Model-Based Approach to Estimate dFRC in the ICU Using Measured Lung Dynamics

摘要

Acute Respiratory Distress Syndrome (ARDS) is characterized by inflammation, filling of the lung with fluid and collapsed lung unit. Mechanical ventilation (MV) is used to treat ARDS/ALI using positive end expiratory pressure (PEEP) to recruit and retain lung units, thus increasing pulmonary volume and dynamic functional residual capacity (dFRC) at the end of expiration. However, simple methods to measure dFRC at the bedside currently do not exist and other methods are invasive and impractical to carry out on a regular basis. Stress-strain theory is used to estimate ΔdFRC, which represents the extra pulmonary volume due to PEEP, utilizing readily available patient data from a single breath. The model uses commonly controlled or measured parameters (lung compliance, plateau airway pressure, PV data) to identify a parameter β_1 as a function of PEEP and tidal volume. A median β_1 value is calculated for each PEEP level over a cohort and is hypothesised as a constant throughout the population for the particular PEEP. Estimated ΔdFRC values are then compared to measured values to assess accuracy of the model. ΔdFRC was calculated for 9 patients and compared to the measured values. The median percentage error was 40.29% [IQR: 14.20-55.39] for PEEP = 5cmH_2O, 31.12% [IQR: 10.53-192.71] for PEEP = 10cmH_2O, 20.8% [IQR: 7.51-81.06] for PEEP = 15cmH_2O, 15.44% [IQR: 11.92-36.18] for PEEP = 20cmH_2O, 19.7% [IQR: 4.79-20.76] for PEEP = 25cmH_2O and 11.78% [IQR: 2.99-27.5] for PEEP = 30cmH_2O. Linear regression between estimated and measured ΔdFRC produced R~2 = 0.862. The model-based approach offers a simple and non-invasive method which does not require interruption of MV to estimate dFRC. The clinical accuracy of the model is limited but was able to track the impact of changes in PEEP and tidal volume on dFRC, on a breath-by-breath basis for each PEEP.