Awake Prone Positioning in COVID-19 Hypoxemic Respiratory Failure: Exploratory Findings in a Single-center Retrospective Cohort Study.

BACKGROUND: Awake prone positioning has been widely used in patients with COVID-19 respiratory failure to avoid intubation despite limited evidence. Our objective was to evaluate if prone positioning is associated with a reduced intubation rate when compared to usual care.

METHODS: This was a retrospective cohort study in the emergency department of a large quaternary hospital in Sao Paulo. We retrieved data from all admitted patients in need of oxygen supplementation (>3 L/min) and tachypnea (>24 ipm) from March 1 to April 30, 2020, excluding those who had any contraindication to the prone position or who had an immediate need for intubation. The primary endpoint was endotracheal intubation up to 15 days. Secondary outcomes included a 6-point clinical outcome ordinal scale, mechanical ventilation-free days, admission to the intensive care unit, and need of hemodialysis and of vasoactive drugs, all assessed at or up to 15 days. We analyzed unadjusted and adjusted effect estimates with Cox proportional hazards models, logistic regression, quantile regression, and sensitivity analyses using propensity score models.

RESULTS: Of 925 suspected COVID-19 patients admitted off mechanical ventilation, 166 patients fulfilled inclusion and exclusion criteria: 57 were exposed to prone positioning and 109 to usual care. In the intervention group, 33 (58%) were intubated versus 53 (49%) in the control group. We observed no difference in intubation rates in the univariate analysis (hazard ratio = 1.21, 95% confidence interval [CI] = 0.78 to 1.88, p = 0.39) nor in the adjusted analysis (hazard ratio = 0.90, 95% CI = 0.55 to 1.49, p = 0.69). Results were robust to the sensitivity analyses. Secondary outcomes did not differ between groups.

CONCLUSIONS: Awake prone positioning was not associated with lower intubation rates. Caution is necessary before widespread adoption of this technique, pending results of clinical trials.