High-Flow Oxygen and Other Noninvasive Respiratory Support Therapies in Bronchiolitis: Systematic Review and Network Meta-Analyses

Abstract

Objectives: We present a systematic review on the effectiveness of noninvasive respiratory support techniques in bronchiolitis.

Data sources: Systematic review with pairwise meta-analyses of all studies and network meta-analyses of the clinical trials.

Study selection: Patients below 24 months old with bronchiolitis who require noninvasive respiratory support were included in randomized controlled trials (RCTs), non-RCT, and cohort studies in which high-flow nasal cannula (HFNC) was compared with conventional low-flow oxygen therapy (LFOT) and/or noninvasive ventilation (NIV).

Data extraction: Emergency wards and hospitalized patients with bronchiolitis.

Data synthesis: A total of 3,367 patients were analyzed in 14 RCTs and 8,385 patients in 14 non-RCTs studies. Only in nonexperimental studies, HFNC is associated with a lower risk of invasive mechanical ventilation (MV) than NIV (odds ratio, 0.49; 95% CI, 0.42-0.58), with no differences in experimental studies. There were no differences between HFNC and NIV in other outcomes. HFNC is more effective than LFOT in reducing oxygen days and treatment failure. In the network meta-analyses of clinical trials, NIV was the most effective intervention to avoid invasive MV (surface under the cumulative ranking curve [SUCRA], 57.03%) and to reduce days under oxygen therapy (SUCRA, 79.42%), although crossover effect estimates between interventions showed no significant differences. The included studies show methodological heterogeneity, but it is only statistically significant for the reduction of days of oxygen therapy and length of hospital stay.

Conclusions: Experimental evidence does not suggest that high-flow oxygen therapy has advantages over LFOT as initial treatment nor over NIV as a rescue treatment.