Abstract
BACKGROUND: Prone positioning (PP) has demonstrated its potential for improving outcomes in patients with ARDS who require invasive mechanical ventilation. However, the ability of prolonged proning to reduce mortality in patients with COVID-19 specifically, sessions lasting > 24 h remains uncertain. METHODS: In this retrospective cohort study, we examined 158 subjects with COVID-19 pneumonia who required mechanical ventilation due to moderate-to-severe ARDS. Seventy-six subjects were placed in standard PP and 82 in extended PP, defined as prone sessions lasting at least 32 h. Our primary aim was to evaluate the effect of EPP on 90-d survival in subjects with COVID-19 with acute severe respiratory failure. To ensure the reliability of our findings and to minimize bias, we applied 3 adjustment approaches: cardinality matching (CM), matching weighting (MW), and inverse probability of treatment weighting with stabilized and trimmed weights (SW). We used Kaplan-Meier curves and Cox proportional hazard models to analyze the effects of EPP on 90-d mortality and sensitivity analysis by calculating E-values. RESULTS: The overall crude 90-d mortality rate was 31.7%. The unadjusted 90-d mortality rates were 19.5% in the EPP group and 44.7% in the SPP group (hazard ratio [HR] 0.35 [95% CI 0.19- 0.63], P < .001). After adjustment for confounding factors using CM, MW, and SW, baseline covariates were balanced between the 2 groups. Subjects in the EPP group exhibited lower 90-d mortality rates after adjustment using CM (HR 0.42 [95% CI 0.23-0.79], P = .007), MW (HR 0.45 [95% CI 0.21-0.95], P = .036), or SW (HR 0.29 [95% CI 0.15-0.56], P < .001). CONCLUSIONS: Extended PP was associated with improved 90-d survival in subjects with COVID-19 undergoing mechanical ventilation for severe ARDS. These findings suggest the potential benefit of EPP in the management of COVID-19-related respiratory failure. Further research and prospective studies are warranted to confirm and elucidate the underlying mechanisms of this association.