ObjectiveTo evaluate the usefulness of transthoracic electrical bioimpedance in sedated and paralyzed patients with acute lung injury during mechanical ventilation with and without early application of positive end-expiratory pressure (PEEP).DesignProspective, repeated-measures study.SettingUniversity-affiliated intensive care center.PatientsTen patients with acute lung injury.InterventionsSimultaneous, three-paired cardiac output (CO) measurements by transthoracic electrical bioimpedance (TEB) and thermodilution (TD) were made at 0 and 15 cm H2O of PEEP.Measurements and Main ResultsThe average of the TD-CO measurements was 7.22 +/- 2.12 (SD) L/min during 0 cm H2O of positive end-expiratory pressure (ZEEP), and 6.91 +/- 1.72 L/min during PEEP (NS). The average of the TEB-CO measurements was 4.48 +/- 1.37 L/min during ZEEP, and 6.03 +/- 2.03 L/min during PEEP (p < .05). For each level of PEEP, bias and precision between methods were calculated. Bias calculations between TD-CO and TEB-CO ranged from -1.54 +/- 7.02 L/min at ZEEP to -2.52 +/- 4.28 L/min at PEEP, and -2.47 +/- 6.09 L/min for mixed data at ZEEP and PEEP. There was no significant correlation between the percent change with PEEP in TEB-CO and TD-CO (r2= .05, NS).ConclusionsIn patients with acute lung injury: a) the agreement between TEB-CO and TD-CO measurements is poor; b) agreement is not clinically improved by application of PEEP; and c) TEB cannot monitor trends in CO. (Crit Care Med 1998; 26:1441-1445)