Endotracheal Tube Size Adjustments Within Seven Days of Neonatal Intubation Free

This study used registry-based data collected in 21 academic neonatal centers in the United States, Canada, Germany, the United Kingdom, Austria, Singapore, and Australia.

This was a retrospective multicenter cohort study using prospectively collected TI practice and outcome data in the National Emergency Airway Registry for Neonates (NEAR4NEOS).¹  TI data were collected using standardized operational definitions at each participating site and were entered into a centralized online secure database hosted at the Children’s Hospital of Philadelphia.^15,16  The institutional review boards at each site approved NEAR4NEOS data collection or deemed data collection exempt as a quality improvement initiative.

The NEAR4NEOS registry organizes TI events into discrete TI encounters. Each TI encounter may include multiple TI attempts if the first attempt is unsuccessful, and the size of ETT for each attempt is recorded. We included infants who had their first TI encounter recorded in the NEAR4NEOS registry between August 2016 and September 2022. Only infants with a first TI encounter performed via the oral or nasal route using exclusively uncuffed ETTs were evaluated. We excluded infants with a first TI encounter performed by a non-neonatology provider (otolaryngology, anesthesia, surgery), infants with a first TI encounter that had an ETT larger than 4.0 mm or missing ETT size data for the first or final attempt, infants with airway or cranial anomalies, and infants with a first TI encounter that did not lead to the successful placement of an ETT by the final attempt. The successful placement of an ETT during the first encounter was necessary to determine the outcome measures of downsizing and upsizing.

The ETT size used for the first attempt of the first TI encounter was categorized according to the patient’s weight at the time of TI recorded in the NEAR4NEOS registry. Downsizing of the ETT was defined as a TI encounter in which the final successful TI attempt was completed using a smaller ETT than was selected for the first attempt. Upsizing was defined as a subsequent TI encounter that occurred within 7 days and concluded with the placement of a larger ETT than was placed at the end of the first TI encounter captured in the NEAR4NEOS registry. A frequency of upsizing or downsizing >10% for a particular ETT size within individual 200 g weight subgroups was a priori defined as suboptimal based on investigator consensus.

Our primary outcome was any TI associated event (TIAE), and the secondary outcome was severe oxygen desaturation, following NEAR4NEOS operational definitions.¹  TIAEs occurred during the intubation or up to 20 minutes afterward; TIAEs were not the indication for the intubation. Multiple TIAEs could occur during an intubation; however, this would only be counted once for the composite outcome of “any TIAE.” TIAEs were further classified as severe and nonsevere per NEAR4NEOS; we examined severe TIAEs and nonsevere TIAEs as exploratory outcomes. Severe TIAEs were defined as any of the following: cardiac arrest, esophageal intubation with delayed recognition, hypotension requiring therapy, cardiac compressions <1 minute, laryngospasm, vomit with aspiration, gum or dental trauma, pneumothorax and/or pneumomediastinum, and direct airway injury. Nonsevere TIAEs were defined as any of the following: esophageal intubation with immediate recognition, dysrhythmia (including any duration of bradycardia <60 beats per minute without chest compressions performed), medication error, mainstem bronchial intubation, vomit without aspiration, hypertension requiring therapy, pain and/or agitation requiring additional medications causing a delay in intubation, epistaxis, and lip trauma. Severe oxygen desaturation was defined as ≥20% decrease in the oxygen saturation from the highest oxygen saturation immediately before the first TI attempt to the lowest saturation during the procedure.

Patient characteristics and intubation practice measures were summarized using standard descriptive statistics. We used Wilcoxon Rank Sum tests and χ-square tests to compare patient characteristics, intubation characteristics, and procedural adverse outcomes between TI encounters where the ETT size selected did or did not adhere to NRP recommendations. To examine the relationship between ETT size and the study outcomes, we stratified the cohort into 200 g weight subgroups. The frequency of ETT downsizing and upsizing were reported for the 2 most frequently selected ETT sizes in each weight subgroup. We analyzed the association between downsizing and procedural adverse outcomes for the entire cohort using χ-squared tests.

Within patient weight subgroups, we compared procedural adverse outcome rates between the 2 most common ETT sizes selected for the first attempt. We used logistic regression to evaluate the independent association between ETT size and any TIAE, and severe oxygen desaturation, nonsevere TIAEs, and severe TIAEs, controlling for variables that have previously been shown to be associated with decreased risk of adverse events during intubation (use of a paralytic medication, video laryngoscope, and first airway provider).^1,17 –19  Location of intubation (delivery room versus NICU) was not included in the model, as this is not associated with differences in procedural outcomes.²⁰  A sensitivity analysis was performed to evaluate whether the TI route (oral versus nasal) impacted the results of the logistic regression model. A robust sandwich variance estimator was used to calculate standard errors in all regression models to account for clustering by site. A P value < .05 was considered statistically significant. Statistical analyses were performed using Stata 17.0 (Stata Corp, College Station, TX).

A total of 8653 unique patients representing 21 centers that participate in the NEAR4NEOS registry were intubated during the study period, and 7293 patients were eligible for inclusion in the study cohort (Supplemental Fig 3). The ETT size for the first TI attempt was consistent with NRP recommendations in 5752 (78.9%) of these. Site level adherence with NRP recommendations for the first ETT size ranged from 46% to 100% (Supplemental Fig 4). Many patient and practice characteristics varied between TIs based on adherence with NRP recommendations for the first ETT size as shown in Table 1, but procedural outcomes did not differ between these groups in aggregate (Table 2).

For many 200 g weight subgroups examined, the first ETT size selected varied substantially (Fig 1). This variability was most pronounced in the weight subgroups of 1000 to 1199 g and 2000 to 2199 g, both of which are immediately above the NRP recommended threshold cutpoints of 1000 g and 2000 g. Clinicians selected an initial ETT size of 2.5 mm (NRP recommends a 3.0 mm) in 260 of 686 (37.9%) infants weighing 1000 to 1199 g and 3.0 mm (NRP recommends a 3.5 mm ETT) in 206 of 329 (62.6%) infants weighing 2000 to 2199 g.

When assessed in the full study cohort, the ETT was downsized from the first selected size in 5.0% of TI encounters and was upsized within 7 days in 1.5% TI encounters. Figure 2 demonstrates the frequency of downsizing and upsizing for the weight subgroups that surround the NRP recommended weight thresholds used to transition between ETT sizes. The rate of ETT downsizing was highest when NRP-recommended ETT sizes were used in weight subgroups that were immediately above NRP-based weight thresholds. The ETT was downsized from a 3.0 mm to a smaller ETT in 53 of 420 (12.6%) infants weighing 1000 to 1190 g and from a 3.5 mm to smaller ETT in 20 of 117 (17.1%) of infants weighing 2000 to 2199 g (Fig 2). Across the cohort, procedural adverse outcomes were more common in TI procedures where downsizing occurred compared with procedures without downsizing: any TIAE (32.9% vs 16.8%, P < .001) and severe oxygen desaturation (60.3% vs 40.1%, P < .001).

Among infants weighing 1000 to 1199 g, initial use of a 2.5 mm ETT versus the NRP-recommended 3.0 mm ETT was associated with lower adjusted odds of any TIAE (adjusted odds ratio [aOR] 0.62, 95% confidence interval [CI] 0.41–0.94) and severe oxygen desaturation (aOR 0.53, 95% CI 0.38–0.75) (Table 3). In the weight subgroup of 2000 to 2199 g, use of a 3.0 mm ETT versus the NRP-recommended 3.5 mm ETT was associated with lower adjusted odds of severe oxygen desaturation (aOR 0.55, 95% CI 0.34–0.89). Nonsevere TIAEs and severe TIAEs stratified by weight subgroup and initial endotracheal tube size are shown in the Supplemental Table 4. Results were similar in sensitivity analyses including route of intubation (oral versus nasal) in the model (Supplemental Tables 5 and 6).

We used a large contemporary multicenter quality improvement research database to identify data-supported weight-based thresholds for ETT size selection in infants undergoing TI. There are several important findings for infants weighing up to 200 g above the current NRP-based thresholds of 1000 g and 2000 g. First, within these weight subgroups, clinicians frequently chose ETTs sized 0.5 mm below NRP recommendations or downsized the ETT below NRP recommendations during the TI encounter. Second, upsizing the ETT within 7 days was uncommon when smaller than NRP-recommended ETT sizes were placed in these subgroups. Finally, using an ETT 0.5 mm smaller than the recommended size was independently associated with a reduction in the odds of important procedural adverse outcomes. These findings suggest that for infants weighing up to 200 g above the NRP-based thresholds, use of an ETT that is 0.5 mm smaller than presently recommended may be appropriate.

At present, there is no gold standard to determine the most appropriate ETT size for a given infant. In this study, we used both the frequency of ETT size change (downsizing and upsizing) and procedural adverse outcomes to identify optimal ETT sizes for individual weight subgroups. Downsizing was used as a clinical signal that the airway provider believed a smaller ETT was more appropriate after attempting to place a larger sized tube, presumably because of difficulty passing the ETT into the airway or concern for possible airway injury owing to perceived anatomic size mismatch. Upsizing served as a balancing measure indicating that a smaller ETT was not clinically satisfactory within 7 days after placement. We chose a 7-day interval by investigator consensus, as upsizing within this timeframe was deemed to be most likely attributable to a suboptimal ETT size rather than interval patient growth.

Based on the data collected through NEAR4NEOS, we are unable to ascertain the indication for downsizing or upsizing. However, the fact that the ETT was downsized or upsized, irrespective of the medical reason, serves as a clinically relevant indicator given that current tube size guidelines are not evidence-based. Additionally, given the medical implications of ETT size changes (association with procedural adverse outcomes or subjecting the patient to another TI procedure), identifying an ETT size that avoids a size change is important.

The observed association between downsizing and TIAEs is likely multifactorial. In some cases, the larger ETT can directly lead to trauma, an esophageal intubation, or epistaxis (during nasal intubation). In other cases, an initial failed attempt because of the use of an ETT that is too large can result in adverse events owing to prolongation of the TI encounter, predisposing the patient to physiologic deterioration, and the need for additional TI attempts.

Many intubation procedures in our cohort did not conform with NRP recommendations for ETT size. This phenomenon occurred most frequently among infants weighing just above 1000 g and 2000 g. We observed site-level variation in the proportion of intubations that followed NRP recommendations, but we were unable to determine why providers selected ETT sizes for individual intubations.

To our knowledge, this is the first large scale study to evaluate the optimal ETT size for infants undergoing TI. Previous published literature referenced guidelines as expert opinion or usual practice. Compared with previously published guidelines, NRP generally recommends larger ETTs based on weight. Previous guidelines had the upper weight limit for a 2.5 mm ETT ranging from 1000 to 2500 g and the upper weight limit for a 3.0 mm ETT ranging from 2500 to 4000 g.^7,12 –14  Our data reinforce that selecting an ETT larger than current NRP recommendations should be avoided given the suboptimal downsizing rates and associations with procedural adverse outcomes. Despite this, we found that larger than NRP-recommended ETTs are occasionally selected, especially near the current NRP weight thresholds. Autopsy airway measurements in premature infants have been shown to underestimate the size of endotracheal tube that can pass through the vocal cords, likely because of airway elasticity.¹⁰  This highlights the importance of the clinical data in this study to inform decisions around ETT sizing for TIs.

We chose to focus on infant weight as a basis for determining data-driven thresholds. In autopsy measurements, both gestational age and patient weight correlate strongly with airway size.¹⁰  However, given potential imprecision in gestational age estimates depending on when dating is performed during pregnancy, we chose to use patient weight to define patient subgroups. We acknowledge that patient weight is sometimes estimated in the delivery room; however, most TIs in our study were performed in the NICU, where patient weight has been measured.

This study has limitations. Data collected in NEAR4NEOS are self-reported and could be subject to reporting bias. We attempted to focus on the noninstrumented airway to best understand the ETT size thresholds. Therefore, we limited our study to the first TI for each patient in the NEAR4NEOS database to reduce the possibility of airway swelling or trauma from a previously placed ETT. Some infants were previously intubated at another institution that did not participate in NEAR4NEOS and therefore the first TI captured by NEAR4NEOS was not their first TI. ETTs are measured by the internal diameter, and the outer diameter may vary by manufacturer.²¹  We did not document the manufacturer of ETTs used for this study. For the minority of ETTs placed nasally, the size of the nares could have impacted ETT size selection; including nasal route, but in our models, it did not substantially change the observed results. The NEAR4NEOS database does not document the ETT leak post intubation, and postextubation airway outcomes that may be relevant, such as stridor or subglottic stenosis, were not consistently available for this cohort. Some infants with a large ETT leak may not have undergone ETT upsizing as the clinical team may have chosen to tolerate the degree of leak or to extubate to noninvasive respiratory support. Finally, it is possible that infants experienced other adverse outcomes related to ETT sizing that were not captured in this study.

Strengths of the study include the use of clinical data from a large international cohort. This is the first study to our knowledge to use clinical data to guide ETT size selection for neonatal TI. The observed variation in the initial ETT size selection within weight subgroups facilitated a robust analysis of outcomes for different ETT sizes within the same weight groups. Finally, the NEAR4NEOS data are well suited for this study. ETT size is collected for each TI attempt, which allowed us to determine instances of downsizing, and multiple TI encounters for the same patient are linked with a unique identifier in NEAR4NEOS, which allowed us to determine instances of upsizing within 7 days as a balancing measure.

For infants weighing 1000 to 1199 g and 2000 to 2199 g, the recommended ETT sizes of 3.0 mm and 3.5 mm, respectively, were commonly downsized during the procedure, and 0.5 mm smaller ETT sizes were commonly selected, rarely upsized within 7 days, and independently associated with reduced odds of procedural adverse outcomes. These results may inform evidence-based recommendations for ETT size selection during neonatal intubation.

aOR

adjusted odds ratio

CI

confidence interval

ETT

endotracheal tube

NEAR4NEOS

National Emergency Airway Registry for Neonates

NRP

Neonatal Resuscitation Program

TI

tracheal intubation

TIAE

tracheal intubation associated event