient characteristics were assessed at the encounter level, provider and practice characteristics were evaluated at the attempt level, as these could change between attempts within the same encounter. The association between procedural characteristics and success was assessed using Pearson’s χ2 test for categorical variables and non-parametric Mann-Whitney U test for continuous variables. A p value of <0.05 was considered statistically significant for all tests.
Results
A total of 109 VL NTI encounters including 164 individual attempts were recorded between October 2021 and January 2023. Of the 109 encounters, 71 (65%) were successful on the first attempt, and all encounters were ultimately successful. Table 1 summarises patient characteristics. The median gestational age at birth was 32 weeks (IQR 26–36 weeks) with a median age at time of intubation of 34 days (IQR 4–74 days). The median weight at NTI was 2780 g (IQR 1900–3920 g). Table 2 summarises attempt level provider and practice characteristics. Non-physician advanced practice providers (nurse practitioners and physician assistants) performed the majority of NTIs (57%), followed by neonatology fellows (postgraduate year 4–6 physicians in training) (33%) and attending neonatologists (9%).
Patient characteristics (n=109)
Attempt-level provider and practice characteristics (n=164)
Successful attempts (n=109) had significantly shorter median total duration (36 vs 60 s, p<0.001), shorter median time to obtain view (20 vs 26 s, p=0.009) and shorter median time inserting the ETT (12 vs 30 s, p<0.001) compared with failed attempts (n=55) (table 3). Compared with unsuccessful intubation attempts, successful intubations more frequently had a CL grade 1 (63% vs 18%, p<0.001) and a left or not visualised tongue location (76% vs 56%, p=0.009). There were no significant differences between the groups in terms of suctioning or presence of a gastric tube.
Procedural characteristics of successful attempts versus failure (n=164)
Of the 164 intubation attempts, suctioning during VL was performed 19 times (12%). Of these 19 suctioning events, 12 (63%) were characterised as unnecessary because an adequate view of the airway was obtained prior to attempt at suctioning and only 10 (53%) were successful in clearing the targeted secretions present at the time of suctioning. The duration of suctioning ranged from 6 to 19 s (mean 10 s).
Discussion
VL improves NTI outcomes, but the optimal way to use information available on the VL screen during NTI procedures is undefined. This is the first study to evaluate procedural characteristics identifiable on the VL screen as they pertain to NTI procedural success. We found that an improved CL grade, left or non-visualised tongue position and laryngoscope blade location lifting the epiglottis were associated with procedural success.
O’Shea et al found that neonatology trainees are unsuccessful for a variety of reasons including oesophageal intubation, failure to recognise airway anatomy and difficulty directing the tracheal tube despite having obtained an adequate view of the airway.17 That study only assessed unsuccessful attempts. Our study adds to this literature by identifying specific procedural characteristics seen on the VL screen. By including both successful and unsuccessful procedures, we were able to determine the association between observable VL NTI characteristics and procedural success.
Establishing adequate glottic exposure is essential for successful tracheal intubation. Consistent with studies of adult and paediatric patients,14 18 19 improved CL grade was associated with successful NTI in our study. However, CL grade does not capture all details of an adequate airway view. We attempted to identify specific and observable characteristics on the VL screen that contribute to improved glottic exposure.
We found an association between laryngoscope blade tip location and NTI success. Infants often have a long U-shaped epiglottis that is narrow, floppy and angled posteriorly.20 All NTIs in our study were performed with a Miller laryngoscope blade. The Miller blade is narrow, predominately flat and has a slight curve at the distal end that provides improved glottic exposure when compared with the Macintosh laryngoscope blade.20–22 In practice, the Miller laryngoscope blade tip can be used to lift the epiglottis or it can be placed in the epiglottic vallecula, where the hypoepiglottic ligament is engaged to expose the glottis. While previous research has found no significant difference in glottic exposure when the Miller blade was positioned to lift the epiglottis or placed in the vallecula, our study found an association between success and a blade tip positioned to lift the epiglottis. These results are hypothesis generating. It is possible the Miller blade should be consistently used to lift the epiglottis for neonatal intubation. Alternatively, it is possible that providers in our study who placed the blade in the vallecula did not engage the hyoepiglottic ligament adequately.
Classic tracheal intubation teaching recommends inserting the laryngoscope blade into the right side of the patient’s mouth so advancing the blade towards the posterior oropharynx displaces the tongue to the left side of the mouth, making room to advance the ETT.23 Our findings support this instruction and establish an association between left-sided or non-visible tongue position and procedural success. Interestingly, we observed a right-sided tongue position in 50 of 169 (30%) intubation attempts. Attention to tongue position and correcting a right-sided tongue is critical in neonates, where the tongue is often proportionally larger and has the potential to be more difficult to control and more obstructive than in adults.24
O’Shea et al found that suctioning rarely improved the airway view.17 We found similar results in our cohort; suctioning was rare and was typically not indicated when performed. On occasion, suctioning was attempted for secretions that did not impede the airway view. Additionally, when suctioning was performed it was often ineffective at clearing targeted secretions. Suctioning adds valuable time to a VL NTI attempt and has the potential to elicit a vagal response and trigger bradycardia and oxygen desaturation in the neonatal patient.25 These results suggest avoiding suctioning when the airway view is established and proper suctioning technique when indicated would likely reduce the duration of NTI attempts and may improve NTI outcomes.
The median attempt duration for successful NTI in our cohort (36 s) exceeds Neonatal Resuscitation Program (NRP) recommendation to complete intubation within 30 s.26 O’Donnell et al similarly found that most NTI are not performed within 30 s.4 Our results suggest that the NRP recommendations for intubation time may not be a realistic target. Because increasing number of intubation attempts is also associated with increasing risk for tracheal intubation adverse events, clinicians must weigh the risk of adverse events from a prolonged NTI attempt with the risk of subsequent intubation attempts in real time.6 7
We divided the total procedural time into distinct phases: time spent obtaining the view and time spent placing the ETT, and we assessed the ratio between these phases. In successful procedures, a greater proportion of the total procedural time was spent obtaining the view (reflected by the ratio of 1.6), In contrast, this ratio was 0.7 in unsuccessful procedures, suggesting those proceduralists allocated less time to obtaining the view and more time trying to place the ETT. It is possible that allocating more time and attention to achieving an optimal view would improve procedural success and reduce the amount of time spent trying to insert an ETT with a suboptimal view.
Limitations
This was a single-centre study at a level IV N/IICU. Although we excluded infants with known airway malformations, our study population may not be representative of all NTIs. It is possible that supervisors provided corrective feedback to some providers during intubation; however, we were unable to determine this because VL recordings did not include audio. Given the procedural location and unit practice to perform DL in extremely low birthweight (ELBW) infants, these findings may not represent intubations in the delivery suite or in ELBW infants. In addition, intubation premedications including sedation and neuromuscular blockade are routinely used in our unit and our findings may not generalise to NTIs performed without premedication.
Strengths
Our study has several strengths. This is the first study to identify observable procedural characteristics in VL NTI. We included both successful and unsuccessful neonatal VL NTI attempts, which allowed us to identify specific characteristics associated with procedural success. Because our study was performed in a training institution where airway providers have varied experience levels, results may be applicable across other training settings.
Conclusion
We identified procedural characteristics visible on the VL screen that are associated with NTI procedural success. Intubators should allocate more procedural time to obtaining an adequate airway view by positioning the laryngoscope blade to optimise glottic exposure, lifting the epiglottis and ensuring the tongue is not on the right side of the image before attempts at passing the ETT. The identification of modifiable factors may improve how VL is used to teach and perform neonatal intubation.
Data availability statement
Data are available on reasonable request. De-identified data (individual participant data, study protocol, statistical analysis plan) that underlie the results reported in this article are available immediately after publication and ending 36 months after article publication to researchers who provide a methodologically sound proposal in order to achieve aims in an approved proposal. Proposals should be directed to corderw@chop.edu. After 36 months, the data will be available in our university’s data warehouse but without investigator support other than deposited metadata.
Ethics statements
Patient consent for publication
References
Footnotes
X @wilcorder
Contributors WC: guarantor, writing—review and editing, writing—original draft, resources, methodology, investigation, formal analysis, data curation, conceptualisation. TN: writing—review and editing, validation, formal analysis. AMA: writing—review and editing, conceptualisation. JF: writing—review and editing, conceptualisation. EL: writing—review and editing, conceptualisation. RDG: writing—review and editing, conceptualisation. LS: writing—review and editing, data curation. EEF: writing—review and editing, supervision, resources, methodology, formal analysis, data curation, conceptualisation.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.