Abstract
BACKGROUND: Patients who are admitted to the intensive care unit (ICU) for severe acute exacerbations of chronic obstructive pulmonary disease (COPD) have poor outcomes. Although international clinical practice guidelines cautiously recommend the routine use of systemic corticosteroids for COPD exacerbations, data are scarce and inconclusive regarding their benefit for most severe patients who require mechanical ventilation in the ICU. Furthermore, corticosteroids may be associated with an increased risk of infection, ICU-acquired limb weakness, and metabolic disorders. METHODS AND ANALYSIS: This study is an investigator-initiated, multicentre, randomized, placebo-controlled, double-blind trial comparing systemic corticosteroids to placebo during severe acute exacerbations of COPD in patients who require mechanical ventilation in French ICUs. A total of 440 patients will be randomized 1:1 to methylprednisolone (1 mg/kg) or placebo for 5 days, and stratified according to initial mechanical ventilation (non-invasive or invasive), pneumonia as triggering factor, and recent use of systemic corticosteroids (< 48 h). The primary outcome is the number of ventilator-free days at day 28, defined as the number of days alive and without mechanical invasive and/or non-invasive ventilation between randomization and day 28. Secondary outcomes include non-invasive ventilation (NIV) failure rate, duration of mechanical ventilation (invasive and/or NIV), circulatory support (vasopressor), outcomes related to corticosteroid adverse events (severe hyperglycaemia, gastrointestinal bleeding, uncontrolled arterial hypertension, ICU-acquired weakness, ICU-acquired infections, and delirium), lengths of ICU and hospital stay, ICU and hospital mortality, day 28 and day 90 mortality, number of new exacerbation(s)/hospitalization(s) between hospital discharge and day 90, and dyspnoea and comfort at randomization, ICU discharge, and day 90. Subgroup analyses for the primary outcome are planned according to stratification criteria at randomization.