Abstract
OBJECTIVE: To compare the effects of four positive end-expiratory pressure titration methods on respiratory system compliance over the first 3 days of mechanical ventilation and to analyze the agreement between derived positive end-expiratory pressure and compliance values among these methods immediately after randomization. METHODS: Single-center, randomized study acute respiratory distress syndrome patients were assigned to one of four groups based on positive end-expiratory pressure titration methods: electrical impedance tomography, transpulmonary pressure measured via an esophageal catheter, the best compliance approach with daily positive end-expiratory pressure titration, and a control group using a low positive end-expiratory pressure/ fraction of inspired oxygen table with adjustments as necessary. The primary outcome was mean respiratory system compliance over the first 3 days of mechanical ventilation. Immediately post-randomization, the best positive end-expiratory pressure according to each method was assessed for every patient, and within-patient agreement of titrated positive end-expiratory pressure and compliance for pairs of methods was calculated with the Bland-Altman method. RESULTS: Forty-nine patients participated. Compared to control, the mean difference in compliance was 0.03mL/cmH2O (95%CI -2.74 to 2.8) in the Electrical impedance tomography Group; 1.90mL/cmH2O (95%CI -0.98 to 4.78) in the Catheter Group, and 1.42mL/cmH2O (95%CI -1.35 to 4.19) in the best compliance group. Within-patient agreement of titrated positive end-expiratory pressure and compliance was poor, with 95% limits of agreement ranging from -9.3 to 9cmH2O for positive end-expiratory pressure and from -8.5 to 11.4mL/cmH2O for compliance. CONCLUSION: No significant differences in mean respiratory system compliance were found among positive end-expiratory pressure titration methods compared to control. The agreement between titrated positive end-expiratory pressure and respiratory system compliance using different methods was low.