High flow nasal oxygen therapy has been widely used but guidelines as to the optimal starting flow rate and oxygen percentage are not available. Prolonged exposure to an inappropriately high oxygen concentration should be avoided as there is increasing evidence that the production of oxygen free radicals can lead to lung damage. This pilot dose finding study will determine the optimal flow rate and oxygen concentration using HFNO2 for patients with respiratory failure requiring low, medium or high oxygen concentration from conventional low flow devices. An assessment will also be made of comfort and compliance with HFNO2.

The study will recruit a total of between 180 and 270 patients across three hospital sites mainly from high dependence units.

The patients recruited will be stratified into 3 groups according to their oxygen (O2) requirements at recruitment (low, medium and high).

All recruited patients will then be given supplementary O2 using the Optiflow system. Patients from all three recruitment oxygen requirements (low, medium and high) will be randomised to 30 litres, 45 litres or 60 litres per minute with initial a fractional inspired O2 concentration(FiO2) of 0.90 (90% oxygen). The FiO2 is then decreased in 0.05 increments every 5 minutes and the minimum oxygen saturation recorded until it reaches 94%.

End points:

Minimal FiO2 required to maintain saturation above 94% for patients with low, medium and high recruitment oxygen requirements for 30 litres, 45 litres and 60 litres per minute. Use Kruskal Wallis to compare the median FiO2 requirements of those patients who recruited with low, medium and high O2 requirement.

Obtain dose response curve 02 saturation against FiO2 for 3 different flow rates for each of the recruitment 02 requirements and use regression modelling to find best fit dose FiO2 response O2 saturation curves

Patient satisfaction scores and number of patients unable to tolerate the three flow rates in each of the three starting groups and flow rates.

Inclusion Criteria:

Patient aged 18-80 at time of recruitment to study) Able to give their own consent (4 "A" Test score of 0 if capacity questionable) Acute respiratory failure but no significant deterioration in the previous hour Receiving conventional oxygen therapy (Hudson, Venturi, non-rebreathing mask or equivalent) up to a maximum of 90% Oxygen saturation ≥ 94% and respiratory rate ≤30

Exclusion Criteria:

Type II respiratory failure (chloride ≤95 and bicarbonate ≥35, on an arterial blood gas or venous sample within 4 weeks) Chronic respiratory disease limiting functional capacity (MRC breathlessness grade IV or V) Pregnancy Heart failure (NYHA Grade III or IV) Decreased GCS Cardiovascular instability (systolic BP ≤90 or heart rate ≥130) Pulmonary embolism Unable to give informed consent Contraindication to receiving high flow nasal oxygen

• Nasal obstruction
• Previous bleomycin administration
• Base of skull fracture

Data not relevant to individual patient after episode of respiratory failure