Background:

Optimal adjustment of mechanical ventilation in intensive care patients is crucial to ensure effective ventilation and to reduce the risk of ventilator-associated complications such as pneumonia and ventilator-induced lung injury. Current monitoring methods may often be insufficient to achieve truly optimal ventilator settings in routine clinical practice. By applying a specific breathing pattern, additional key ventilation-related parameters can be monitored and adjusted using a novel carbon dioxide-based method. This approach has the potential to substantially improve ventilation in critically ill patients.

The carbon dioxide-based method utilizes a modified intensive care ventilator (Servo-i®, Getinge) equipped with CE-marked research software. The software modifies the breathing pattern by introducing three slightly prolonged expiratory pauses in three out of nine breaths. This ventilator has been used in multiple large-animal studies and clinical trials, including two conducted by the present research group.

Study Design:

The study will include adult patients receiving mechanical ventilation in a neurointensive care unit due to brain injury. Patients must be in a stable phase with regard to neurological status and circulation and require controlled mechanical ventilation. Informed consent will be obtained from the patient after recovery when applicable.

Part 1 (Pilot Study):

This is an observational study including 15 patients, aiming to establish reference values for parameters obtained using the carbon dioxide-based method, such as end-expiratory lung volume and mixed venous oxygen saturation. Patients will be switched to the research ventilator for 15-20 minutes, during which measurements from the carbon dioxide-based method and standard vital sign monitoring will be collected. Ventilator settings will be identical to those used on the patient's conventional ventilator. After completion of the protocol, patients will be switched back to their standard ventilator.

Part 2 (Main Study):

This is an interventional study with a within-subject (self-controlled) design, including 15 mechanically ventilated patients with stable brain injury. The aim is to optimize ventilator settings using the carbon dioxide-based method to achieve more lung-protective ventilation.

Patients will be transferred to the research ventilator using the same procedure as in the pilot study, initially maintaining the same ventilator settings. After a 10-minute stabilization period, ventilatory parameters will be recorded and an arterial blood gas sample obtained (baseline). The specific breathing pattern will then be initiated, and ventilation will be adjusted based on additional parameters provided by the capnodynamic method (end-expiratory lung volume, cardiac output, and mixed venous oxygen saturation). Measurements and arterial blood gases will be repeated immediately after initiation of the breathing pattern and again 15 minutes after ventilator adjustments. The study will then conclude, and the patient will be returned to their standard ventilator using the optimized ventilator settings to potentially ensure patient benefit.

The primary outcome measure is a reduction in driving pressure, as it is well established that a driving pressure >15 cmH₂O is associated with an increased risk of lung injury (ventilator-induced lung injury, VILI). Clinical observations in the intended study population show driving pressures of approximately 19 cmH₂O with an estimated standard deviation of 2 cmH₂O (study baseline). When these values are entered into a power calculation for a paired two-sided t-test, a sample size of approximately 10 patients is obtained. To account for potential dropouts, the study group size is set at 15 patients. The significance level is <0.05, with a power of 80%.

Inclusion Criteria:

• Stable with regard to intracranial pressure
• Hemodynamically stable

Exclusion Criteria:

• Ongoing instability with regard to intracranial pressure and circulation
• Spontaneous breathing / patient-triggered ventilation
• If the patient subsequently declines consent