Patients presenting to the emergency department (ED) may require breathing support with machines depending on the condition. Throughout the breathing support, the settings on the breathing machines will be tailored to the patient's requirements. These settings are manually adjusted by trained physicians. Currently, there are machines which can automatically change the settings based on real-time specific information obtained from the patient. This study aims to compare the use of machines which require manual adjustments (open-loop conventional ventilators) and machines which can automatically change the settings (closed-loop automated ventilators). Patients will be carefully selected to ensure no harm is caused whilst delivering the best care. This study will look into the duration when patients are receiving optimum settings and levels of oxygen and carbon dioxide in the blood. The outcomes of this study would allow us to identify methods to improve patient care.

Invasive mechanical ventilation is a lifesaving intervention for patients with respiratory failure in the emergency department (ED). Recent technological advancements have introduced closed-loop automated ventilators as a potential alternative to open-loop conventional ventilators. However, the efficacy and safety of closed-loop automated ventilators in the emergency setting remains understudied. This research aims to evaluate the efficacy and safety of closed-loop automated ventilator compared to open-loop conventional ventilator in intubated and ventilated patients in the ED.

A randomized controlled trial will be conducted in an ED of a tertiary university-affiliated hospital. Eligible patients are 18 years or older, decision made by treating physicians to intubate and mechanically ventilate. Some of exclusion criteria are pregnancy, heart failure, metabolic acidosis, circulatory shock, life-threatening asthma and morbid obesity. The primary measure of efficacy is the duration of ventilation within a predefined range of acceptable respiratory parameters between automated and conventional ventilation. Secondary outcome measures are; number of manual adjustments required to attain targeted settings in automated and conventional ventilators, PaO2/FiO2 ratio (PF ratio), arterial blood gas results, vital signs, breath-by-breath analysis, and rate of ventilator dyssynchrony. The ventilator used in the intervention arm is the closed-loop automated ventilator Hamilton C6s INTELLIVENT-ASV (Hamilton Medical AG, Switzerland). Hamilton C1 ASV is chosen as the open-loop conventional comparator as it is similar to Hamilton C6s INTELLIVENT-ASV, without the INTELLIVENT software.

Based on Lellouche et al, the calculated total sample size with a dropout rate of 10% is 132. The data is analysed based on the intention-to-treat (ITT) and per-protocol (PP) principles. The primary endpoint measurements are reported as areas under the curves (AUC) within the predefined range of acceptable respiratory parameters. Between-group differences in continuous variables are analysed using independent t-test or Mann-Whitney U test. Between-group differences in categorical variables are analysed using chi-square test.

Inclusion Criteria:

• Decision made by treating physicians to intubate and mechanically ventilate

Exclusion Criteria:

1. Suspected or confirmed pregnancy.

2. Known right ventricular heart failure upon assessment for recruitment.

3. Severe metabolic acidosis upon intubation (pH <7.2 or bicarbonate <12 mmol/L)

4. Circulatory shock requiring noradrenaline more than 0.5 mcg/kg/min upon assessment for recruitment.

5. Severe or acute life-threatening asthma.

6. Patients with chest wall deformities that would affect ventilation (e.g. severe kyphoscoliosis, diaphragmatic hernia, flail chest, trauma, pectus excavatum or carinatum, ankylosing spondylitis

7. Patients with previous lobectomy or pneumonectomy.

8. Patients with pneumothorax or other condition that requires chest drainage tube.

9. Patients with body mass index > 40 kg/m2.

10. Manufacturer's contraindications:

    • Difference in oxygen saturation between pulse oximetry (SpO2) and arterial sample (SaO2) of more than 5% (due to unreliable sensor).
    • Difference in carbon dioxide level between end-tidal sensor (ETCO2) and arterial sample (PaCO2) of more than 5 mm Hg (due to unreliable sensor).
    • Known pneumothorax and bronchopulmonary fistula upon assessment for recruitment.

11. Participation in another interventional trial.

12. Do-not-attempt-resuscitation (DNAR) order.

Individual participant data for this study will not be made publicly available. However, summary results and study outcomes will be transparently reported in the form of aggregated data and CSV files. This approach ensures that key findings are accessible while respecting participant confidentiality.