Comparison of HACOR score and ROX index for predicting noninvasive ventilation failure in acute hypoxemic respiratory failure: a prospective study.

Abstract

Background: Acute hypoxemic respiratory failure (AHRF) is commonly managed with noninvasive ventilation (NIV) in the emergency department (ED), but delayed identification of NIV failure worsens outcomes. We compared the HACOR (Heart rate, Acidosis, Consciousness, Oxygenation, and Respiratory rate) score and the ROX (Respiratory rate-OXygenation) index for predicting NIV failure in ED patients with AHRF.

Methods: This prospective observational study enrolled adult patients with AHRF receiving NIV in the ED between December 2024 and December 2025. HACOR and ROX were calculated at baseline and 1 h post-NIV; score changes (ΔHACOR, ΔROX) were also assessed. The primary outcome was NIV failure requiring intubation within 24 h. Predictive performance was evaluated via receiver operating characteristic (ROC) curves and DeLong comparisons.

Results: Among 127 patients, 66 (52.0%) failed NIV. Non-survivors/failures had significantly worse scores at both timepoints. The 1-hour HACOR showed the highest discriminative ability (AUC 0.878, 95% CI 0.808-0.930), outperforming baseline HACOR, all ROX parameters, and ΔROX (all p ≤ 0.05). A cut-off > 5 yielded 96.7% sensitivity and a negative likelihood ratio of 0.05.

Conclusions: In ED patients with AHRF treated with NIV, the 1-hour HACOR score was superior to all ROX-based metrics for predicting NIV failure. Early HACOR reassessment may enable timely escalation of respiratory support and improve clinical decision-making.