A Study on the Assessment of Difficult Weaning in Patients With Acute Respiratory Failure Undergoing Mechanical Ventilation Based on Machine Learning EIT-pendelluft: A Multicenter Retrospective Study.
A Study on the Assessment of Difficult Weaning in Patients With Acute Respiratory Failure Undergoing Mechanical Ventilation Based on Machine Learning EIT-pendelluft: A Multicenter Retrospective Study.
Invasive mechanical ventilation (IMV) is a common method used in the ICU to assist breathing and promote gas exchange. When a patient's condition improves, the weaning process (WP) should be initiated. There is a risk of failure during weaning, especially for elderly patients with prolonged ICU stays or those suffering from respiratory or neuromuscular diseases. Therefore, accurately assessing the likelihood of successful weaning before initiating the process holds significant clinical importance.
The weaning process involves gradually reducing the intensity of ventilatory support, continuous control of respiratory failure, restoration of oxygen indices, reduction of inhaled oxygen concentration, improvement of various laboratory indicators, and objective clinical assessment. Weaning is generally categorized into three types: simple weaning, difficult weaning, and delayed weaning. Successful weaning is defined as a patient breathing for 48 hours without respiratory support and without signs of respiratory failure. If a patient fails the spontaneous breathing trial (SBT) and requires re-intubation within 48 hours or dies within 48 hours, it is considered a failure of weaning. Successful extubation is defined as stable breathing for 48 hours or more without the need for IMV after weaning. It is related to multiple factors, including the patient's general condition, control of related diseases, ventilation-perfusion ratio, cough and sputum clearance ability, lung compliance, and diaphragm function. Timely assessment of these factors is crucial for determining the optimal timing for weaning, thereby avoiding premature weaning or unnecessarily prolonging IMV to reduce complications and prevent adverse clinical outcomes. Recent studies have identified new predictive factors, but their predictive performance has not been satisfactory. Various methods have been attempted in clinical practice to predict successful weaning, such as using pressure support ventilation or T-tube for SBT. However, there are numerous factors influencing the success of weaning, some of which have not been adequately considered, and published clinical studies have not yielded satisfactory results. Currently, the most widely used SBT trial still has a failure rate of approximately 20%.
This study intends to apply electrical impedance tomography (EIT), which can be used at the bedside, as a non-radiative, real-time monitoring and diagnostic tool. It is currently widely used to monitor regional lung ventilation distribution, lung volume, and respiratory mechanical characteristics, assisting clinicians in more accurately observing real changes in lung ventilation and lung volume at the bedside. This study aims to monitor patients' respiratory oscillations through EIT, clarify their correlation with difficulties, and use it as a convenient, practical, and effective new predictive factor for assessing the success rate of weaning in patients with severe acute respiratory failure.
Inclusion Criteria:
- 1. It is anticipated that a spontaneous breathing trial (SBT) will be conducted within 48 hours for patients with acute respiratory failure who are receiving pressure support ventilation (PSV).
2. Prolonged mechanical ventilation time (ventilation for ≥7 days).
Exclusion Criteria:
- 1. Cardiac and pulmonary instability. 2. Severe neurological deficits. 3. Bilateral phrenic nerve injury. 4. Previously admitted to the ICU with a life expectancy of less than three months.
5. Presence of contraindications for EIT examination.