Coronavirus disease (COVID-19) can result in severe hypoxemic respiratory failure thatultimately may require invasive mechanical ventilation in the Intensive Care Unit (ICU).Although lifesaving, invasive mechanical ventilation is associated with high mortality,severe discomfort for patient, long-term sequelae, stress to loved-ones and high costsfor society. During the ongoing pandemic high number of invasively ventilated COVID-19patients overwhelmed ICU capacity.Non-invasive respiratory support, such as high flow nasal oxygen (HFNO) or non-invasiveventilation (NIV) have the potential to reduce the risk for invasive mechanicalventilation and in selected cases ICU admission. However, data from different studies areconflicting and studies performed in COVID-19 patients are of limited quality.Furthermore, identification of early predictors of HFNO/NIV treatment failure may preventunnecessary delay of initiation of invasive ventilation, which may be associated withadverse clinical outcome. The development and validation of a prediction model, thatincorporates readily available clinically data may prove pivotal to fine-tunenon-invasive respiratory support.The overall aim of the NORMO2 project is to investigate the role and risks of HFNO andNIV to improve outcome in hospitalized hypoxemic COVID-19 patients.