NIV Modality Change in “do not intubate” patients with hypoxemic ARF: a physiological pilot study

Introduction: The physiological impact of switching non-invasive ventilation (NIV) modality after initial failure in patients with hypoxemic acute respiratory failure (hARF) and “do not intubate” (DNI) order is unexplored (Rochwerg B, ERJ 2017;50). Aims and Objectives: To assess whether changing NIV modality improves blood gas parameters in this patient population. Methods: Retrospective study of patients with severe COVID-19 admitted to a respiratory intermediate care unit (Nov 2020 - Mar 2021). Inclusion criteria: diagnosis of hARF (PaO₂ < 60 mmHg room air and PaO₂/FiO₂ 45 mmHg; hemodynamic instability; Glasgow Coma Scale < 8; ongoing best supportive care. Variable distribution assessed by Shapiro–Wilk test; Wilcoxon signed‑rank test for non-normal data. Results: Ten patients (median 76y, IQR:73–81) included. Baseline [median (IQR)]: PaO₂/FiO₂ 106 (59–133.5), PaO₂ 56.5 mmHg (49.2–80.5), FiO₂ pre-NIV 66.5% (44.5–85%). Mean HACOR and APACHE II scores at admission: 5.2 ± 1.48 and 11.3 ± 2.11. Mean duration of NIV support: 9.6 ± 4.6 days. Modality changes were heterogeneous, not protocolized. No differences observed between blood gas data collected within 24 hours before/after modality change: PaO₂ (p = 0.91), PaCO₂ (p = 0.62), PaO₂/FiO₂ (p = 0.26), HCO₃⁻(p = 0.26), SpO₂/FiO₂ (p = 0.29). Ventilator settings unchanged: IPAP (p = 0.51), EPAP (p = 0.77), FiO₂ (p = 0.34). Conclusions: Changing NIV modality after failure did not improve gas exchange in patients with hARF and DNI order. Exploratory design limits generalizability.