Metabolomics and Clinical Data to Predict High-Flow Nasal Cannula Failure and Optimize Respiratory Support in Acute Hypoxemic Respiratory Failure

This observational study aims to evaluate whether metabolomic signatures combined with routinely collected clinical and physiological data can improve the prediction of high-flow nasal cannula (HFNC) failure in patients with acute hypoxemic respiratory failure (AHRF). HFNC is widely used as first-line non-invasive respiratory support in AHRF; however, delayed recognition of treatment failure may lead to worse clinical outcomes, including delayed intubation and increased morbidity and mortality.

The study will prospectively enroll adult patients with AHRF treated with HFNC. Clinical variables, respiratory parameters, laboratory results, and patient outcomes will be collected during routine clinical care. Blood samples will also be obtained for metabolomic analysis to identify molecular profiles associated with HFNC success or failure.

The primary objective is to identify metabolomic and clinical predictors associated with HFNC failure and escalation of respiratory support. Secondary objectives include evaluating the association between metabolomic patterns and relevant clinical outcomes such as endotracheal intubation, duration of respiratory support, intensive care unit (ICU) length of stay, and mortality. The study also aims to develop predictive models integrating biological and clinical data to support personalized respiratory management strategies in AHRF.