Noninvasive Respiratory Support in Pediatric Critical Asthma: What to Start and Where to Go?

Pediatric asthma creates a tremendous burden on children, families, and medical systems. In the United States, one of every 14 individuals below the age 18 suffers from asthma; in total, over 5 million children are affected. For children with asthma exacerbations, there are approximately 750,000 emergency department visits and 74,000 hospital admissions/y, contributing to the nearly $6 billion in yearly costs attributable to pediatric asthma.^1,2 Whereas asthma is rarely fatal in children, many children with an asthma exacerbation will develop severe respiratory distress that requires respiratory support.^3-5 Noninvasive respiratory support (NRS) devices, including CPAP, noninvasive ventilation (NIV), and high-flow nasal cannula (HFNC), are increasingly utilized forms of respiratory support to treat respiratory distress in lieu of intubation and invasive mechanical ventilation. In general, CPAP and NIV offer a higher form of respiratory support, whereas HFNC is thought to be more comfortable for patients and result in less barotrauma. However, as the use of NRS has increased, respiratory therapists and physicians have been left to make decisions about its implementation without high-quality evidence to guide practice. Multiple prior retrospective studies have compared the use of NIV with HFNC in pediatric asthma exacerbations.^6-8 However, prior studies have been small, with < 100 subjects; and overall, evidence to guide use of HFNC versus NIV in pediatric asthma is limited.

In this issue of Respiratory Care, Russi et al⁹ describe a large cohort of children with critical asthma managed with NRS. Using a high-quality international database of pediatric ICU (PICU) patients, the authors conducted a descriptive retrospective study of children ages 2–17 admitted to 125 different PICUs from 2017–2021 for asthma requiring NRS. Patients were excluded if they had tracheostomy dependence, congenital heart disease, pulmonary hypertension, acute bronchiolitis, acute chest syndrome, acute laryngotracheobronchitis, or intubation prior to NRS. The study analyzed 10,083 encounters and found that HFNC was the most common modality used on admission (65.1%), followed by NIV (31.4%) and CPAP (3.5%). Use of HFNC increased during the study period, reaching a peak of 71.9% in 2021, while use of NIV and CPAP decreased. The authors found high variability among centers, with SD of 29.6, 27.1, and 3.1 for HFNC, NIV, and CPAP, respectively. Those initiated on HFNC were younger (NIV 9 y, CPAP 7.6 y, HFNC 6.9 y, P < .001) and less sick (Pediatric Risk of Mortality scores: NIV 0.6%, CPAP 0.5%, HFNC 0.3%, P < .001) compared to those on CPAP or NIV. Subjects supported on NIV had a higher intubation rate (NIV 2.4%, CPAP 2.2%, HFNC 0.8%) and longer PICU length of stay (LOS), potentially related to greater illness severity. There was, however, no difference in mortality or pneumothoraces across the groups. Subjects initiated on HFNC were escalated to NIV or mechanical ventilation in 7.5% of the cases, and HFNC was associated with NRS failure in the authors’ multivariable regression model (adjusted odds ratio 2.4, P < .001).

The authors of this study readily acknowledge its limitations. The retrospective nature of the study limits causal inference between type of respiratory support and outcomes. However, the largest limitation comes from the lack of granularity in the reported data. The study does not report the reasons individuals were started on the different respiratory support modalities, the NRS parameters (pressures, flows, F_IO2 ) that were used, or the criteria for escalating respiratory support. Additionally, this study is not able to separate HFNC initiated solely for delivery of aerosolized medications such as albuterol versus its application for respiratory distress. Without these additional data, it is difficult to interpret the trends in HFNC failure. Apart from the respiratory support given, the study also does not address the various pharmacologic interventions administered to these subjects, which is especially important to interpreting results when considering the known variability in medication choice in critical asthma.¹⁰

The data Russi et al presents support two conflicting conclusions. First, HFNC was frequently used by critical care physicians in the initial management of critical asthma, and the vast majority (92.7%) of children did not require further escalation. Of those subjects who failed HFNC, only a small percentage (4.9%) then required further escalation to invasive mechanical ventilation. This suggests that critical asthma can in most cases be safely managed with the use of HFNC. However, this study also indicates that initial management with HFNC carries a higher failure rate than CPAP or NIV and thus may expose children to increased risks and/or respiratory distress as a result. These risks include intubation, and that those escalated from HFNC to CPAP or NIV required intubation at higher rates than those subjects initially started on NIV (4.9% vs 2.4%) suggests potential harm with delayed initiation of NIV, though this comparison was not statistically evaluated.

Overall, the growing use of HFNC in pediatric critical asthma noted by Russi et al seems to have outstripped the evidence to guide its use. Whereas small observational studies by Baudin¹¹ and Gates¹² and an emergency department–based randomized controlled trial by Ballestero¹³ showed potential benefit of HFNC versus standard oxygen therapy in clinical parameters, LOS was the same or longer. Additionally, a small observational study by Pilar⁶ of subjects with critical asthma has findings that echo Russi’s, showing potential harm in delay of NIV use with 40% of their cohort experiencing HFNC failure compared to 0% in the NIV group. Another factor that complicates interpretation is that prior studies have suggested ineffective delivery of nebulized bronchodilators when used with flows > 6 L/min.^14-16 Whereas Russi et al do not report HFNC prescriptions, previous studies have demonstrated that 1–2 L/min/kg flows were the most commonly used.⁹ This suggests that HFNC is often used in conjunction with aerosolized therapies in a way that reduces their efficacy and may be an explanation for the greater risk of treatment failure observed. These mixed data demonstrate that it remains unclear how to best deploy HFNC in those with severe asthma exacerbations and which flows are the most beneficial.

How then should the bedside provider respond to a child with asthma with worsening gas exchange and respiratory distress? HFNC remains a tempting choice due to the perceived greater comfort it provides, and the results of this study would suggest that trialing it is safe in most patients. However, providers who do trial HFNC in this population will have to stay vigilant in their selection of patients and be mindful of the flows they use in conjunction with continuous β₂ agonist nebulization to minimize the risk for treatment failure. Those who initiate patients with CPAP will have to be likewise watchful as it also demonstrated a high failure rate (21%), which may be associated with its low and declining use, either because of inadequate clinical efficacy or changing provider preferences. Finally, whereas NIV appears to be more favorable in those with more severe disease per Russi’s data, clinicians are still without clear evidence to guide their titration of this treatment to best improve patient outcomes. These questions will remain unresolved until high-quality comparative effectiveness trials are completed for the different NRS modalities and their variations of use. This is a challenging prospect in a population whose interactions with the PICU are short (median LOS 3.1 d) and will require close partnership with emergency medicine and primary care physicians to create pathways of care that are effective and evidence-based.

The authors of this study should be commended for their contributions to the difficult clinical question of NRS use in pediatric critical asthma. This study shows on a large scale the variations that exist in treating respiratory failure in these children and points our field toward a needed greater research focus on improving their clinical care in the PICU. We look forward with excitement to the increased attention and enthusiasm in our community toward finding the best treatments and modalities for children with critical asthma and hope that future national guidelines and prospective multi-center studies will improve and standardize care across different institutions.