Background

Sedated and awake tracheal intubation approaches are considered safest in adults with difficult airways, but little is known about the outcomes of sedated intubations in children. The primary aim of this study was to compare the first-attempt success rate of tracheal intubation during sedated tracheal intubation versus tracheal intubation under general anesthesia. The hypothesis was that sedated intubation would be associated with a lower first-attempt success rate and more complications than general anesthesia.

Methods

This study used data from an international observational registry, the Pediatric Difficult Intubation Registry, which prospectively collects data about tracheal intubation in children with difficult airways. The use of sedation versus general anesthesia for tracheal intubation were compared. The primary outcome was the first-attempt success of tracheal intubation. Secondary outcomes included the number of intubation attempts and nonsevere and severe complications. Propensity score matching was used with a matching ratio up to 1:15 to reduce bias due to measured confounders.

Results

Between 2017 and 2020, 34 hospitals submitted 1,839 anticipated difficult airway cases that met inclusion criteria for the study. Of these, 75 patients received sedation, and 1,764 patients received general anesthesia. Propensity score matching resulted in 58 patients in the sedation group and 522 patients in the general anesthesia group. The rate of first-attempt success of tracheal intubation was 28 of 58 (48.3%) in the sedation group and 250 of 522 (47.9%) in the general anesthesia group (odds ratio, 1.06; 95% CI, 0.60 to 1.87; P = 0.846). The median number of intubations attempts was 2 (interquartile range, 1 to 3) in the sedation group and 2 (interquartile range, 1, 2) in the general anesthesia group. The general anesthesia group had 6 of 522 (1.1%) intubation failures versus 0 of 58 in the sedation group. However, 16 of 58 (27.6%) sedation cases had to be converted to general anesthesia for successful tracheal intubation. Complications were similar between the groups, and the rate of severe complications was low.

Conclusions

Sedation and general anesthesia had a similar rate of first-attempt success of tracheal intubation in children with difficult airways; however, 27.6% of the sedation cases needed to be converted to general anesthesia to complete tracheal intubation. Complications overall were similar between the groups, and the rate of severe complications was low.

Editor’s Perspective
What We Already Know about This Topic
  • The incidence of difficult intubation is about 1.5% in children

  • In most cases of anticipated difficult intubation, general anesthesia is used rather than sedation

What This Article Tells Us That Is New
  • In a retrospective study using the Pediatric Difficult Airway Registry, intubation under sedation had a similar rate of first-attempt success compared to intubation with general anesthesia

  • Nevertheless, 28% of the sedation cases needed to be converted to general anesthesia to complete tracheal intubation, and 1% in the general anesthesia group had failed intubations

  • Complications overall were similar between the groups, and the rate of severe complications was low

The incidence of difficult mask ventilation is approximately 6.6%, and the incidence of difficult tracheal intubation is close to 1.5% in the general pediatric population.1  Difficult tracheal intubation has been estimated to be three times higher in neonates in a recent European multicenter international study.2  Difficult intubation can be associated with significant complications including hypoxemia, airway trauma, cardiac arrhythmias, and cardiac arrest.2,3  Anesthetic strategies to perform tracheal intubation in children with an anticipated difficult airway may differ depending on patient factors, institutional resources, and clinician experience. General anesthesia with volatile or intravenous agents remains the most common approach to perform tracheal intubation in children with difficult airways. However, some anesthesia clinicians prefer intravenous sedation for tracheal intubation.3  Proponents of sedated intubation tout the maintenance of spontaneous ventilation and the ability to emerge the patient if intubation is impossible as advantages over intubation under general anesthesia. Previous data suggest that controlled ventilation with or without neuromuscular blockade is associated with fewer complications than spontaneous ventilation. A sensitivity analysis in that study suggested that the increased complications were related to airway reactivity during tracheal intubation, suggesting that the anesthetic depth may play a role in complications.4 

Because sedation for airway management is not a common practice, there remains a knowledge gap about its efficacy and related complications in patients with difficult airways. It is unlikely that any single center would have enough sedated cases to perform a comparative analysis. The Pediatric Difficult Airway (PeDI) Registry is an international registry that prospectively collects data from pediatric patients with difficult airways.3,5,6  Our study aimed to use data in the Pediatric Difficult Airway Registry to determine whether sedation for tracheal intubation in children with difficult tracheal intubation is associated with a lower first-attempt success rate and more complications than general anesthesia. We hypothesized that sedated intubation would be associated with lower first-attempt success and more complications than general anesthesia. Our study primary outcome was the first-attempt success rate of tracheal intubation.

The Pediatric Difficult Airway Registry

This observational study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.7  The Pediatric Difficult Intubation Registry is an international, multicenter, web-based registry created in 2012 by a special interest group of the Society for Pediatric Anesthesia. The Pediatric Difficult Airway Registry collaborative aims to improve the quality of airway management in children with difficult intubation.3  The registry collects observational data prospectively from 34 international hospitals. The registry collects patient demographics, physical assessment, airway management techniques including devices, pharmacologic and ventilation strategies, and relevant outcomes from children under 18 yr old cared for by anesthesiologists in different hospital locations such as the operating room, the intensive care unit, and the emergency department. Each center was granted institutional review board approval for standardized data collection with the requirement for written informed consent waived. Each participant hospital enters data via a secure, password-protected web-based data entry portal using a centralized Research Electronic Data Capture. Data compliance is ensured by a pediatric anesthesiology attending physician identified as project coordinator at each center, and compliance and data accuracy are audited monthly at the central level, by the data-coordinating center, The Children’s Hospital of Philadelphia. Pediatric Difficult Airway Registry data from a similar time period has been previously reported in a study analyzing standard versus nonstandard videolaryngoscopy blades. That study only examined patients in whom videolaryngoscopy was used, while this current study examines all patients enrolled in the registry.6 

Selection of Patients and Variables

Patients included in the study were children under 18 yr old in whom tracheal intubation is difficult as defined by at least one of the following criteria:

  1. Children with difficult laryngeal exposure as directly assessed by the attending anesthesiologist with direct laryngoscopy (Cormak and Lehane Classification of 3 or higher).8 

  2. Children in whom direct laryngoscopy was physically impossible because of anatomical reasons (e.g., severely limited mouth opening).

  3. Children who failed direct laryngoscopy within the preceding 6 months.

  4. Children in whom the attending anesthesiologist deferred direct laryngoscopy because of a low chance of success or a perceived increased risk of harm.

Intubation encounters from consecutive patients in the registry that occurred from September 2017 to December 2020 were included in the study. Patients with missing demographic data, type of planned anesthesia technique, or reporting awake or no anesthesia technique planned and patients with unanticipated difficult airway were excluded.

Definitions

Planned anesthesia technique was defined as the technique that the anesthesia provider initially attempted to establish successful tracheal intubation. Final anesthesia technique was the anesthesia modality used to achieve successful tracheal intubation. Anesthesia technique refers to the approach used by the anesthesia provider to obtain appropriate conditions for tracheal intubation, including sedation, general anesthesia, and awake or no anesthesia. The technique was determined by the anesthesia attending physician. Failed intubation was defined as the inability to intubate the trachea despite multiple intubation attempts. All decisions on how to manage the airway were made by the attending anesthesiologist and were not dictated by the study design. The groups were categorized based on the planned anesthesia technique between sedation and general anesthesia.

We defined complications as nonsevere and severe using modified operational definitions from the National Emergency Airway Registry for Children (NEAR4KIDS).9,10  Nonsevere complications were hypoxemia (defined as an oxygen saturation of less than 90% for more than 60 s or a 10% decrease in baseline saturation for more than 60 s), minor airway trauma (lips or dental), esophageal intubation with immediate recognition, laryngospasm, bronchospasm, pharyngeal bleed, epistaxis, arrhythmia, and emesis without aspiration. Severe complications included cardiac arrest, severe airway trauma (glottis, subglottis, palatoglossal arch, intraoral), esophageal intubation with delayed recognition, death, aspiration, and pneumothorax.

Propensity Score Matching

We used the nearest-neighbor matching method with the logit of propensity scores as the distance measure to balance baseline characteristics and avoid selection bias between patients who received general anesthesia and those who were sedated. Propensity scores estimated from the logistic regression model represent individual probabilities for being in the sedation group. Twenty-one baseline variables including patients’ demographics (i.e., age, weight, sex, history of prematurity) and medical characteristics data (i.e., American Society of Anesthesiologists [ASA] status, entry criteria, physical exam findings, diagnosed syndrome, first-attempt provider’s role, anticipated difficulties, intubating location), together with the site variable, were assessed for matching. Correlations between matching variables were assessed using variance inflation factor and Spearman correlation coefficient matrix: weight, entry criteria 4 (deferred direct laryngoscopy because of a low chance of success and a perceived increased risk of harm), and physical exam finding (yes/no) were not included for matching because they were highly correlated with age, other entry criteria, types of physical exam findings, respectively (but were included for pre- and postmatching balance assessment). To assess whether the association between age and use of sedation was linear, we divided age into 10 equal-length intervals and plotted log odds of use of sedation over the intervals. The association was found to be nonlinear, and we therefore grouped the patients into five age groups according to developmental classification (i.e., neonates, less than 1 month old; infants, 1 month to 1 yr old; toddlers, 1 to 5 yr old; school-aged, 5 to 13 yr old; and teenagers, 13 to 18 yr old). All variables were included as categorical variables for matching. Type of ventilation was considered a posttreatment variable related to the anesthetic technique and was not included for matching. We used R package MatchIt11  to implement the matching. Before matching, the ratio of the number of patients managed with sedation to general anesthesia was 75:1,764 (~1:23), so we allowed 15 matching iterations to reach the matching ratios from 1 to 1 to 1 to 15 (i.e., up to 15 general anesthesia patients can be matched to each sedated patient). The matching algorithm allowed sedated patients with highest propensity score (i.e., presumably the hardest to find a match for in the general anesthesia group) to be matched first for each iteration. Each iteration paired the general anesthesia patients with the highest digit match to the sedated patient. Once the general anesthesia patient was matched, it was not reconsidered. We also set the caliper width to be 0.1 of the SD of the logit of the propensity score to only allow patients within this distance to be paired. If no general anesthesia patient was found to be within the caliper width of a sedated patient, that sedated patient was left unmatched. On the other hand, if a general anesthesia patient could not be matched to a sedated patient within the specified caliper width within 15 iterations, that general anesthesia patient was left unmatched. To assess the effectiveness of matching, the balance in baseline characteristics between the treatment groups was assessed using absolute standardized mean difference (Cobalt: Covariate Balance Tables and Plots, R package version 4.3.0; https://CRAN.R-project.org/package=cobalt) before and after matching, with the absolute value greater than 0.1 considered imbalanced.

Study Endpoints

The primary outcome was first-attempt success of tracheal intubation. A tracheal intubation attempt was defined as the act of inserting an airway device into the pharynx or naris with the intent to perform tracheal intubation.

Secondary outcomes included number of intubation attempts, severe and nonsevere complications throughout all attempts, and complications during the first attempt of tracheal intubation. We also assessed devices used during first attempt to intubate the trachea and technical difficulties encountered between the sedation and the general anesthesia groups.

Statistical Analysis

Descriptive analyses were conducted by treatment groups (sedation vs. general anesthesia). The frequencies and percentages were presented for categorical variables, while median and interquartile ranges were used for numeric variables. To account for the clustering within site and within matching pairs, we used marginal models (i.e., generalized estimating equation) with exchangeable working correlation structure to establish the association between treatment (sedation vs. general anesthesia) and outcomes. Interaction of site and matching pair was included as the cluster variable. Binomial distribution and logit link function were specified for binary outcomes (i.e., first-attempt success, failed intubation, and complications), and odds ratios and 95% CIs were computed where the model converged. For rare outcomes, defined as incidence being less than 5% in total, Gaussian distribution and identity link function were also specified, and absolute risk difference and 95% CIs were computed as a post hoc sensitivity analysis.12  Incidence rate ratios with 95% CI (instead of odds ratios) were used to establish relations between treatment and count outcomes (i.e., number of attempts) when a Poisson distribution and a log-link function were specified in the generalized estimating equation model. Generalized estimating equation models incorporating matching weights were conducted for matched data, and baseline characteristics found to be imbalanced after matching were included as covariates in postmatching models (except site since it was included as cluster variable). To assess difference in devices used between treatment groups, use of supplemental oxygenation, and use of nasal devices, a chi-square test or Fisher’s exact test was used, as appropriate. Since these variables were decided after the anesthesia technique was chosen (posttreatment variable), the analysis was conducted for prematched data only. No a priori power calculations were conducted. The analysis was conducted using Statistical Analysis Software version 9.4 (SAS Institute Inc, USA) and R software version 3.5.1 (R Core Team, Vienna, Austria), and propensity score matching was performed with the MatchIt package.11  A two-tailed test was conducted for all hypothesis testing, and a P value less than 0.05 was considered statistically significant.

We collected data from difficult airway encounters in 34 hospitals between January 2017 and December 2020. We assessed 2,649 patients, of which 1,839 patients with anticipated difficult airway met the inclusion criteria for the study (fig. 1). Of those 1,839 patients, 75 received sedation, and 1,764 patients received general anesthesia for tracheal intubation. Compared to the general anesthesia group, patients receiving sedation included a higher percentage of teenagers, ASA status E, anticipated difficult mask ventilation, and difficult intubation, and a higher percentage of them were intubated in the intensive care unit (table 1). The matching resulted in 58 sedated patients matched to at least one general anesthesia patient, with a total number of 522 general anesthesia patients being matched. Each sedated patient was matched to a median number of 10 (interquartile range, 2 to 15) general anesthesia patients. Among 58 sedated patients, 21 (36.2%) were matched to 1 to 5 general anesthesia patients, 13 (22.4%) were matched to 6 to 14 general anesthesia patients, and 24 (41.4%) were matched to 15 general anesthesia patients. Matched sedation patients were more likely to be older (toddlers, school-aged, and teenagers), heavier, ASA classification I or II, less likely to have a diagnosed syndrome or airway-related diagnosis, and more frequently intubated in the operating room than the intensive care unit compared to unmatched sedation patients. Additionally, there were 34 sites in the prematched data set and 16 sites left after matching. Postmatching absolute standardized mean differences were greater than 0.1 for type of provider (i.e., trainee, attending) during the first intubation attempt (absolute standardized mean difference = 0.104), and this variable was added as a covariate in postmatching generalized estimating equation models (fig. 2). The patient characteristics and case data are presented in table 1.

Table 1.

Demographic Data and Standardized Mean Difference of the Cohort by Anesthesia Technique (Sedation vs. General Anesthesia) for Tracheal Intubation

Demographic Data and Standardized Mean Difference of the Cohort by Anesthesia Technique (Sedation vs. General Anesthesia) for Tracheal Intubation
Demographic Data and Standardized Mean Difference of the Cohort by Anesthesia Technique (Sedation vs. General Anesthesia) for Tracheal Intubation
Fig. 1.

Flow diagram of patients in the Pediatric Difficult Intubation (PeDI) Registry included in the study.

Fig. 1.

Flow diagram of patients in the Pediatric Difficult Intubation (PeDI) Registry included in the study.

Close modal
Fig. 2.

Absolute standardized mean difference for patients of variables between patients receiving sedation versus general anesthesia before and after propensity score matching. Nearest neighbor matching with caliper width equal to 0.1 of the standard deviation of the logit of the propensity score was performed. To account for the clustering within site and within matching pairs, we used marginal models with exchangeable working correlation structure to establish the association between treatment (sedation vs. general anesthesia) and outcomes. Interaction of site and matching pair was included as the cluster variable. ASA, American Society of Anesthesiologists; ICU, intensive care unit.

Fig. 2.

Absolute standardized mean difference for patients of variables between patients receiving sedation versus general anesthesia before and after propensity score matching. Nearest neighbor matching with caliper width equal to 0.1 of the standard deviation of the logit of the propensity score was performed. To account for the clustering within site and within matching pairs, we used marginal models with exchangeable working correlation structure to establish the association between treatment (sedation vs. general anesthesia) and outcomes. Interaction of site and matching pair was included as the cluster variable. ASA, American Society of Anesthesiologists; ICU, intensive care unit.

Close modal

The rate of first-attempt success of tracheal intubation after applying propensity score matching was 28 of 58 (48.3%) in patients who had sedation and 250 of 522 (47.9%) in patients with general anesthesia (odds ratio, 1.06; 95% CI, 0.60 to 1.87; P = 0.846). The number of attempts was not significantly different between the groups with a calculated incidence rate ratio of 0.86 (95% CI, 0.58 to 1.28; P = 0.460; table 2). No intubation failures were reported in the sedation group, while 6 of 522 (1.1%) patients in the general anesthesia group had intubation failure. In 2 patients, the planned procedure was performed with a supraglottic airway, while 4 patients were awakened and their planned procedures were canceled. The rate of conversion from sedation to general anesthesia was 16 of 58 (27.6%; 18 of 75 [24.0%] in prematched data); all of these patients were eventually intubated using general anesthesia with no failed intubations reported.

Table 2.

Outcomes before and after Propensity Score Matching: Tracheal Intubation First-attempt Success, Complications, and Number of Attempts

Outcomes before and after Propensity Score Matching: Tracheal Intubation First-attempt Success, Complications, and Number of Attempts
Outcomes before and after Propensity Score Matching: Tracheal Intubation First-attempt Success, Complications, and Number of Attempts

The complications encountered and comparisons between the sedation and general anesthesia groups before and after propensity score matching are presented in table 2. The complications overall were similar between the sedation group and the general anesthesia group: 15 of 58 (25.9%) versus 90 of 521 (17.3%; odds ratio, 1.41; 95% CI, 0.72 to 2.76; P = 0.323). Nonsevere complications altogether in the sedation group and the general anesthesia group were not significantly different: 14 of 58 (24.1%) versus 90 of 521 (17.3%; odds ratio, 1.28; 95% CI, 0.65 to 2.55; P = 0.478). The most common nonsevere complication was hypoxemia, but it did not significantly differ between the groups (fig. 3). Severe complications were not different between the groups (table 2). When analyzing complications only during the first attempt, no significant differences were encountered between the sedation group and the general anesthesia group.

Fig. 3.

Forest plot showing the primary and secondary outcomes after propensity score matching and generalized estimating equation analysis. NA, generalized estimating equation model did not converge or less than five patients had the outcome; the upper limit of the confidence interval was cut by 25. The x axis is on the log scale. *Odds ratios in the forest plot are generated from matched data using marginal model (i.e., generalized estimating equation) with binomial distribution and logit link function.

Fig. 3.

Forest plot showing the primary and secondary outcomes after propensity score matching and generalized estimating equation analysis. NA, generalized estimating equation model did not converge or less than five patients had the outcome; the upper limit of the confidence interval was cut by 25. The x axis is on the log scale. *Odds ratios in the forest plot are generated from matched data using marginal model (i.e., generalized estimating equation) with binomial distribution and logit link function.

Close modal

Post hoc sensitivity analysis for rare events showed that patients in the sedation group had lower risk of epistaxis (0 of 58 [0%] vs. 13 of 521 [2.5%]; risk difference, ˗2.2%; 95% CI, ˗ 3.9% to ˗0.5%; P = 0.012) than the general anesthesia group. When analyzing complications only during the first attempt, patients in the sedation group had a lower risk of minor airway trauma (0 of 58 [0%] vs. 14 of 521 [2.7%]; risk difference, −3.5%; 95% CI, −5.8% to −1.2%; P = 0.002), esophageal intubation with immediate recognition (0 of 58 [0%] vs. 5 of 521 [1.0%]; risk difference, −0.6%; 95% CI, −1.1% to −0.1%; P = 0.026), and epistaxis (0 of 58 [0%] vs. 9 of 521 [1.7%]; risk difference, −1.1%; 95% CI, −1.8% to −0.3%; P = 0.003) than the general anesthesia group.

Table 3 presents descriptive data about airway devices, anesthetic technique, and technical difficulties between the groups using prematched data. Additionally, post hoc comparison of techniques for supplemental oxygenation, nasal endotracheal intubation, and use of nasopharyngeal airway during the first attempt are presented in table 3. A variety of drugs were used for sedation, the most common being midazolam, ketamine, dexmedetomidine, fentanyl, and propofol; common combinations of drugs used included opioids and propofol ± midazolam, dexmedetomidine and ketamine ± midazolam, and fentanyl and ketamine ± midazolam. Sevoflurane was the most common agent used for general anesthesia. Glycopyrrolate was used in 23 (30.7%) patients in the sedation group and 192 (10.9%) patients in the general anesthesia group. Topical lidocaine was used in 28 (37.3%) of the sedated patients and 121 (6.9%) of the general anesthesia patients. In the general anesthesia group, 48.0% (847 of 1,764) of the patients received neuromuscular blockade, and 5 of 18 of the patients in the sedation group that were converted to general anesthesia also received a neuromuscular blocking drug.

Table 3.

Devices Used during First Attempt, Technical Difficulties, and Use of Supplemental Oxygenation Reported (Data before Propensity Score Matching)

Devices Used during First Attempt, Technical Difficulties, and Use of Supplemental Oxygenation Reported (Data before Propensity Score Matching)
Devices Used during First Attempt, Technical Difficulties, and Use of Supplemental Oxygenation Reported (Data before Propensity Score Matching)

In this propensity score–matched cohort of pediatric patients receiving sedation versus general anesthesia to facilitate tracheal intubation, there was no difference in first-attempt tracheal intubation success between the two groups. The complications overall were low and similar between the groups. Post hoc sensitivity analysis showed that patients in the general anesthesia group had an increased risk of minor airway trauma, epistaxis, and esophageal intubation with immediate recognition.

Previous reports from the Pediatric Difficult Airway collaborative have shown a higher proportion of complications as the number of intubation attempts increases, leading to guidance to limit the number of intubation attempts.3  A previous study from the Pediatric Difficult Airway collaborative analyzing ventilation techniques found that spontaneous ventilation was associated with more complications in children with difficult airway, but sensitivity analysis suggested that the relationship was mediated by airway reactivity rather than the ventilation technique itself.4  This finding implied a possible relationship between lighter anesthetic depth and complications. The definitive effect of neuromuscular blockade and ventilation technique on complications is challenging to evaluate due to the overlap between the two techniques and the retrospective nature of our study; certainly, the general anesthesia group includes patients paralyzed during the first attempt, while none of the sedation group patients were paralyzed. The type of ventilation was dependent on the anesthetic technique, and our study specifically compares the outcomes of the two different anesthetic techniques (sedation vs. general anesthesia) using propensity score matching to account for such confounding variables. It is possible that complications in the sedated patients were related to airway reactivity, which is more likely in spontaneously ventilating patients. In this propensity score–matched cohort, there was no difference in the median number of tracheal intubation attempts between the sedation and general anesthesia groups. Although the rate of complications in our study was low, there were some interesting findings. There were no reported intubation failures (0 of 58) in the sedation group versus 6 of 522 in the general anesthesia group. The patients with failed intubation either were rescued and successfully oxygenated and ventilated with a supraglottic airway or emerged from anesthesia and the surgical case was canceled (table 2), but no emergent front-of-neck access was performed in this cohort. In the sedation group, 16 of 58 (27.5%) of patients had to be converted to general anesthesia before they were successfully intubated. Factors that could explain this need to convert to general anesthesia include the inability of some pediatric patients to follow commands during sedation and the quality of airway topicalization with local anesthetics; as mentioned before, previous work from our group identified airway reactivity as a contributor to complications in this population. In our study, the number of sedated patients that received topical local anesthetic was rather low (37.3%), and that may have influenced our results. Interestingly, the general anesthesia group had an increased risk of esophageal intubation with immediate recognition, minor airway trauma, and epistaxis, which seems remarkable since the sedation group had a higher rate of nasotracheal intubations and use of a nasopharyngeal airway as an adjunct. The causes for these associations are likely multifactorial; unfortunately, we can only hypothesize about possible contributing factors. There is limited data regarding sedation for tracheal intubation in children with difficult airways. Péan et al.13  studied the use of sevoflurane versus propofol in adults with difficult airway and found similar success rates during intubation attempts and similar technical difficulties but higher incidence of tachycardia and hypertension in the sevoflurane group. Our study does not support a preferred approach for tracheal intubation in children with difficult airway based on the first-attempt success rate and the rate of complications. Clinicians should select an approach based on their skill and patient factors.

We did not include the use of neuromuscular blockade as a matching variable since only the patients in the general anesthesia and failed sedation groups received it. Previous work from the Pediatric Difficult Airway Registry has reported on the effect of neuromuscular blockade,4  but no significant differences in outcomes have been related to neuromuscular blockade. Sedation is a continuum that can be difficult to clearly establish, even by experienced clinicians.14  Unlike in adults, infants and children may be unable to follow directions during stressful situations such as an awake or sedated tracheal intubation, making the use of sedation challenging. Several sedation scales have been developed to assess the level of the sedation by clinicians such as the Observer’s Assessment of Alertness/Sedation,15  the Pediatric Sedation State Scale,16  the Vancouver Sedative Recovery Scale,17  and the University of Michigan Sedation Scale.18  Nonetheless, some of these scales may present limited ability to differentiate deeper states of sedation, and interprovider variability remains a problem. In this study, there was significant variability in the frequency of the use of sedation to facilitate tracheal intubation by center. It is possible that providers who do not use sedation often may have more complications than those providers with more familiarity with the technique. Additionally, this cohort had a significant number of trainees as the first-attempt clinician, and this variable remained imbalanced after matching, which may have affected our results. We added the type of clinician as a covariate to the final analysis to best control for this issue. Patients in the sedation group were more likely to receive supplemental oxygenation during intubation than those under general anesthesia, yet there was no significant difference in the incidence of hypoxemia between the groups.

Our study has to be considered in the context of its limitations. First, although sedation is one of the options for induction of anesthesia and airway management, the registry does not capture the level of sedation used by the anesthesia provider, which precludes our ability to classify according to the level of sedation. Additionally, sedation was not a commonly used technique in this cohort, leading to only a small number of sedation cases. Because our study is retrospective, we lose some granularity about why clinicians chose general anesthesia versus sedation. Although we matched physical exam findings, it is impossible to know all the factors that may have influenced the decision to choose one technique over another. It is possible that the patients who were sedated had more concerning physical exam findings and were anticipated to be more challenging to intubate. We also lack details about the timing and dosages of the various drugs used for sedation. The degree and duration of hypoxemia is not captured in detail by the registry, nor is the duration of supplemental oxygenation. We did not include this as part of the analysis of the complications that may have underscored the effect of supplemental oxygenation on hypoxemia in this cohort. Additionally, we cannot account for the different degrees of anticipated difficulty that were perceived by clinicians in these patients; it is possible that those patients with more severe syndromic features/anatomical abnormalities were planned for sedation, and we are unable to adjust for this. Despite these limitations, this study provides further insight to a small but challenging population that is difficult to study. Propensity score matching was used to minimize selection bias and the effect of baseline characteristics in the selected outcome but only accounts for measured variables. It is possible that unmeasured variables still may have influenced the studied outcomes.

In conclusion, the rate of first-attempt success of tracheal intubation was similar in children with difficult airways intubated under general anesthesia versus sedation. However, 27.6% of sedation cases needed to be converted to general anesthesia to complete tracheal intubation. The rate of nonsevere and severe complications were low and similar in both groups. Post hoc sensitivity analysis demonstrated that sedation was associated with a lower risk of minor airway trauma, esophageal intubation with immediate recognition, and epistaxis.

Acknowledgments

The authors thank Heather Griffis‚ Ph.D.‚ and Steve Ampah‚ Ph.D.‚ from the Department of Biomedical and Health Informatics, Data Science and Biostatistics Unit at the Children’s Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.

Research Support

Supported by internal funding from the Department of Anesthesiology and Critical Care, Children’s Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.

Competing Interests

Dr. Kovatsis is a medical advisor to Verathon, Inc. (Bothell, Washington). Dr. Fiadjoe is a Board member for the American Board of Anesthesiology (Raleigh, North Carolina) and received past grant funding from the Anesthesia Patient Safety Foundation (Rochester, Minnesota). The other authors declare no competing interests.

1.
Valois-Gómez
T
,
Oofuvong
M
,
Auer
G
,
Coffin
D
,
Loetwiriyakul
W
,
Correa
JA
:
Incidence of difficult bag-mask ventilation in children: A prospective observational study.
Paediatr Anaesth
2013
;
23
:
920
6
2.
Disma
N
,
Virag
K
,
Riva
T
,
Kaufmann
J
,
Engelhardt
T
,
Habre
W
;
NECTARINE Group of the European Society of Anaesthesiology Clinical Trial Network
:
Difficult tracheal intubation in neonates and infants. NEonate and Children audiT of Anaesthesia pRactice IN Europe (NECTARINE): A prospective European multicentre observational study.
Br J Anaesth
2021
;
126
:
1173
81
3.
Fiadjoe
JE
,
Nishisaki
A
,
Jagannathan
N
,
Hunyady
AI
,
Greenberg
RS
,
Reynolds
PI
,
Matuszczak
ME
,
Rehman
MA
,
Polaner
DM
,
Szmuk
P
,
Nadkarni
VM
,
McGowan
FX
, Jr
,
Litman
RS
,
Kovatsis
PG
:
Airway management complications in children with difficult tracheal intubation from the Pediatric Difficult Intubation (PeDI) registry: A prospective cohort analysis.
Lancet Respir Med
2016
;
4
:
37
48
4.
Garcia-Marcinkiewicz
AG
,
Adams
HD
,
Gurnaney
H
,
Patel
V
,
Jagannathan
N
,
Burjek
N
,
Mensinger
JL
,
Zhang
B
,
Peeples
KN
,
Kovatsis
PG
,
Fiadjoe
JE
;
PeDI Collaborative
:
A retrospective analysis of neuromuscular blocking drug use and ventilation technique on complications in the Pediatric Difficult Intubation Registry using propensity score matching.
Anesth Analg
2020
;
131
:
469
79
5.
Burjek
NE
,
Nishisaki
A
,
Fiadjoe
JE
,
Adams
HD
,
Peeples
KN
,
Raman
VT
,
Olomu
PN
,
Kovatsis
PG
,
Jagannathan
N
,
Hunyady
A
,
Bosenberg
A
,
Tham
S
,
Low
D
,
Hopkins
P
,
Glover
C
,
Olutoye
O
,
Szmuk
P
,
McCloskey
J
,
Dalesio
N
,
Koka
R
,
Greenberg
R
,
Watkins
S
,
Patel
V
,
Reynolds
P
,
Matuszczak
M
,
Jain
R
,
Khalil
S
,
Polaner
D
,
Zieg
J
,
Szolnoki
J
,
Sathyamoorthy
K
,
Taicher
B
,
Riveros Perez
NR
,
Bhattacharya
S
,
Bhalla
T
,
Stricker
P
,
Lockman
J
,
Galvez
J
,
Rehman
M
,
Von Ungern-Sternberg
B
,
Sommerfield
D
,
Soneru
C
,
Chiao
F
,
Richtsfeld
M
,
Belani
K
,
Sarmiento
L
,
Mireles
S
,
Bilen Rosas
G
,
Park
R
,
Peyton
J
;
PeDI Collaborative Investigators
:
Videolaryngoscopy versus fiber-optic intubation through a supraglottic airway in children with a difficult airway: An analysis from the multicenter Pediatric Difficult Intubation Registry.
Anesthesiology
2017
;
127
:
432
40
6.
Peyton
J
,
Park
R
,
Staffa
SJ
,
Sabato
S
,
Templeton
TW
,
Stein
ML
,
Garcia-Marcinkiewicz
AG
,
Kiss
E
,
Fiadjoe
JE
,
von Ungern-Sternberg
B
,
Chiao
F
,
Olomu
P
,
Zurakowski
D
,
Kovatsis
PG
;
PeDI Collaborative Investigators
:
A comparison of videolaryngoscopy using standard blades or non-standard blades in children in the Paediatric Difficult Intubation Registry.
Br J Anaesth
2021
;
126
:
331
9
7.
von Elm
E
,
Altman
DG
,
Egger
M
,
Pocock
SJ
,
Gøtzsche
PC
,
Vandenbroucke
JP
;
STROBE Initiative
:
The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies.
Lancet
2007
;
370
:
1453
7
8.
Cook
TM
:
A new practical classification of laryngeal view.
Anaesthesia
2000
;
55
:
274
9
9.
Graciano
AL
,
Tamburro
R
,
Thompson
AE
,
Fiadjoe
J
,
Nadkarni
VM
,
Nishisaki
A
:
Incidence and associated factors of difficult tracheal intubations in pediatric ICUs: A report from National Emergency Airway Registry for Children: NEAR4KIDS.
Intensive Care Med
2014
;
40
:
1659
69
10.
Nishisaki
A
,
Turner
DA
,
Brown
CA
, 3rd
,
Walls
RM
,
Nadkarni
VM
;
National Emergency Airway Registry for Children (NEAR4KIDS); Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network
:
A national emergency airway registry for children: Landscape of tracheal intubation in 15 PICUs.
Crit Care Med
2013
;
41
:
874
85
11.
Ho
D
,
Imai
K
,
King
G
,
Stuart
EA
.
MatchIt: Nonparametric preprocessing for parametric causal inference
.
J Stat Softw
2011
;
42
:
28
.
12.
Pedroza
C
,
Truong
VT
:
Performance of models for estimating absolute risk difference in multicenter trials with binary outcome.
BMC Med Res Methodol
2016
;
16
:
113
13.
Péan
D
,
Floch
H
,
Beliard
C
,
Piot
B
,
Testa
S
,
Bazin
V
,
Lejus
C
,
Asehnoune
K
:
Propofol versus sevoflurane for fiberoptic intubation under spontaneous breathing anesthesia in patients difficult to intubate.
Minerva Anestesiol
2010
;
76
:
780
6
14.
Tobias
JD
:
Sedation of infants and children outside of the operating room.
Curr Opin Anaesthesiol
2015
;
28
:
478
85
15.
Chernik
DA
,
Gillings
D
,
Laine
H
,
Hendler
J
,
Silver
JM
,
Davidson
AB
,
Schwam
EM
,
Siegel
JL
:
Validity and reliability of the Observer’s Assessment of Alertness/Sedation Scale: Study with intravenous midazolam.
J Clin Psychopharmacol
1990
;
10
:
244
51
16.
Cravero
JP
,
Askins
N
,
Sriswasdi
P
,
Tsze
DS
,
Zurakowski
D
,
Sinnott
S
:
Validation of the Pediatric Sedation State Scale.
Pediatrics
2017
;
139
:
e20162897
17.
Macnab
AJ
,
Levine
M
,
Glick
N
,
Susak
L
,
Baker-Brown
G
:
A research tool for measurement of recovery from sedation: The Vancouver Sedative Recovery Scale.
J Pediatr Surg
1991
;
26
:
1263
7
18.
Malviya
S
,
Voepel-Lewis
T
,
Tait
AR
,
Merkel
S
,
Tremper
K
,
Naughton
N
:
Depth of sedation in children undergoing computed tomography: Validity and reliability of the University of Michigan Sedation Scale (UMSS).
Br J Anaesth
2002
;
88
:
241
5

Appendix: PeDI Collaborative Investigators

The following collaborators do not meet all authorship criteria but contributed substantially to the work reported in this article.

From the Department of Anesthesiology, Alberta Children’s Hospital, Calgary, Alberta, Canada:

David Lardner, M.B.B.S., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesia and Pain Management, Perth Children’s Hospital, Nedlands, Western Australia, Australia:

Britta S. von Ungern-Sternberg, M.D., Ph.D., contributed as an investigator, collected data, and provided care for study patients.

David Sommerfield, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesia and Pain Management, Royal Children’s Hospital, Melbourne, Victoria, Australia:

Chris Holmes, M.D., contributed as an investigator, collected data, and provided care for study patients.

Stefano Sabato, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Pain Medicine, UC Davis Children’s Hospital, Davis, California:

Niroop Ravula, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Perioperative and Pain Medicine, Stanford University

Medical Center, Stanford, California:

Christine Jette, M.D., contributed as investigator, collected data, and provided care for study patients.

Sam Mireles, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, Hospital for Sick Children, Toronto, Ontario, Canada:

Clyde Matava, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Pharmacology and Therapeutics, British Columbia Children’s Hospital, Vancouver, British Columbia, Canada:

Simon Whyte, M.B.B.S., F.R.C.A., F.R.C.P.C., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, School of Medicine, Pontifical Catholic University of Chile, Santiago, Chile:

Eduardo Vega, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Perioperative Care, West China Medical Center of Sichuan University, Chengdu, China:

Lei Yang, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Instituto de Ortopedia Infantil Roosevelt, Bogotá, Colombia:

Piedad Echeverry-Marin, M.D., contributed as investigator, collected data, and provided care for study patients.

Carolina Pérez-Pradilla, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Pediatric Anesthesiology, Ann and Robert H. Lurie Children’s Hospital of Chicago, Feinberg School of Medicine, Northwestern University, Chicago, Illinois:

Narasimhan Jagannathan, M.D., M.B.A., contributed as an investigator, collected data, and provided care for study patients.

Nicholas E. Burjek, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Children’s Hospital of Colorado, Aurora, Colorado:

David Polaner, M.D., contributed as an investigator, collected data, and provided care for study patients.

Elizabeth Starker, M.D., contributed as an investigator, collected data, and provided care for study patients.

Judit Szolnoki, M.D., contributed as an investigator, collected data, and provided care for study patients.

Melissa Brooks-Peterson, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Pain and Perioperative Medicine, Children’s National Medical Center, Washington, D.C.:

Angela Lee, M.D., contributed as investigator, collected data, and provided care for study patients.

Eugenie Heitmiller, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, Johns Hopkins All Children’s Hospital, St. Petersburg, Florida:

Mohamed Rehman, M.D., contributed as an investigator, collected data, and provided care for study patients.

Allison Fernandez, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Spectrum Health Partners Maine, South Portland, Maine:

Jonathan Meserve, M.D., contributed as an investigator, collected data, and provided care for study patients.

Charles (Ted) Lord, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University, Baltimore, Maryland:

John McCloskey, M.D., contributed as an investigator, collected data, and provided care for study patients.

Nicholas Dalesio, M.D., M.P.H., contributed as an investigator, collected data, and provided care for study patients.

Rahul Koka, M.D., M.P.H., contributed as an investigator, collected data, and provided care for study patients.

Robert Greenberg, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Perioperative and Pain Medicine, Children’s Hospital of Boston, Harvard Medical School, Boston, Massachusetts:

Raymond Park, M.D., contributed as an investigator, collected data, and provided care for study patients.

James Peyton, M.B.Ch.B., M.R.C.P., F.R.C.A., contributed as an investigator, collected data, and provided care for study patients.

Mary Lyn Stein, M.D., contributed as an investigator, collected data, and provided care for study patients.

Chinyere Egbuta, M.D., contributed as an investigator, collected data, and provided care for study patients.

Stephen Flynn, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, Critical Care and Pain, Massachusetts General Hospital, Boston, Massachusetts:

Somaletha Bhattacharya, M.B.B.S., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Pediatric Anesthesiology, University of Michigan Health Center, Ann Arbor, Michigan:

Paul Reynolds, M.D., contributed as an investigator, collected data, and provided care for study patients.

Ian Lewis, M.D., contributed as an investigator, collected data, and provided care for study patients.

Bishr Haydar, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, National Institute of Pediatrics, Mexico City, Mexico:

Lina Sarmiento, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of Minnesota, Minneapolis, Minnesota:

Martina Richtsfeld, M.D., contributed as an investigator, collected data, and provided care for study patients.

Kumar Belani, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of Mississippi Medical Center, Jackson, Mississippi:

Sara Robertson, M.D., contributed as an investigator, collected data, and provided care for study patients.

Madhankumar Sathyamoorthy, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Washington University School of Medicine in St. Louis, St. Louis, Missouri:

Charles Schrock, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, Erasmus Medical Center Sophia’s Children Hospital Rotterdam, Rotterdam, The Netherlands:

Jurgen C. de Graaff, M.D. Ph.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of New Mexico, Albuquerque, New Mexico:

Codruta Soneru, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Perioperative Care and Pain Medicine, New York University School of Medicine, New York, New York:

Neeta Singh, D.O., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, New York Presbyterian Hospital–Weill Cornell Medical College, New York, New York:

Franklin Chiao, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Duke University, Durham, North Carolina:

Brad Taicher, D.O., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Wake Forest School of Medicine, Wake Forest, North Carolina:

Thomas Templeton, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Pain Management, Children’s Hospital of Cleveland Clinic, Cleveland, Ohio:

Pilar Castro, M.D., contributed as an investigator, collected data, and provided care for study patients.

N. Ricardo Riveros Perez, M.D., contributed as an investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Pain Medicine, Nationwide Children’s Hospital, Ohio State University, Columbus, Ohio:

Vidya T. Raman, M.D., contributed as investigator, collected data, and provided care for study patients.

Ralph Beltran, M.D., contributed as investigator, collected data, and provided care for study patients.

Tarun Bhalla, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Critical Care Medicine, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania:

Benjamin B. Bruins, M.D., contributed as investigator, collected data, and provided care for study patients.

Paul Stricker, M.D., contributed as investigator, collected data, and provided care for study patients.

Justin L. Lockman, M.D., M.S.Ed., contributed as investigator, collected data, and provided care for study patients.

Brian Struyk, M.D., contributed as investigator, collected data, and provided care for study patients.

Christopher Ward, M.D., contributed as investigator, collected data, and provided care for study patients.

Akira Nishisaki, M.D., M.S.C.E., contributed as investigator, collected data, and provided care for study patients.

Ramesh Kodavatiganti, M.D., contributed as investigator, collected data, and provided care for study patients.

Rodrigo J. Daly Guris, M.D., contributed as investigator, collected data, and provided care for study patients.

Mark S. Teen, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Instituto de Ortopedia Infantil Roosevelt, Bogotá, Colombia:

Piedad C. Echeverry Marín, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Vanderbilt University School of Medicine, Nashville, Tennessee:

Scott Watkins, M.D., contributed as investigator, collected data, and provided care for study patients.

Christy Crockett, M.D., contributed as investigator, collected data, and provided care for study patients.

John Moore, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Pain Management, University of Texas Southwestern, Dallas, Texas; the Children’s Health System of Texas, Dallas, Texas; and the Outcome Research Consortium, Cleveland, Ohio:

Tally Goldfarb, M.D., contributed as investigator, collected data, and provided care for study patients.

Patrick Olomu, M.D., contributed as investigator, collected data, and provided care for study patients.

Peter Szmuk, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Texas Children’s Hospital, Baylor College of Medicine, Houston, Texas:

Paul Hopkins, M.D., contributed as investigator, collected data, and provided care for study patients.

Chris Glover, M.D., M.B.A., contributed as investigator, collected data, and provided care for study patients.

Kim Nguyen, M.D., contributed as investigator, collected data, and provided care for study patients.

Thomas L. Shaw, M.D., contributed as investigator, collected data, and provided care for study patients.

Olutoyin Olutoye, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of Texas Medical School at Houston, Houston, Texas:

Ranu Jain, M.D., contributed as investigator, collected data, and provided care for study patients.

Maria Matuszczak, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology and Pain Medicine, Seattle Children’s Hospital, University of Washington School of Medicine, Seattle, Washington:

Agnes Hunyady, M.D., contributed as investigator, collected data, and provided care for study patients.

Adrian Bosenberg, M.D., contributed as investigator, collected data, and provided care for study patients.

See Tham, M.D., contributed as investigator, collected data, and provided care for study patients.

Daniel Low, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin:

Guelay Bilen-Rosas, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, University of Washington in St. Louis, St. Louis, Missouri:

James Fehr, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesia, University of California Los Angeles, Los Angeles, California:

Lisa K. Lee, M.D., contributed as investigator, collected data, and provided care for study patients.

Ihab Ayad, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Albert Einstein College of Medicine, New York, New York:

Roshan Patel, M.D., contributed as investigator, collected data, and provided care for study patients.

From the Department of Anesthesiology, Yale New Haven Hospital, New Haven, Connecticut:

Cheryl Gooden, M.D., contributed as investigator, collected data, and provided care for study patients.