A55-49 Crossover to Non-Invasive Ventilation After High-Flow Nasal Cannula Failure: Reanalysis of the Renovate Trial

Rationale

High-flow nasal cannula (HFNC) and noninvasive ventilation (NIV) are commonly used for the management of acute respiratory failure (ARF). However, whether escalating support from HFNC to NIV represents an effective rescue strategy or a potentially harmful delay to invasive ventilation is unknown.

Methods

We conducted an exploratory analysis of the RENOVATE trial, a large multicenter randomized clinical study designed to assess the non-inferiority of high-flow nasal cannula (HFNC) compared with noninvasive ventilation (NIV) across different causes of acute respiratory failure (ARF). The trial enrolled 1,800 patients with five etiologies: hypoxemic ARF in non-immunocompromised patients, hypoxemic ARF in immunocompromised patients, acute cardiogenic pulmonary edema (ACPE), hypercapnic ARF due to chronic obstructive pulmonary disease (COPD), and COVID-19-related ARF. Patients randomized to HFNC who subsequently crossed over to NIV were identified. We performed a Bayesian analysis to evaluate mortality and endotracheal intubation among patients who did versus did not undergo crossover, along with a descriptive analysis of baseline characteristics, timing of crossover, and outcomes.

Results

Ninety patients (5% of the overall cohort) crossed over from HFNC to NIV: 46 with COVID-19, 24 with non-immunocompromised hypoxemic ARF, 10 with COPD-related hypercapnic ARF, 8 with ACPE, and 2 with immunocompromised hypoxemic ARF. Median age was 63.6 years (SD 16.5), median SAPS 3 was 57 [52-64], and 60.5% of patients were enrolled in the ICU. Patients received a median of 2 [1-3] hours of noninvasive ventilation prior to randomization. At baseline, crossover patients had lower PaO₂/FiO₂ ratios (153.2 vs 176.7 mmHg, p = 0.001) and higher PaCO₂ levels (38 vs 36 mmHg, p = 0.02) than those without crossover. Most crossover events occurred within the first 48 hours after HFNC initiation, with the highest frequency on day 1. After adjustment for PaO₂/FiO₂ ratio, SpO₂, PaCO₂, and PaO₂, the composite outcome of endotracheal intubation or death within seven days was higher in the crossover group (51.5% vs 37.6%; marginal odds ratio 1.76, 95% credible interval 1.06-2.57; posterior probability 0.992). Mortality at 28 days was similar between groups (29.5% vs 22.3%).

Conclusion

In this cohort, patients who crossed over from HFNC to NIV experienced high rates of early endotracheal intubation and mortality across different etiologies of acute respiratory failure. Well-designed randomized studies are required to clarify the optimal escalation strategy in acute respiratory failure.

This abstract is funded by: Brazilian Ministry of Health