The investigators propose to conduct a feasibility, multi-centre, randomised controlled trial of targeted oxygen therapy in adult critically ill patients receiving mechanical ventilation via an endotracheal tube as part of their treatment for respiratory failure. Participants will be allocated to either a normal blood oxygen target group or a lower than normal blood oxygen target group. The primary purpose of the study will be to assess the feasibility of recruiting complex patients who lack capacity into a clinical trial in which oxygenation is being assessed, and that the clinicians responsible for these patients are able to deliver the intervention effectively. The safety of using a lower than normal blood oxygen target will also be assessed and blood samples taken for subsequent investigation of the biological mechanisms underlying the observed changes.

Participants will be randomised (1:1) into either an intervention or control group. The intervention in this trial is tightly controlled administration of oxygen to patients to achieve a haemoglobin oxygen saturation (SpO2) of 88-92%. The control group will also have tightly controlled oxygen administration, but to achieve an SpO2 of 96% or above. The target for the control group represents a normal SpO2, whilst that in the intervention group is lower than what is considered to be normal. It should be noted that although lower than normal, this SpO2 is close to what the general public experience when travelling by pressurised aircraft as the fractional inspired oxygen concentration in that situation is only 0.15-0.17 (15-17%).

The controlled oxygen administration would commence as soon as possible after admission to the critical care unit and end following removal of the participant's artificial breathing tube. The researchers and clinical team cannot be blinded to treatment allocation, due to the nature of the intervention. Those analysing the data will be blinded to the intervention.

The investigators aim is to determine whether revising standard targets for blood oxygen levels in patients requiring artificial ventilation on a critical care unit is feasible and whether this affects specific blood biomarker levels.

Investigators in this area of expertise currently lack the information necessary to determine how much oxygen should be given to adult critically ill patients on a mechanical ventilator to assist their breathing in order to achieve the best clinical outcomes i.e. minimal morbidity and mortality. Excessive oxygen administration (both its concentration and duration) and hyperoxaemia (an excessively high blood oxygen level) are known to be harmful to critically ill patients. The reason for this is that oxygen in high concentration is toxic, because it results in the release of molecules called reactive oxygen species (ROS). ROS lead to a state known as 'oxidative stress', in which cells and tissues are rapidly destroyed by these destructive molecules.

The difficulty with treating critically ill patients is that they are usually extremely unwell, requiring artificial ventilation to support their breathing. Unfortunately, these patients appear to be exceptionally prone to the harms of oxidative stress, leading to permanent damage of their lungs. Yet it is precisely this group of patients who require high concentration oxygen to keep their blood oxygen levels within the normal. It has therefore been hypothesised that using lower blood oxygen targets than usual may be beneficial to these patients through the avoidance of excessively high concentrations of oxygen. The investigator proposes that in mechanically ventilated critically ill patients harm (morbidity and mortality) will be reduced by using 'targeted oxygen therapy' (TO2T) to achieve lower blood oxygen levels than normal when compared to standard practice (normal blood oxygen levels). The levels are only moderately lower than normal and well within what most clinicians would consider to be safe in these patients.

Prior to any large-scale study, it is important to understand if this approach to managing critically ill patients is feasible. This prospective, dual site randomised controlled trial will therefore establish whether it is possible to conduct a trial in which blood oxygen levels are carefully titrated by the critical care team, to achieve specific targets. Participants will be allocated into one of two groups: i) normal blood oxygen levels, and ii) low blood oxygen levels. For the time that participants are artificially ventilated by a breathing (endotracheal) tube, they will remain in their allocated treatment group. Information will be collected from the patient's charts and records during their stay on the CCU to assess the impact of the treatment. A series of blood samples will also be collected from participants to quantify the degree of oxidative stress through the measurement of several specific biomarkers. A total of 60 patients will be enrolled at two sites.

As oxygen is a drug that is administered to almost every patient admitted to a critical care unit (CCU), it is imperative that decisions regarding oxygenation are based upon evidence rather than conjecture. Limited work has been undertaken to date to understand whether lowering blood oxygen levels is achievable in this complex patient group. The information from this study will be used to design a subsequent much larger study to fully evaluate whether TO2T to achieve lower blood oxygen levels saves lives in critically ill patients. If lower blood oxygen levels improve survival in critically ill patients implementation of this intervention could have an immediate and cost-effective impact across the entire National Health Service.

Inclusion Criteria:

• Unplanned admission to a critical care unit
• 18 years of age and above (no upper age limit)
• Respiratory failure forms part of the admission diagnosis
• The patient is mechanically ventilated via an endotracheal tube
• The patient is expected to receive mechanical ventilation for > 24 hours

Exclusion Criteria:

• Admission following surgery (elective or unplanned)

• Those patients expected to die within 24 hours of admission to ICU *

• Pregnant females

• Admission post-cardiac arrest

• Admission post trauma (including traumatic brain injury)

• Known sickle cell trait or disease

• Ongoing significant haemorrhage or profound anaemia

• Severe peripheral vascular disease

• Severe pulmonary hypertension

• Other medical conditions where mild hypoxaemia would be contra-indicated ***

• Patients participating in other interventional clinical trials

  • As determined by the responsible clinical team ** As determined by the responsible clinical team and /or research team

Clinicians and other healthcare providers will not be blinded to treatment allocation, nor will the clinical research team. This is because knowing the patient's blood oxygen level is an essential component of the study. Those members of the research team analysing data after the recruitment phase of the study will be blinded to group allocation. Participants will not be blinded but are unlikely to be aware to their allocation due to the severity of their illness.