Early Prediction of Noninvasive Ventilation Failure in Critically Ill Patients with Acute Respiratory Failure: Role of Respiratory Rate, PaO₂/FiO₂ Ratio, and HACOR Score

Background: Noninvasive ventilation (NIV) is widely used for acute respiratory failure in critically ill patients. However, NIV failure requiring invasive mechanical ventilation can occur, and delayed recognition may worsen outcomes. This study evaluated readily available physiological predictors of NIV failure in ICU patients receiving NIV. Material and Methods: This prospective observational study included 120 adult patients with acute respiratory failure admitted to the ICU and initiated on NIV. Initial respiratory rate (RR) and PaO₂/FiO₂ ratio were recorded at NIV initiation, while the heart rate, acidosis, consciousness, oxygenation, and respiratory rate (HACOR) score was assessed after one hour. Patients were categorized into NIV success and failure groups. Results: NIV was successful in 72 (60.0%) patients, while 48 (40.0%) required invasive mechanical ventilation. Initial RR was significantly higher in the failure group than in the success group (38.30 ± 4.59 vs 31.10 ± 3.46 breaths/min, p < 0.001). The initial PaO₂/FiO₂ ratio was significantly lower among patients with NIV failure (190.27 ± 52.95 vs 266.84 ± 42.35, p < 0.001). A one-hour HACOR score >5 was present in 83.3% of failure cases compared with 4.2% of successful cases (p < 0.001). Higher RR independently increased the odds of failure (OR 1.60, 95% CI 1.36 to 1.88), whereas higher PaO₂/FiO₂ reduced the odds (OR 0.69, 95% CI 0.60 to 0.79). HACOR >5 was the strongest predictor (OR 115.00, 95% CI 28.85 to 458.44). AUCs were 0.900, 0.880, and 0.981 for RR, PaO₂/FiO₂, and HACOR, respectively. Conclusion: Initial RR, PaO₂/FiO₂ ratio, and one-hour HACOR score can help identify ICU patients at high risk of NIV failure, with HACOR demonstrating the strongest predictive performance.

Keywords: Noninvasive ventilation, Acute respiratory failure, HACOR score, Respiratory rate, PaO₂/FiO₂ ratio.