Methods for Determination of Individual PEEP for Intraoperative Mechanical Ventilation Using a Decremental PEEP Trial

Abstract

(1) Background: Individual PEEP settings (PEEPIND) may improve intraoperative oxygenation and optimize lung mechanics. However, there is uncertainty concerning the optimal procedure to determine PEEPIND. In this secondary analysis of a randomized controlled clinical trial, we compared different methods for PEEPIND determination. (2) Methods: Offline analysis of decremental PEEP trials was performed and PEEPIND was retrospectively determined according to five different methods (EIT-based: RVDI method, Global Inhomogeneity Index [GI], distribution of tidal ventilation [EIT VT]; global dynamic and quasi-static compliance). (3) Results: In the 45 obese and non-obese patients included, PEEPIND using the RVDI method (PEEPRVD) was 16.3 ± 4.5 cm H2O. Determination of PEEPIND using the GI and EIT VT resulted in a mean difference of −2.4 cm H2O (95%CI: −1.2;−3.6 cm H2O, p = 0.01) and −2.3 cm H2O (95% CI: −0.9;3.7 cm H2O, p = 0.01) to PEEPRVD, respectively. PEEPIND selection according to quasi-static compliance showed the highest agreement with PEEPRVD (p = 0.67), with deviations > 4 cm H2O in 3/42 patients. PEEPRVD and PEEPIND according to dynamic compliance also showed a high level of agreement, with deviations > 4 cm H2O in 5/42 patients (p = 0.57). (4) Conclusions: High agreement of PEEPIND determined by the RVDI method and compliance-based methods suggests that, for routine clinical practice, PEEP selection based on best quasi-static or dynamic compliance is favorable.