Nasal high flow is widely used in critically ill patients admitted to the intensive care unit (ICU) for acute hypoxemic respiratory failure. It has been shown to improve patient comfort, increase oxygenation and reduce need for intubation in some patients. The Respiratory Oxygenation (ROX) index has been established as a simple tool to help clinicians identify those patients who will succeed and those who will fail under nasal high flow and therefore predict the need for intubation. However, when nasal high flow therapy is successful, little is known as to how and when weaning of this device should be performed and what are the predictors of a safe withdrawal of the device. The objectives of this retrospective exploratory study are to identify a cut-off value of the ROX index predictive of success of the withdrawal trial, to describe a one-year use of the withdrawal trial; to describe the ROX value closest to weaning from nasal high flow, and to identify factors associated with success or failure of the withdrawal trial from nasal high flow therapy in patients receiving nasal high flow therapy.

Although nasal high flow is widely used in critically ill patients admitted to the intensive care unit (ICU) for acute hypoxemic respiratory failure, and its benefits have been shown in multiple studies, little is known as to how and when weaning from the technique device should be performed and what are the predictors of a safe withdrawal. The Respiratory Oxygenation (ROX) index was established and tested to predict outcome of nasal high flow therapy in patients with acute hypoxemic respiratory failure. It combines three parameters available at the bedside that reflect patient's respiratory status: respiratory rate, the level of the inspired fraction of oxygen (FiO2) and pulse oximetry (SpO2). Importantly, the combination of these three parameters outweighed the performance of each parameter taken separately to predict nasal high flow therapy outcome. Since its initial publication, several studies coming from different countries have confirmed the potential for the ROX index to predict need for intubation in patients with acute hypoxemic respiratory failure, thus making a valuable tool accessible to everyone, including nurses, to monitor on a routine basis. Because it reflects the patient's respiratory status, the greater the score, the better the patient's condition. Hence, one may hypothesize that, at the other end of the spectrum of respiratory failure (i.e., at the stage of weaning and withdrawal of the high flow device when the patient has considerably improved), the ROX index may also help identify those patients who will be successfully separated from the nasal high flow support. To date, one study has suggested that among several parameters, the value of the ROX index could be associated with successful cessation of nasal high flow. Here, in a retrospective exploratory study, investigators aim to to identify a cut-off value of the ROX index predictive of success of the withdrawal trial, to describe a one-year use of the withdrawal trial (average duration of use of nasal high flow before weaning, number of weaning trials, number of failed and successful weaning trials.); to describe the ROX value closest to weaning from nasal high flow, and to identify factors associated with success or failure of the withdrawal trial from nasal high flow therapy in patients receiving nasal high flow therapy; A successful withdrawal is defined as the withdrawal of nasal high flow without necessity of placing the patient back on nasal high flow within the following 24h. A failed withdrawal is defined by the need to resume nasal high flow within the 24h hours following the withdrawal. The possible identification of a cut-off value may help clinicians initiate earlier weaning from nasal high flow.

Inclusion Criteria:

- ICU patient treated with nasal high flow

Exclusion Criteria:

• age < 18 years
• concomitant treatment with non-invasive ventilation
• acute hypercapnic respiratory failure

ICU patients treated with nasal high flow therapy