Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

Bayes to the Rescue: Continuous Positive Airway Pressure Has Less Mortality Than High-Flow Oxygen

Modesto i Alapont, Vicent MD, PhD1; Khemani, Robinder G. MD, MsCI2; Medina, Alberto MD, PhD3,4; del Villar Guerra, Pablo MD, PhD5; Molina Cambra, Alfred MD1

Pediatric Critical Care Medicine: February 2017 - Volume 18 - Issue 2 - p e92-e99
doi: 10.1097/PCC.0000000000001055
Online Clinical Investigations
Buy
SDC

Objectives: The merits of high-flow nasal cannula oxygen versus bubble continuous positive airway pressure are debated in children with pneumonia, with suggestions that randomized controlled trials are needed. In light of a previous randomized controlled trial showing a trend for lower mortality with bubble continuous positive airway pressure, we sought to determine the probability that a new randomized controlled trial would find high-flow nasal cannula oxygen superior to bubble continuous positive airway pressure through a “robust” Bayesian analysis.

Desing, Setting, Patients, and Interventions: Sample data were extracted from the trial by Chisti et al, and requisite to “robust” Bayesian analysis, we specified three prior distributions to represent clinically meaningful assumptions. These priors (reference, pessimistic, and optimistic) were used to generate three scenarios to represent the range of possible hypotheses. 1) “Reference”: we believe bubble continuous positive airway pressure and high-flow nasal cannula oxygen are equally effective with the same uninformative reference priors; 2) “Sceptic on high-flow nasal cannula oxygen”: we believe that bubble continuous positive airway pressure is better than high-flow nasal cannula oxygen (bubble continuous positive airway pressure has an optimistic prior and high-flow nasal cannula oxygen has a pessimistic prior); and 3) “Enthusiastic on high-flow nasal cannula oxygen”: we believe that high-flow nasal cannula oxygen is better than bubble continuous positive airway pressure (high-flow nasal cannula oxygen has an optimistic prior and bubble continuous positive airway pressure has a pessimistic prior). Finally, posterior empiric Bayesian distributions were obtained through 100,000 Markov Chain Monte Carlo simulations.

Measurements and Main Results: In all three scenarios, there was a high probability for more death from high-flow nasal cannula oxygen compared with bubble continuous positive airway pressure (reference, 0.98; sceptic on high-flow nasal cannula oxygen, 0.982; enthusiastic on high-flow nasal cannula oxygen, 0.742). The posterior 95% credible interval on the difference in mortality identified a future randomized controlled trial would be extremely unlikely to find a mortality benefit for high-flow nasal cannula oxygen over bubble continuous positive airway pressure, regardless of the scenario. Interpreting these findings using the “range of practical equivalence” framework would recommend rejecting the hypothesis that high-flow nasal cannula oxygen is superior to bubble continuous positive airway pressure for these children.

Conclusions: For children younger than 5 years with pneumonia, high-flow nasal cannula oxygen has higher mortality than bubble continuous positive airway pressure. A future randomized controlled trial in this population is unlikely to find high-flow nasal cannula oxygen superior to bubble continuous positive airway pressure.

1Pediatric Intensive Care Unit, Hospital Universitari i Polotècnic, València. Spain.

2Anesthesiology and Critical Care Medicine, Children's Hospital Los Angeles, University of Southern California Keck School of Medicine, Los Angeles, CA.

3Pediatric Intensive Care Unit, Area de Gestión Clínica de Pediatría, Hospital Universitario Central de Asturias, Oviedo, Spain

4CIBER-Enfermedades Respiratorias, Instituto de Salud Carlos III, Spain.

5Department of Pediatrics, Segovia General Hospital, Segovia. Spain.

Drs. Modesto i Alapont, Khemani, and Medina conceived the study, analyzed the data, interpreted the results, and wrote the initial drafts of the article. Drs. del Villar Guerra and Molina Cambra revised the article and approved submission. All authors read and approved the final article.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/pccmjournal).

The authors have disclosed that they do not have any potential conflicts of interest.

For information regarding this article, E-mail: amedinavillanueva@gmail.com

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies