ABSTRACT
BACKGROUND: Prone positioning (PP) is an established and commonly used lung recruitment method for intubated patients with severe acute respiratory distress syndrome, with potential benefits in clinical outcome. The role of PP outside the intensive care unit (ICU) setting is debated. OBJECTIVES: We aimed at assessing the role of PP in death and ICU admission in non-intubated patients with acute respiratory failure related to COronaVIrus Disease-19 (COVID-19) pneumonia. DESIGN: This is a retrospective analysis of a collaborative multicenter database obtained by merging local non-interventional cohorts. METHODS: Consecutive adult patients with COVID-19-related respiratory failure were included in a collaborative cohort and classified based on the severity of respiratory failure according to the partial arterial oxygen pressure to fraction of inspired oxygen ratio (PaO2/FiO2) and on clinical severity by the quick Sequential Organ Failure Assessment (qSOFA) score. The primary study outcome was the composite of in-hospital death or ICU admission within 30 days from hospitalization. RESULTS: PP was used in 114 of 536 study patients (21.8%), more commonly in patients with lower PaO2/FiO2 or receiving non-invasive ventilation and less commonly in patients with known comorbidities. A primary study outcome event occurred in 163 patients (30.4%) and in-hospital death in 129 (24.1%). PP was not associated with death or ICU admission (HR 1.17, 95% CI 0.78-1.74) and not with death (HR 1.01, 95% CI 0.61-1.67) at multivariable analysis; PP was an independent predictor of ICU admission (HR 2.64, 95% CI 1.53-4.40). The lack of association between PP and death or ICU admission was confirmed at propensity score-matching analysis. CONCLUSION: PP is used in a non-negligible proportion of non-intubated patients with COVID-19-related severe respiratory failure and is not associated with death but with ICU admission. The role of PP in this setting merits further evaluation in randomized studies.