Predicting and Preventing Ventilator-Induced Lung Injury

点击翻译按钮获取中文摘要

Project Summary Acute respiratory distress syndrome (ARDS) is a rapid onset respiratory failure that is caused by factors ranging from pneumonia to sepsis. The impact of ARDS is substantial with more than 200,000 cases per year in the United States and an estimated mortality rate of 40%. All ARDS patients are mechanically ventilated to overcome the derangements in lung function caused by pulmonary edema, surfactant inactivation, and alveolar collapse. However, this essential mechanical ventilation can cause additional ventilator-induced lung injured (VILI) through tissue overdistension (volutrauma), the cyclic collapse and reopening of small airways and alveoli (atelectrauma), and inflammatory effects (biotrauma). Since VILI is a risk in all ARDS patients, and a significant contributor to ARDS mortality, improvements in ventilatory management are a key step in improving ARDS survival. However, further refinement of ventilation protocols to reduce VILI is challenging because of differences between patients and the changes in lung function that occur over time as ARDS worsens or resolves. Because of this inter- and intra-patient variability, ventilation that is beneficial in one person can be harmful in another. To overcome this challenge, we postulate that ventilation should be guided using a VILI cost function that provides real-time feedback of ventilation safety by describing the amount of VILI that is occurring. Our study will define such VILI cost functions based on the changes in lung function, structure, and inflammation that are the result of injurious ventilation. Using the cost function as a guide, the optimally safe ventilation for each patient could be determined by manually adjusting the ventilator settings. However, given the large number of permutations of ventilation adjustments this is not a practical approach. Instead, we will develop a mathematical model to predict optimal ventilation for each patient. These simulations will be personalized by fitting to real time pressure-flow measurements and then used to find the ventilation pattern that minimizes the VILI Cost Function. The predicted optimally safe ventilation will then be applied, and the process repeated to account for changes in lung function over time. The potential benefits of the proposed study are substantial. The VILI cost functions we define will provide an essential measurement of ventilation safety. Our innovative approach to optimize lung-protective ventilation using predictive models may lead to decreased ARDS mortality by protecting the injured lung while, at the same time, reducing provider workload. The proposed system also represents a paradigm shift in the way that ventilation strategies are established. Instead of testing a strategy in animal models and then in the heterogeneous ARDS patient population, where the effect may be beneficial to some patients and harmful to others, focus may be directed towards identifying algorithms that predict and prevent VILI independent of ARDS phenotype and lung mechanical function.