Use of Prone Positioning for the Acute Respiratory Distress Syndrome Increased During the Coronavirus Disease 19 Pandemic

Rationale: Prone positioning in acute respiratory distress syndrome (ARDS) improves patient outcomes but has been underutilized. In this study, we hypothesize that prone positioning use has increased during the COVID-19 pandemic. Methods: We conducted a retrospective study of patients treated in the medical ICU of a large academic tertiary care hospital in Baltimore, Maryland. Use of prone positioning among patients with COVID-19 ARDS treated from March 20th, 2020 to June 16th, 2020 were compared to patients with ARDS in 2019. Potential participants were identified from a registry of patients admitted with acute hypoxemic respiratory failure. Inclusion criteria required use of mechanical ventilation, the presence of ARDS, and a PaO2/FiO2 of < 150 during the first 72 hours following intubation. The primary outcome was use of prone positioning within 48 hours of the first qualifying PaO2/FiO2. Secondary outcomes were time to prone positioning and in-hospital mortality. The proportions of patients placed in the prone position in 2019 versus 2020 was compared using Fisher's exact test. Logistic regression was used to examine the association of early prone position (within 6 hours) with inhospital mortality in univariate models and models adjusted for age, sex and sequential organ failure assessment (SOFA) score. Results: Of 43 patients with COVID-19 that met inclusion criteria, 35 (81%) were proned within 48 hours of meeting oxygen criteria compared to 5 (25%) of 20 qualifying ARDS cases in 2019 (p<0.001) (Figure 1). Among those patients in whom it was used, prone positioning was used within 6 hours of meeting in oxygen criteria in 37% vs. 10% of patients in the COVID-19 vs. pre-COVID-19 ARDS patients (p=0.04). Overall, 37% of COVID-19 participants and 50% of non-COVID-19 ARDS patients died. Those proned within 6 hours of meeting oxygen criteria had numerically lower mortality compared to those not proned or proned later (33 vs. 44%), but this was not statistically significantly associated with in-hospital mortality in univariate or adjusted logistic regression models (adjusted odds ratio=0.43, 95% CI 0.12-1.57). Conclusions: Use of prone positioning for patients with moderate to severe ARDS markedly increased during the COVID-19 pandemic. Larger studies are needed to define the changes in prone positioning frequency in different settings and to understand why and how this rapid change in practice occurred. This understanding may inform interventions to more broadly implement evidence based ARDS care in a sustained fashion.