Association between ARDS Etiology and Risk of Noninvasive Ventilation Failure

Rationale: The etiology of acute respiratory distress syndrome (ARDS) may play an important role in the failure of noninvasive ventilation (NIV). Objectives: To explore the association between ARDS etiology and risk of NIV failure. Methods: A multicenter prospective observational study was performed in 17 intensive care units in China from September 2017 to December 2019. Patients with ARDS who used NIV as a first-line therapy were enrolled. The etiology of ARDS was recorded at study entry. Results: A total of 306 patients were enrolled. Of the patients, 146 were classified as having pulmonary ARDS (ARDSp) and 160 were classified as having extrapulmonary ARDS (ARDSexp). From initiation to 24 hours of NIV, the respiratory rate, heart rate, arterial oxygen pressure (Pa_O)/fraction of inspired oxygen (Fi_O), and arterial carbon dioxide pressure improved slower in patients with ARDSp than those with ARDSexp. Patients with ARDSp experienced more NIV failure (55% vs. 28%; P < 0.01) and higher 28-day mortality (47% vs. 14%; P < 0.01). The adjusted odds ratios of NIV failure and 28-day mortality were 5.47 (95% confidence interval [CI], 3.04-9.86) and 10.13 (95% CI, 5.01-20.46), respectively. In addition, we combined the presence of ARDSp, presence of septic shock, age, nonpulmonary sequential organ failure assessment score, respiratory rate at 1-2 hours of NIV, and Pa_O/Fi_O at 1-2 h of NIV to develop a risk score of NIV failure. With the increase of the risk score, the rate of NIV failure increased. The area under the curve of the receiver operating characteristic was 0.84 (95% CI, 0.79-0.89) and 0.81 (0.69-0.92) in the training and validation cohorts, respectively. Using 5.5 as cutoff value to predict NIV failure, the sensitivity and specificity was good. Conclusions: Among patients with ARDS who used NIV as a first-line therapy, ARDSp was associated with slower improvement, more NIV failure, and higher 28-day mortality than ARDSexp. The risk score combined presence of ARDSp, presence of septic shock, age, nonpulmonary sequential organ failure assessment score, respiratory rate at 1-2 hours of NIV, and Pa_O/Fi_O at 1-2 hours of NIV has high accuracy to predict NIV failure among ARDS population.