Initial High-Flow Nasal Oxygen versus Noninvasive Ventilation in Acute Hypercapnic Respiratory Failure: A Prospective Observational Study

Abstract The aim of the study is to evaluate the association between the initial respiratory support modality (high-flow nasal oxygen [HFNO] vs. noninvasive ventilation [NIV]) and 30-day mortality among patients presenting to the emergency department with acute hypercapnic respiratory failure (AHRF). This prospective observational cohort study was conducted in the emergency department of a tertiary care teaching hospital between September 25, 2019, and September 25, 2020. Adult patients with AHRF requiring noninvasive respiratory support were consecutively enrolled. Patients were analyzed according to the respiratory support modality ultimately continued after the early clinical reassessment. The primary outcome was 30-day mortality. A total of 141 patients were included; 71 (50.4%) received HFNO and 70 (49.6%) received NIV. Thirty-day mortality did not differ significantly between the HFNO and NIV groups (29.6 vs. 31.4%; p = 0.956). Similarly, treatment failure requiring invasive mechanical ventilation (52.1 vs. 55.7%; p = 0.795) and ICU admission (69.0 vs. 65.7%; p = 0.812) did not differ significantly between groups. After adjustment for age, sex, baseline pH, baseline PaCO2, and COPD status, the initial respiratory support modality was not independently associated with 30-day mortality (adjusted OR, 1.14; 95% CI, 0.52–2.51) or treatment failure (adjusted OR, 1.36; 95% CI, 0.64–2.86). The adjusted mortality model demonstrated limited discrimination while the treatment failure model showed modest discrimination. In this prospective observational cohort, initial HFNO was not associated with statistically significant differences in 30-day mortality, treatment failure, or ICU admission compared with NIV after adjustment for measured confounders. However, these findings should be interpreted cautiously because of the observational study design, residual confounding, and post-treatment selection bias. Randomized controlled trials are needed to determine the comparative effectiveness of HFNO and NIV in AHRF.