Dynamic ROX-based index incorporating heart rate and flow for predicting high-flow nasal cannula outcomes in acute respiratory failure: A retrospective single-centre cohort study.

BACKGROUND

The ROX index is widely used to assess response to high-flow nasal cannula (HFNC) therapy in acute respiratory failure (ARF), but static measurements may not adequately reflect evolving physiological changes during treatment.

OBJECTIVES

To derive and evaluate the performance of a dynamic, physiology-informed HFNC risk index incorporating heart rate (HR) and HFNC flow, and to compare its performance with conventional ROX-based indices.

METHODS

We conducted a retrospective cohort study of adults with ARF initiated on HFNC in a mixed intensive care unit between 2020 and 2023. Modified ROX-based indices incorporating the SpO₂/FiO₂ ratio, respiratory rate, HR, and HFNC flow were evaluated at baseline and at 2, 6, and 12 h after HFNC initiation. Discriminative performance for HFNC outcomes was assessed using the area under the receiver operating characteristic curve (AUROC), with pairwise comparisons performed using DeLong's test. Associations with HFNC failure were examined using Cox proportional-hazards regression.

RESULTS

Among 145 patients, 91 (62.8%) achieved HFNC success. At 6 h after HFNC initiation, the dynamic index ΔROX-HR-Flow demonstrated excellent discrimination for HFNC outcomes (AUROC 0.97; 95% CI 0.94-1.00). ΔROX-HR-Flow showed numerically higher performance than other ROX-based indices, although differences compared with the conventional ROX index were not statistically significant. Higher ΔROX-HR-Flow values were independently associated with a lower hazard of HFNC failure.

CONCLUSIONS

A dynamic ROX-based index incorporating HR and HFNC flow showed promising discriminative performance using non-invasive bedside parameters. Prospective multicentre validation is required before clinical adoption.