Regional ventilation at EIT-guided PEEP and mortality in severe ARDS: a prospective observational study

Severe acute respiratory distress syndrome (ARDS) is characterized by regional mechanical heterogeneity and systemic inflammation. Electrical impedance tomography (EIT) can individualize positive end-expiratory pressure (PEEP), but the prognostic value of EIT-derived variables measured at an overdistension/collapse (OD/CL)-optimized PEEP remains uncertain. This single-center prospective observational cohort enrolled 69 adults with severe ARDS requiring invasive mechanical ventilation from January 2025 to March 2026. Patients underwent decremental EIT-guided PEEP titration, and optimal PEEP was defined as the level that minimized the combined burden of regional overdistension and collapse. Except for an exploratory trajectory analysis across PEEP levels, all EIT-derived variables used for outcome and correlation analyses were measured at the OD/CL-selected PEEP. The primary outcome was 28-day mortality. Multivariable logistic regression was supported by Firth penalized regression, bootstrap internal validation, and a sensitivity analysis excluding early deaths. Non-survivors had a lower center of ventilation than survivors (45.35 [43.44–46.48]% vs. 49.40 [46.49–55.53]%, p < 0.001), indicating a more ventral distribution of tidal ventilation, and higher interleukin-6 concentrations (1020.9 [414.6-23712.8] vs. 236.2 [64.4–2608.0] pg/mL, p = 0.001). During decremental titration, EIT-derived variables changed systematically with PEEP. In the primary model, APACHE II score, lower center of ventilation, and log₁₀ interleukin-6 were associated with mortality. Adding center of ventilation and interleukin-6 increased the AUC from 0.838 for APACHE II alone to 0.943; the optimism-corrected AUC was 0.931. As the study was single-center and exploratory, the model performance estimates should be interpreted cautiously. In this single-center exploratory cohort of patients with severe ARDS, lower center of ventilation measured after OD/CL-guided PEEP selection was associated with 28-day mortality. These findings suggest that EIT-derived regional ventilation may characterize clinically relevant physiological heterogeneity, but they do not establish clinical utility or treatment guidance and require external validation.