Early high-flow group (n=739) | Standard oxygen±rescue high-flow group (n=733) | |

A. Initial intervention cost per infant | ||

Capital cost of high-flow equipment, mean (SD) | 3.53 (3.23) | – |

Consumables, mean (SD) | 136.61 (0.00) | 2.64 |

(A) Total intervention cost per infant (capital+consumables), mean (SD) | 139.21 (11.73) | 2.64 |

B. Costs for rescue high-flow therapy (n=165) | ||

Capital (high-flow equipment) for those receiving rescue high-flow, mean (SD) | – | 4.32 (3.97) |

Consumables for those receiving rescue high-flow, mean (SD) | – | 136.61 (0.00) |

(B) Total costs associated with rescue high-flow therapy, mean (SD) | – | 141.01 (19.30) |

C. Costs associated with length of stay for bronchiolitis episodes of care, mean (SD) | ||

Cost of hospital length of stay—non-ICU stay, mean (SD) | 5552 (1593) | 5596 (1970) |

Cost of ICU stay for those with an ICU stay, mean (SD, n) | 13 415 (8679, n=87) | 13 889 (11 823, n=65) |

(C) Total costs associated with bronchiolitis episode of care, mean (SD) | 7131 (5469) | 6827 (5701) |

D. Hospital transfer and retrieval cost, mean (SD, n) | 2442 (221, n=20) | 2406 (142, n=15) |

E. Total cost per infant (A+B+C+D) | ||

Mean (SD) | 7314 (5586) | 6893 (5809) |

Median (IQR) | 5568 (5567–5572) | 5431 (5431–5569) |

Differential mean cost (95% CrI)* | 420 (−176 to 1002) | |

F. Total cost up to treatment failure or until discharge if responder | ||

Mean (SD) | 5681 (1575) | 5492 (584) |

Median (IQR) | 5568 (5567–5572) | 5431 (5431–5569) |

Differential mean cost (95% CrI)* | 189 (67 to 311) | |

G. Effectiveness | ||

Treatment failure rate | 0.118 (0.094 to 0.141) | 0.228 (0.197 to 0.258) |

Differential treatment effect (95% CrI)* | −0.118 (−0.072 to −0.148) | |

H. Cost-effectiveness ratio, ie, mean incremental cost per treatment failure avoided, F/G (95% CrI) | 1778 (207 to 7096) |

The incremental cost per treatment failure avoided in the table is close to but not exactly equal to the difference in cost divided by the difference in effectiveness, due to the bootstrapping method employed and rounding error.

*95% non-parametric credible interval based on 10 000 bootstrap replications.

ICU, intensive care unit; IQR, interquartile range; SD, standard deviation.