Predictors of HFNC Failure in Patients With AHRF Using Diaphragm Ultrasound

Predictors of High-Flow Nasal Cannula Failure (HFNC) in Patients With Acute Hypoxemic Respiratory Failure (AHRF) Using Diaphragm Ultrasound Parameters

Ultrasound of the diaphragm is an easily reproducible bedside tool that provides a non-invasive measurement of inspiratory muscle function and respiratory efficiency. The diaphragmatic thickness correlates with the strength and muscle shortening and is reflective of its contribution to the respiratory workload. Diaphragm thickening fraction (DTF) has been shown to be predictive of extubation failure in ventilated patients. However, whether measurements of DTF are predictive of high flow nasal cannula (HFNC) outcomes in patients with acute hypoxemic respiratory failure (AHRF) remains unknown. The objective of this study is to identify predictors of HFNC failure by diaphragm ultrasound and to compare its performance with the well-established ROX index.

Inclusion Criteria:

• Age ≧ 18; AND
• Acute hypoxemic respiratory failure (AHRF), defined by respiratory rate of greater than 25 breaths per minute and a ratio of the PaO2 to the fraction of inspired oxygen (PaO2 / FiO2) of less than 300, and use of accessory muscles of respiration or paradoxical abdominal motion; AND
• Required ventilatory support with high-flow nasal cannula (HFNC)

Exclusion Criteria:

• Patients suffering from hypercapnic respiratory failure; OR
• AHRF secondary to conditions that are indicated for non-invasive ventilation (NIV) (e.g. acute exacerbation of COPD, cardiogenic pulmonary edema); OR
• Use of NIV or invasive mechanical ventilation (IMV) prior to HFNC initiation; OR
• Patients with imminent need for endotracheal intubation and invasive mechanical ventilation (IMV); OR
• Patients with known or suspected diaphragm paralysis; OR
• Pregnancy; OR
• Patients with abdominal compartment syndrome; OR
• Use of HFNC for more than 12 hours prior to ICU admission; OR
• Patients with suboptimal diaphragm ultrasound image quality for data processing; OR