PaCO₂ responders of prone positioning in acute respiratory distress syndrome had better survival rates: A multicenter cohort study in Taiwan

Background: Prone positioning has demonstrated decreased mortality in moderate to severe acute respiratory distress syndrome (ARDS) patients. This study aimed to investigate the gas exchange effect of prone positioning in moderate to severe ARDS patients. Methods: This retrospective study includes eight tertiary referral centers. Demographic data, laboratory data, treatment records, ventilator setting data, and outcomes were collected. Arterial blood gas data were collected before and 24 h after the initiation of prone positioning and under the supine position. PaO₂ responders were defined as the PaO₂/FiO₂ ratio increasing by ≥ 20 % or ≥20 mm Hg, while PaCO₂ responders were defined as PaCO₂ decreasing by ≥ 1 mm Hg after prone positioning for one day. Results: One hundred thirty-eight patients receiving prone positioning were enrolled, with 86 (62.3 %) being PaO₂ responders and 67 (48.6 %) being PaCO₂ responders. Mortality rates were not significantly different between PaO₂ responders and non-responders at day 28 (29.1 % vs. 38.5 %, P = 0.254) or day 60 (39.5 % vs. 50.0 %, P = 0.229). PaCO₂ responders had a significantly lower mortality rate at day 28 (20.9 % vs. 43.7 %, P = 0.004) and day 60 (31.3 % vs. 54.9 %, P = 0.005). The PaCO₂ responders had significant improvement in minute ventilation (10.2 ± 2.5 vs. 10.9 ± 2.8 L/min, P = 0.024), dynamic driving pressure (18.7 ± 4.1 vs. 17.6 ± 4.0 cmH₂O, P = 0.004), and lung compliance [21.4(17.8–27.6) vs. 23.5(19.9–29.0) L/cmH₂O, P = 0.002] after prone positioning. Conclusions: In this multicenter retrospective cohort study of patients with ARDS receiving prone positioning, PaCO₂ responders had better clinical outcomes than PaCO₂ non-responders.