RESUMO
BACKGROUND: Multiple breath washout (MBW) is a lung function test that identifies the degree of ventilation inhomogeneity (VI) in the lungs. In vitro validation of MBW devices is recommended. So far, plastic lung models for MBW validation ignored variable degrees of VI. Our primary aim was to create a plastic lung model applicable for physiological lung volumes and variable VI. METHODS: A plastic box divided in two chambers was filled with water and ventilated in various lung volumes and respiratory rates. A ventilator was used for efficient gas distribution (model with low VI). An additional divider was inserted to create a model with high VI. The model was connected to commercial MBW devices and measurements were performed using different tracer gases and conditions. Primary outcome was the precision of generated functional residual capacity (FRC) and the ability to generate variable VI. The latter was estimated by lung clearance index (LCI) and expiratory phase III slopes (SIII). LCI was also compared to a mathematical model. FINDINGS: The intra-test variability for FRC was minimal, mean(SD) coefficient of variation 0.96(0.63)%, using different tracer gases under different conditions. Compared to the model with low VI, in the model with high VI LCI and washout SIII were significantly increased. LCI compared well to the mathematical model. INTERPRETATION: This novel lung model shows excellent precision in lung volumes and VI estimates independent of tracer gases and conditions. The model can mimic the lungs of patients with uneven gas distribution.