Response from the Authors

Truth Has Nothing to Do with the Conclusion, and Everything to Do with the Methodology

High-Flow Nasal Cannula versus Continuous Positive Airway Pressure in Critical Bronchiolitis: A Randomized Controlled Pilot

We thank Dr. Alapont and colleagues for their interest in our recent publication.[1] [2] Nevertheless, we find their attempt to mischaracterize our study somewhat disappointing. Had Dr. Alapont and colleagues paid careful attention to the text, they would have noticed that this was a “pilot study” and we were cautious not to overstate our findings. In fact, we were particularly transparent in our discussion by stating that our relatively small sample size “…precluded us from drawing more definitive conclusions regarding equivalence of high flow nasal cannula (HFNC) compared with continuous positive airway pressure (CPAP).”[1] In other words, although treatment failure rates were “similar” between the groups, we could not categorically state that HFNC was “noninferior” to CPAP.

The objectives of our pilot study were simple and clear: (1) to assess the feasibility of a larger international noninferiority trial comparing HFNC to CPAP in children with critical bronchiolitis in a developing country; and (2) to measure treatment effect size in that population to perform sample size calculations to properly power such a trial. We believe that we met both the objectives.

Based on prior studies and the preponderance of data presented in our introduction, we believed that a null hypothesis with equivalence between HFNC and CPAP was most appropriate. We deliberately selected a clinically meaningful primary outcome, the need for escalation to either bilevel noninvasive ventilation or endotracheal intubation, since this is an outcome of consequence to the patient; it is also how we define failure of the initial strategy in our practice. Most importantly, we sought to avoid the pitfalls of the TRAMONTANE study,[3] which defined treatment failure as a nominal worsening of at least one out of four clinical criteria, resulting in an inflated failure rate that has since been challenged.[4]

The rate of treatment failure (primary outcome) in our study was 35.7% (10 out of 28 patients) in the CPAP group and 37.1% (13 out of 35 patients) in the HFNC group. In addition, CPAP and HFNC also performed similarly for the predetermined secondary outcomes of pediatric intensive care unit (PICU) length of stay (5 vs. 5 days, p = 0.46, respectively), and hospital length of stay (8 vs. 9 days, p = 0.95, respectively). In light of these findings, we believe that it is appropriate to state that “… both modalities had similar rates of treatment failure” and that “…HFNC resulted in a rate of treatment failure similar to CPAP.” Please note that “we did not claim” that HFNC was noninferior to CPAP. Such a statement could only be made had comparable results been found in a noninferiority trial large enough to show that HFNC outcomes did not cross the prespecified noninferiority margin, as described in the methodology and discussion of our a posteriori sample size calculations.

Lastly, we appreciate the suggestion by Dr. Alapont and colleagues that a Bayesian analysis might be useful in a study such as ours. However, we believe they were deceitful in their calculations by selecting the rate of intubation (not an a priori outcome in our study) as a variable in an ill-concealed forced attempt to strengthen their argument. We deliberately chose the composite primary outcome of escalation to bilevel noninvasive ventilation or intubation due to concerns that, in a resource-limited setting, patients who failed CPAP would more likely be transitioned to bilevel support (same equipment and different mode) while patients who failed HFNC would more likely escalate directly to intubation. When we enter our “real” primary outcome data, a treatment failure rate of 35.7% (10/28) for CPAP versus 37.1% (13/35) for HFNC, in a Bayesian analysis using the binary outcome beta-Bernoulli's model with a noninformative prior distribution (success probability of 0.5/0.5), the probability that the failure rate is higher in the HFNC group is 0.51 (or 0.49 in the CPAP group). We doubt many would attempt to argue, in good faith, that these probabilities are not similar.

Publication History

Received: 21 April 2020

Accepted: 14 May 2020

Article published online:
21 July 2020

• References

• 1 Cesar RG, Bispo BRP, Félix PHCA. et al. High-flow nasal cannula versus continuous positive airway pressure in critical bronchiolitis: a randomized controlled trial. J Pediatr Intensive Care 2020; 9 (04) 248-255
• 2 Modesto I, Alapont V, Medina A, Del Villar-Guerra P. Truth has nothing to do with the conclusion, and everything to do with the methodology. J Pediatr Intensive Care 2021; 10 (01) 83-84
• 3 Milési C, Essouri S, Pouyau R. et al; Groupe Francophone de Réanimation et d'Urgences Pédiatriques (GFRUP). High flow nasal cannula (HFNC) versus nasal continuous positive airway pressure (nCPAP) for the initial respiratory management of acute viral bronchiolitis in young infants: a multicenter randomized controlled trial (TRAMONTANE study). Intensive Care Med 2017; 43 (02) 209-216
• 4 Shein SL, Slain KN, Rotta AT, Milési C, Cambonie G. High-flow nasal cannula flow rate in young infants with severe viral bronchiolitis: the question is still open. Intensive Care Med 2019; 45 (01) 134-135