Model-based PEEP selection in mechanically ventilated patients — First clinical trial results

Background: Current methods to treat acute respiratory distress syndrome (ARDS) and acute lung injury (ALI) do not provide standardised approaches to optimise positive end expiratory pressure (PEEP). Mathematical models provide a means to optimise PEEP selection. A clinical trial is conducted to validate mathematical models that enable PEEP selection based on alveolar recruitability. Methods: Patients diagnosed with ARDS and ALI were enrolled in a single centre study. A specialised recruitment manoeuvre was performed starting from zero PEEP (ZEEP) in increments of 5 cmH2O, finishing with PEEP that produced a peak airway pressure = 45 cmH2O. The trial used volume controlled ventilation using low tidal volumes of 6 ml per kilogram (max 500 mls). Raw pressure volume (PV) data measured by a pneumatachometer were used in a mathematical model to estimate alveolar recruitment, alveolar recruitability, and to then determine an optimal PEEP. Data from 3 patients recruited to date are used to provide initial validation and define a novel means of selecting PEEP based on current and predicted alveolar recruitment. This study was approved by the New Zealand Upper South Island A Regional Ethics Committee. Results: Optimal PEEP levels for Patients 1 and 3 ranged from 15 to 20 cmH2O. Optimal PEEP for Patient 2 ranged more broadly from 10 to 20 cmH2O. Clinically selected values were 10, 12 and 10 cmH2O respectively. The model also allowed the clinician to detect some mechanical failures. Conclusions: The choice of optimal PEEP for ARDS/ALI patients is highly debated, clinician-specific, and variable, resulting in variable care, length of stay and outcome. This paper provides a model-based method to protocolise the optimization of PEEP based on accepted clinical goals and patient-specific physiological response. Further validation on