Grant: Operating Grant

Knowing when to liberate patients from mechanical ventilation (i.e. removal of breathing or endotracheal tube or extubation) is critically important, as both prolonged ventilation and failed extubation are both associated with harm and risk of death. Our objective is to improve the safety of extubation by harnessing hidden information contained in the patterns of variation of heart and respiratory rate measured over intervals-in-time. Currently, to assess a patient's ability to be extubated, a spontaneous breathing trial (SBT) is routinely performed, where the level of ventilator support is reduced, and their response is observed in order to help predict if they will tolerate extubation (i.e. complete removal of ventilator support). Given that health is associated with a high degree of variation of physiologic parameters (e.g. heart and respiratory rate), and illness & stress are associated with a loss of variability, we aim to uncover the loss of variation as a measure of stress during SBT's. We hypothesize that maintaining stable heart rate and respiratory rate variability (HRV and RRV) throughout the SBT will predict subsequent successful extubation, and conversely, a reduction in either HRV or RRV manifest during a SBT predicts extubation failure. A pilot study has demonstrated feasibility, and compelling preliminary results. A website, centralized data storage and analysis, and a trans-disciplinary team of scientists are in place to definitively test this novel technology. Determination of when to extubate critically ill patients remains a high-stakes clinical challenge; and improved prediction of extubation failure has potential to save lives and reduce costs in critically ill patients.