To the Editor: In the article by Franklin et al.(March 22 issue), 1 in which infants with bronchiolitis were assigned to receive low-flow oxygen (standard-therapy group) or high-flow oxygen (high-flow group) through a nasal cannula, the primary outcomes differed between groups. Treatment failure in the standard-therapy group allowed patients to be placed in the high-flow group, whereas failure in the high-flow group resulted in unknown interventions beyond transfer to an intensive care unit (ICU). In addition, other measures of failure (such as transfer or intubation) did not favor the use of high-flow therapy. Standard therapy worked well in 566 patients, suggesting that a large number of patients in the high-flow group may not have required its use. The use of high-flow therapy in such patients increases cost without a clear benefit. Escalation of therapy in the high-flow group occurred at significantly higher respiratory rates, suggesting that provider discomfort with modest levels of tachypnea drove the escalation of care in the low-flow group. In addition, in both groups, approximately the same number of patients who received high-flow therapy did not benefit, which implies that there is no real advantage in initiating high-flow therapy early rather than postponing it until after standard therapy has failed. Although high-flow therapy may help some patients with bronchiolitis and hypoxemia, it is premature to assume that it should be used universally.