Impact of higher versus lower PEEP on mortality in mechanically ventilated patients with Sepsis — A multicenter, multi-cohort observational analysis

BACKGROUND: Optimal ventilator strategies for septic patients remain uncertain, particularly regarding positive end-expiratory pressure (PEEP). While PEEP may affect hemodynamics through changes in intrathoracic pressure, its impact on clinical outcomes in sepsis is not well understood. This multi-cohort study explores the impact of higher versus lower PEEP settings on mortality and hemodynamics in sepsis. METHODS: We conducted a secondary analysis of 844 septic patients requiring mechanical ventilation across three cohorts (one multicentric prospective = Hyspec ICU, one monocentric prospective = SepDataNet, one monocentric retrospective = RetroSep). Patients were stratified into low PEEP groups according to the ARDSnet low PEEP table and high PEEP groups with PEEP levels above mentioned table. Primary endpoint was ICU or 30-day mortality depending on reported outcome data. Secondary endpoints included vasopressor requirement, lactate levels and fluid balance. RESULTS: In two cohorts, higher PEEP was associated with significantly lower mortality (RetroSep: 39.4% vs. 50.4% for ICU-mortality, p = 0.008; Hyspec-ICU: 22% vs. 41.7% for 30-day mortality, p = 0.029). The difference persisted after propensity score matching in the RetroSep cohort (38.4% vs. 49.4%; p = 0.046). Hemodynamic parameters such as lactate and vasoactive-inotropic scores were similar or improved in high PEEP groups. Hyperspectral imaging showed no adverse microcirculatory effects. CONCLUSION: In this multi-cohort analysis, higher PEEP (above ARDSnet low PEEP table) was associated with improved survival without evidence of hemodynamic deterioration. However, due to the observational nature of this analysis and possible confounders, the results should be considered hypothesis-generating and warrant confirmation in randomized trials.