PEEP Titration in a Computational Model of Dynamic Lung Recruitment

Abstract Introduction: Ventilator-induced lung injury is associated with intratidal recruitment and derecruitment (R/D). Lung recruitment may be responsive to positive end-expiratory pressure (PEEP). However, intratidal R/D are pressure- and time-dependent phenomena. Bates et al. developed a mathematical model of R/D represented by a hysteretic process with a time constant [1]. In this study, we investigated the effect of PEEP on intratidal R/D during mechanical ventilation of the model with nonlinear lung compliance. Methods: The R/D model of Bates et al. [1] was coupled to a sigmoid pressure-volume (P-V) relationship for regional lung compliance [2] and a uniform distribution of critical opening pressures. 10,000 virtual patients were simulated with randomly sampled parameters for: 1) maximum critical opening pressure; 2) consistently-recruited lung fraction; 3) consistently-derecruited lung fraction; 4) respiratory system resistance; 5) R/D time constant; and 6) P-V sigmoid parameters. Mechanical ventilation was simulated for 60 breaths using a volume-controlled waveform, with 450 mL tidal volume, 15 min-1 rate, and 1:2 inspiratory:expiratory ratio. PEEP was varied from 0 to 20 cmH2O. Measured outcomes of the simulations were static compliance, end-expiratory recruitment, intratidal R/D, and intratidal overdistension (exceeding 90% of maximal inflation). Results: Across all virtual patients, intratidal R/D decreased with increasing PEEP, albeit with diminishing returns above 7 cmH2O (Figure). Overdistension increased rapidly at PEEP exceeding 12 cmH2O. For low PEEP, intratidal R/D was correlated with fast R/D time constants, small fractions of consistently-recruited lung, and large amounts of lung recruitable at low critical opening pressures. For high PEEP, intratidal R/D was correlated with fast R/D time constants and high maximum critical opening pressures. A subset of virtual patients exhibited local minima of intratidal R/D at PEEP equal to or slightly greater than that which maximized static compliance. Conclusions: This model of dynamic recruitment/derecruitment during ventilation highlights the sensitivity of optimal PEEP settings (i.e., balancing R/D and overdistension) to patient-specific distributions of lung recruitability and mechanics. Results are consistent with imaging studies of intratidal R/D and overdistension in injured lungs [3,4]. Furthermore, we demonstrate a mechanism whereby intratidal R/D increases with high driving pressure due to low compliance, resulting from either low end-expiratory recruitment or overdistension at high PEEP. References: [1] JHT Bates et al., Crit Care Explor 2(12):e0299, 2020. [2] JG Venegas et al., J Appl Physiol 84(1):389-395, 1998. [3] ARS Carvalho et al., Crit Care 11(4):R86, 2007. [4] AH Jonkman et al., Am J Resp Crit Care Med 208(1):25-38, 2023.