Abstract Background Heterogeneity is an inherent nature of ARDS. Recruitment-to-inflation ratio has been developed to identify the patients who has lung recruitablity. This technique might be useful to identify the patients that match specific interventions, such as higher positive end-expiratory pressure (PEEP) or prone position or both. We aimed to evaluate the physiological effects of PEEP and body position on lung mechanics and regional lung inflation in COVID-19-associated ARDS and to propose the optimal ventilatory strategy based on recruitment-to-inflation ratio. Methods Patients with COVID-19-associated ARDS were consecutively enrolled. Lung recruitablity (recruitment-to-inflation ratio) and regional lung inflation (electrical impedance tomography [EIT]) were measured with a combination of body position (supine or prone) and PEEP (low 5 cmH2O or high 15 cmH2O). The utility of recruitment-to-inflation ratio to predict responses to PEEP were examined with EIT. Results Forty-three patients were included. Recruitment-to-inflation ratio was 0.68 (IQR 0.52–0.84), separating high recruiter versus low recruiter. Oxygenation was the same between two groups. In high recruiter, a combination of high PEEP with prone position achieved the highest oxygenation and less dependent silent spaces in EIT (vs. low PEEP in both positions) without increasing non-dependent silent spaces in EIT. In low recruiter, low PEEP in prone position resulted in better oxygenation (vs. both PEEPs in supine position), less dependent silent spaces (vs. low PEEP in supine position) and less non-dependent silent spaces (vs. high PEEP in both positions). Recruitment-to-inflation ratio was positively correlated with the improvement in oxygenation and respiratory system compliance, the decrease in dependent silent spaces, and was inversely correlated with the increase in non-dependent silent spaces, when applying high PEEP. Conclusions Recruitment-to-inflation ratio may be useful to personalize PEEP in COVID-19-associated ARDS. Higher PEEP in prone position and lower PEEP in prone position decreased the amount of dependent silent spaces (suggesting lung collapse) without increasing the amount of non-dependent silent spaces (suggesting overinflation) in high recruiter and in low recruiter, respectively.
