High-flow nasal cannula in critical asthma: Time for the respiratory therapists to drive care!

Critical asthma is a common reason for admission to the pediatric intensive care unit (PICU)1, 2 that is associated with significant morbidity and high healthcare costs.3 Children with critical asthma generally require treatment with continuous bronchodilators, supplemental oxygen, and noninvasive respiratory support.1 Continuous bronchodilators and supplemental oxygen historically were administered via an aerosol mask interface. More recently, high-flow nasal cannula (HFNC) has been increasingly used both as a support modality in critical asthma and as a means of delivering nebulized medications.4, 5 The use of HFNC in critical asthma offers several potential advantages including decreased work of breathing, increased ventilation efficiency (anatomical dead space washout effect), administration of conditioned (heated and fully humidified) inspired gas, more precise control of FiO2, high tolerability and comfort, the ability to eat and drink during treatment, and reduced staff workload.6 A potential disadvantage of HFNC is that, under certain conditions, it has been shown to adversely affect aerosol deposition,7 although some in vitro studies have suggested the opposite.8, 9 Aerosol delivery during HFNC therapy is affected by flow, nebulizer type, prong size and its relation to the cross-sectional area of the nostril, and the HFNC system used.10, 11 Although several retrospective and observational studies have demonstrated the feasibility of HFNC therapy in children with critical asthma,12-15 most did not find observable differences in outcomes associated with its use.12, 13, 15 However, one study14 reported an association between HFNC use and a delay in initiating noninvasive ventilation (NIV) and a recent propensity-scored matched study using an administrative database noted HFNC use was associated with longer hospital length of stay.16 Despite the lack of high-quality evidence, HFNC is now ubiquitous to PICUs and its frequency of use in children with critical asthma is increasing.5, 17 In this issue of Pediatric Pulmonology, Maue et al.18 used standard quality improvement (QI) methodology combined with multivariable linear regression to evaluate the effect of adding a respiratory therapist (RT)-driven HFNC weaning protocol to their existing continuous albuterol weaning pathway for children with critical asthma.19 The primary outcome was duration of HFNC therapy. Secondary outcomes included PICU length of stay (LOS), hospital LOS, and duration of continuous albuterol therapy. The rates of PICU and hospital readmission, and rates of NIV and invasive mechanical ventilation use were included as balancing measures. A total of four PDSA cycles were performed related to HFNC. The first cycle was the implementation of a RT-driven HFNC protocol; the second cycle allowed continuous albuterol to be weaned simultaneously to HFNC, as patients on continuous albuterol had previously been excluded from their HFNC weaning protocol. The final two cycles included more rapid weaning of HFNC flow and incorporation of a HFNC holiday during weaning. To maximize uptake and increase the ease of use for RTs, their electronic medical record (EMR) was updated so interventions could be appropriately documented without overburdening the staff. They enrolled a total of 410 children over the course of the study and noted a decrease in HFNC treatment duration of 8.6 h during PDSA cycle 2, which was sustained throughout the subsequent PDSA cycles. PICU LOS also decreased by 10 h and hospital LOS decreased by 18.5 h during PDSA cycle 2, and each remained stable in subsequent cycles for the duration of the study. Invasive mechanical ventilation use decreased and readmission rates were higher during PDSA cycle 1, while other balancing measures remained stable over time. The lack of a meaningful improvement observed by Maue et al.18 on duration of HFNC therapy, PICU and hospital LOS, and duration of continuous albuterol treatment after PDSA cycle 2 was likely multifactorial. First, the team may have already fully realized the potential benefit from their protocol and HFNC weaning was not a significant driver of LOS. Second, as is the case in our PICUs at Duke University Medical Center, Riley Hospital for Children has many well-established RT-driven pathways in place, so RTs may already have felt empowered to accelerate weaning of HFNC before the rollout of a more aggressive weaning protocol during PDSA cycle 3. Lastly, despite the relatively short LOS, the rate of NIV use in their cohort was considerably higher (21% vs. 8%–10%) than in previous reports,12, 20, 21 thus potentially limiting the impact of a HFNC protocol. In their study, both the HFNC weaning protocol and the continuous albuterol weaning pathway were managed via a RT-driven process. RT-driven critical asthma pathways have been shown to either decrease hospital length of stay (LOS)20, 22 or equivalence19, 21 compared to physician management of critical asthma. Importantly, none of these prior studies include protocolized management of HFNC. It should be noted that RT protocols have been shown to have a positive impact on RTs' perception of their role in the PICU.19 This is an important consideration, since RT-driven protocols can lead to increase RT workload that has been associated with burnout,23 while satisfaction with their role may be protective against burnout.24 No clinical study is perfect, so the fact that the one by Maue et al.18 has a few notable limitations is unsurprising. First, the study did not report when HFNC treatment was used in relation to the clinical trajectory. Therefore, the reader is left to wonder how often HFNC was used as the initial modality as opposed to on the weaning phase following NIV. Second, different scoring systems were used for continuous albuterol and HFNC weaning, which could have confounded the results. It is also unclear whether the clinical scoring system used for HFNC in their study is valid in critical asthma, as it includes respiratory rate and work of breathing but no other parameters. Finally, they arbitrarily defined HFNC as ≥1 L/kg/min for children under 10 kg and >10 L/min for those above 10 kg, which is has not been validated. The lack of an accepted definition of what constitutes HFNC and how to set or adjust flow continue to pose a barrier to the systematic study of this treatment modality in various disease processes.4 The selection of a meaningful yet attainable primary outcome presents a significant challenge in the study of pediatric critical asthma. While the need for intubation to facilitate invasive mechanical ventilation could be considered a consequential primary outcome, its rarity with contemporary asthma management would necessitate prohibitively large sample sizes to adequately power such a study. Therefore, most investigators opt to assess surrogate markers of patient trajectory, such as treatment duration or LOS. Clinicians may be primarily interested in reducing PICU LOS to optimize resource utilization, but patients and their families may place greater importance on overall hospital LOS. Consequently, reductions in treatment duration or PICU LOS, without concurrent improvements in hospital LOS, may hold lesser importance than those that effectively reduce total hospitalization time. In this study, Maue et al.18 chose duration of HFNC use as the primary outcome, whereas Gates et al.12 used hospital LOS, Brennan et al.20 used PICU LOS, and others used time on continuous albuterol.19, 21 These differing primary outcomes make it challenging to directly compare findings across studies. Additional challenges in comparing various studies on HFNC in pediatric critical asthma arise from variations in the entry criteria used to define the study cohorts. For instance, Gates et al.12 evaluated HFNC in a critical asthma cohort with an Modified Pulmonary Index Score (MPIS) ≥8 and demonstrated a longer duration of continuous albuterol by 6 h and longer PICU LOS by 14 h compared to the study conducted by Maue et al.18 However, interestingly, the overall hospital LOS was shorter in the Gates study.12 Furthermore, other studies have also reported shorter durations of continuous albuterol, PICU LOS, and hospital LOS in comparison to the current report.20, 21 These observed differences can likely be attributed to varying inclusion criteria, distinct protocol designs, and the use of different scoring systems across studies, making planning of a multicenter clinical trial notably challenging. Maue et al.18 should be applauded for their implementation of a standardized HFNC protocol for critical asthma management. Their study contributes to the body of evidence that children with critical asthma treated with HFNC have favorable outcomes. Most importantly, their findings provide further support for the efficacy of RT-driven clinical protocols. Moving forward, the logical next step is to conduct a prospective pragmatic clinical trial with protocolized management of bronchodilators and HFNC. Before more robust and definitive data on the effectiveness of these interventions, centers should at least incorporate RT-driven protocols to facilitate more uniform and efficient critical asthma management. Andrew G. Miller: Conceptualization; writing—review & editing; writing—original draft. Kyle J. Rehder: Conceptualization; writing—review & editing. Alexandre T. Rotta: Conceptualization; writing—review & editing. None. Mr. Miller is a section editor for Respiratory Care and has received honorarium for Saxe Communications, S2N Health, and Fisher and Paykel. Dr. Rotta discloses relationships with Breas US, and Elsevier. The remaining author declares no conflict of interest.