High Flow Oxygen Therapy


I have been a “fanboy” for high flow oxygen therapy (HFOT) for a couple of decades, particularly once high flow nasal canula (HFNC) became available. While this was a bit of a cottage industry, coveted by those of us in critical care (and to a lesser extent in anesthesiology), once the COVID 19 pandemic took hold, high flow was everywhere. And everyone, it seemed, had an (ill informed) opinion about this therapy. So, before I introduce this tutorial, about which I procastinated for years, I have to register a disclaimer: the evidence to support a lot of the “beliefs” about high flow oxygen is scant. Most of the claimed “benefits” beyond treating hypoxemia are industry generated hypotheses without rigorous scientific data. Nevertheless, this put me in a difficult predicament when constructing the tutorial – if I limit the discussion to just the facts that I am certain about – it would be very short. Conversely, by describing alternative “benefits” I take the risk of hyping hypotheses (e.g. CO2 clearance) that may be incorrect…..

High Flow Oxygen Therapy (HFOT), particularly when delivered by nasal cannula (HFNC) has revolutionized the management of the patient with hypoxic respiratory failure – in particular in those patients whose lung pathology has plateaued or those resposive to medical treatment (antibiotics, steroids etc). High flow systems have been available for decades – they involve the use of a high pressure oxygen source, and oxygen air blender (air can be entrained into this device), a high flow flowmeter, a humidifier, a heated delivery tube and a delivery device: CPAP mask, T-Piece with PEEP valve, Tracheostomy or specially designed nasal cannula.

In this tutorial I describe the various devices configurations that are available – ranging from very straightforward standalone machines, to full mechanical ventilators. Regardless of device the major goal is to deliver sufficient flow to meet patient demand – resolving the problem of peak flow and separating out the FiO2 from the flow rate. I postulate that, at flows in excess of 30L per minute, and depending on the diameter of the nasal cannula, the patient’s anatomy and whether the mouth is open (and by how much!) – the patient likely receives a couple of cmH2O of pressure support and 3-5cmH2O of PEEP. So it represents mild CPAP (certainly a CPAP device delivering high flow at 5cmH2O will outperform HFNC). There is a dearth of non industry funded data on how HFOT may benefit the patient. Certainly these devices are very effective at targeting SpO2 and reducing the work of breathing. Certainly they increase non hypoxic apneic duration. Conversely – purported impacts on dead space washout, alveolar ventilation and CO2 clearance are currently unproven. I describe how this may work in the tutorial, but point out that this is principally a belief not a fact. HFNO may also improve mucociliary clearance – due to the high flow of humidified gas passing into the airways. However no-one, to my knowledge, has addressed whether constant flow of heated humidified gas for prolonged periods damages the lung mucosa.

In the second part of the tutorial I talk about how HFOT should be used in clinical practice and the scenarios in which it is beneficial (hypoxemia, weaning and liberation) and when it is not (hypercarbic respiratory failure, post op respiratory failure secondary to atelectasis).

I guarantee that you will learn something.


Oxygen Therapy: Variable vs Fixed Performance Devices

Oxygen is probably the most used and misused drug in a hospital. The purpose of oxygen therapy is to restore the PaO2 or SpO2 to a safe level for that patient. One of the major issues with targeted oxygen therapy is the problem of peak inspiratory flow.

During peak inspiration the FiO2 must be constant. That means that flow delivery must meet flow demand. Oxygen therapy can be delivered with variable or fixed performance devices. Variable performance devices include nasal cannula and simple (“Hudson”) facemasks. In both cases oxygen and air are blended in or near the airway. Nasal cannula are remarkably efficient and can deliver low inspired oxygen concentrations. Due to issues with dead space and rebreathing, simple facemasks are unreliable below 35% (5L). Both devices struggle where there is rapid breathing, particularly with large tidal volumes.

Venturi devices, which are really jets use a narrow injection port to entrain and blend oxygen and air proximal to the facemask. They are more precise but less efficient (in terms of total flow) than variable performance devices. Performance is remarkably robust between 24% and 40% inspired oxygen. They perform less well with rapid deep breathing particularly at high FiO2 levels. Non rebreather facemasks use a reservoir to store fresh gas during expiration and facilitate the delivery of FiO2 of approximately 80% with 10 to 15 liters of flow. As such they are highly efficient, although unreliable and non titratable. These devices can be used with modest oxygen flows for transporting hypoxic patients, but are short term remedies. @ccmtutorials http://www.ccmtutorials.org