Assessment of Factors Related to Auto-PEEP.

BACKGROUND Previous physiological studies have identified factors that are involved in auto-PEEP generation. In our study, we examined how much auto-PEEP is generated from factors that are involved in its development. METHODS One hundred eighty-six subjects undergoing controlled mechanical ventilation with persistent expiratory flow at the beginning of each inspiration were enrolled in the study. Volume-controlled continuous mandatory ventilation with PEEP of 0 cm H2O was applied while maintaining the ventilator setting as chosen by the attending physician. End-expiratory and end-inspiratory airway occlusion maneuvers were performed to calculate respiratory mechanics, and tidal flow limitation was assessed by a maneuver of manual compression of the abdomen. RESULTS The variable with the strongest effect on auto-PEEP was flow limitation, which was associated with an increase of 2.4 cm H2O in auto-PEEP values. Moreover, auto-PEEP values were directly related to resistance of the respiratory system and body mass index and inversely related to expiratory time/time constant. Variables that were associated with the breathing pattern (tidal volume, frequency minute ventilation, and expiratory time) did not show any relationship with auto-PEEP values. The risk of auto-PEEP ≥5 cm H2O was increased by flow limitation (adjusted odds ratio 17; 95% CI: 6-56.2), expiratory time/time constant ratio <1.85 (12.6; 4.7-39.6), respiratory system resistance >15 cm H2O/L s (3; 1.3-6.9), age >65 y (2.8; 1.2-6.5), and body mass index >26 kg/m(2) (2.6; 1.1-6.1). CONCLUSIONS Flow limitation, expiratory time/time constant, resistance of the respiratory system, and obesity are the most important variables that affect auto-PEEP values. Frequency expiratory time, tidal volume, and minute ventilation were not independently associated with auto-PEEP. Therapeutic strategies aimed at reducing auto-PEEP and its adverse effects should be primarily oriented to the variables that mainly affect auto-PEEP values.