COVID-19: How high-flow nasal oxygen is saving lives and sparing some patients the trauma of intubationPHOTO: Zdenko Zivkovic/Flickr (CC BY 2.0)

COVID-19: How high-flow nasal oxygen is saving lives and sparing some patients the trauma of intubation

News & Features

While the scientific understanding of the SARS-CoV-2 virus is constantly improving, so is our understanding of practical methods to treat this life-threatening disease.

Doctors at two of the Western Cape’s most impacted COVID-19 hospitals say they have seen positive outcomes from using high-flow nasal oxygen (HFNO), a non-invasive oxygen therapy, instead of mechanical ventilation. They say they are now using HFNO for many patients who would have been intubated in the early days of the epidemic.

Dr Saadiq Kariem, Chief of Operations for the Western Cape Health Department, said in a media briefing last week that the province had 121 HFNO machines available at hospitals, with another 42 on order, bringing the province’s total to 163.

“It’s made a real difference in Groote Schuur and Tygerberg and preparing for high-flow oxygen as an alternative to people having to be on ventilators,” says Kariem. “[HFNO] can be provided not only in a high-care bed, but we have also been providing [it] in a normal, acute bed. That relieves some pressure on our high-care beds, and that’s generally the strategy at all our hospitals and facilities.”

A letter published in the South African Medical Journal (SAMJ) in early May called for the wider use of HFNO to treat respiratory failure in COVID-19 patients. The letter, submitted by intensive care and infectious disease specialists from Tygerberg Hospital, cites clinical outcomes from the hospital’s ICU that could suggest a lower mortality rate for patients on HFNO, as opposed to ventilation.

Professor Coenie Koegelenberg co-manages the Tygerberg COVID-19 ICU and also co-authored the SAMJ letter. Speaking to Spotlight, he says that more recent data from the hospital’s first 70 ICU patients shows a mortality rate of roughly 40 percent, compared to roughly 85-90 percent for patients who did not receive HFNO and were mechanically ventilated.

“Internationally the outcome once you intubate is not good,” says Koegelenberg. “Once intubated most will not survive, and those that do survive take about three to four weeks to be liberated from the ventilator and to be discharged and that’s a very long time. It means that they occupy a bed for approximately a month, at least three weeks in ICU, whereas with HFNO [according to Tygerberg’s available data] it’s less than a week.”

“We have about a 60 percent survival rate and we are heading towards our 200th ICU admission, so there are many success stories, fortunately,” he says.

Along with Tygerberg, Cape Town’s other leading COVID-19 hospital Groote Schuur (GSH), has also been using HFNO as a first resort for patients with severe disease. Koegelenberg says that in early April, Tygerberg and Groote Schuur clinical staff were already in discussions on how to limit intubation by using this method.

Tygerberg administers HFNO in their ICU, while Groote Schuur provides it in medical wards, and only admits ventilated patients to ICU.

How it works

“It’s important to demystify [the HFNO machine],” says Professor Greg Calligaro, a specialist pulmonologist at GSH working on the frontlines.

He explains that the machine blends and humidifies a combination of oxygen and room air (normal air in the room) to give a certain percentage of breathable oxygen to the patient. This oxygen is flowing at a very high rate (about 60 litres per minute), and because of that, it has the ability to increase oxygenation by washing out what is called ‘dead space’ in the lungs, the parts that do not take part in oxygen exchange.

“The lungs are like bellows, the conducting airways fill up exhaled breath with a higher concentration of carbon dioxide in it so it’s going to be relatively stale air. HFNO fills up all that space with oxygen enriched gas, so basically it allows you to always be breathing very high concentrations of oxygen and not any of the air which would normally be diluted with carbon dioxide from exhaled breath,” says Calligaro.

“It’s a very comfortable form of oxygen administration, so a lot of people feel a lot better on it.”

‘Stiff lungs’ or Acute Respiratory Distress Syndrome

 For COVID-19 patients with severe disease, doctors find that oxygen levels in these patients are dangerously low, or in medical terms, patients are ‘hypoxic’, and classified as having what is called Acute Respiratory Distress Syndrome (ARDS).

“ARDS can be caused by many things [pneumonia being the most common cause], but essentially lung water builds up in between the air sacs of the lung and eventually fluid goes into [these air sacs, called alveoli] and then the lung becomes stiff, like a very stiff balloon and it’s very difficult to inflate,” says Koegelenberg. Stiff lungs like this are considered ‘non-compliant’, meaning that they cannot breathe without the help of pressure from a ventilator.

“Lungs are normally very compliant, it takes very little effort for [a healthy person] to breathe. ARDS is an extreme form where the lung is so stiff that just by blowing in oxygen it won’t work, you need to give it pressure as well and that’s where intubation will always be needed,” says Koegelenberg.

“When the disease first started, all the serious COVID-19 patients fulfilled technically [a clinical definition of ARDS], so if you were following pre-COVID medicine, you would intubate all of them. Initially globally, people were under the impression that this was just another ARDS and should be managed as such, but actually this is a very unique disease and many of the patients’ lungs are still compliant which is not what we usually see with ARDS,” he says.

Hypoxia and getting oxygen into the small sacs in our lungs

 Calligaro says that the major driver for respiratory distress in COVID-19 is the fact that the lung is not able to exchange oxygen, causing patients to become hypoxic. The amount of oxygen somebody gets into their lungs is determined by the amount that actually gets into healthy functioning alveoli and comes in contact with blood, he explains.

“Giving any form of oxygen non-invasively is complicated by the fact that whatever you give at the mouth is mixed with the exhaled breath, so it dilutes the amount of oxygen that you give. The amount that it gets diluted by depends on how fast you’re breathing and how deep your breaths are. You can see that once you’ve got somebody who is breathing really fast, they’re the kind of patient that dilutes the amount of oxygen you’re giving them the most.”

Calligaro explains that HFNO works by continuously flushing out the amount of oxygen that a patient breathes at the level of the alveoli. “That’s why you need the 60l flow, which is much more than what ventilation is in litres, so it’s washing out the lungs with this high concentration of oxygen,” he says.

“The patients we have treated [at GSH] all meet the criteria for severe [ARDS], so four months ago they all would have been intubated, without question. This is an extremely sick group of patients and a lot of them will still go on to be intubated, but [HFNO] is much more comfortable, it’s a very light-fitting [device] on the face, it’s not claustrophobic, it doesn’t require the cost and resources of ICU care, and it may result in at least 1 in 3 survivors.”

Proning and HFNO

Both Tygerberg and GSH have patients on HFNO prone themselves, meaning patients will rotate their bodies every two hours to allow oxygenated blood into different parts of the lung. For this, a patient will lie on their back, their left and right side, and on their stomach for two hours at a time. For ventilated patients, ICU staff are responsible for manually proning patients.

“Proning is a very effective way of pushing up oxygen levels,” says Koegelenberg.

Calligaro says that the understanding of how prone positioning improves oxygenation is complex, and not completely understood. “Blood flow through the lung is not uniform and nor is ventilation,” he says.

Changing a patient’s orientation, Calligaro says, is like turning a wet sponge over, which allows the more compressed parts at the bottom to be filled with air again, better matching blood flow to ventilation.

“The lungs [are] shaped like triangles, with the bases along your back, so if you’re able to lie on your stomach, it’s kind of like suspending this inverted pyramid, it stretches out the lung units and particularly the ones that were at the bottom are now at the top and can open up, and that’s where the best blood supply is, so the oxygenation increases,” he says.

While HFNO and proning have lent themselves to decreasing mortality rates, Calligaro says they are not a cure for COVID-19. Together, these two methods treat low oxygen levels in patients with severe COVID-19 disease, but do not actively treat the virus itself.

A paediatric perspective

Some are calling children bulletproof to COVID-19, and doctors and scientists are still uncertain why the vast majority of children are not developing severe disease. Professor Andrew Argent, Head of the Department of Paediatrics and Child Health at the Red Cross War Memorial Children’s Hospital (RCH), says that as of 12 June the Western Cape had very few children admitted to ICU with COVID-19.

Looking at data from 12 June, Argent said that persons under the age of 20 made up roughly 6 percent of the Western Cape’s overall cases, and hospital admissions were low. On that date, Argent said RCH had only 14 children admitted for COVID-19, while Tygerberg had 5, and of that total number, only 4 were admitted to ICU.

HFNO has been used to treat respiratory illness in children for years, says Argent, and while kids may not be needing it for COVID-19, they definitely need it for other illnesses.

“We’ve used the high-flow systems on thousands of children over the last few years largely for respiratory disease ranging from asthma to bronchiolitis to pneumonia and other conditions, so we’ve used it quite extensively. It’s not just used for COVID-19,” he adds.

Argent says that HFNO therapy is child-friendly and largely tolerated by these young patients.

“One of the beauties about HFNO is that it can be applied to the nose easily, and most patients seem to feel very comfortable with the system. But we will also start at lower flows and then turn it up, so you don’t put it on with massive flow which [can] feel weird, it’s just to put it on and then slowly turn up the flow as they get comfortable with it,” says Argent.

Challenges with delivering HFNO

With HFNO becoming a common form of treatment for severe COVID-19, Calligaro raises concerns that hospitals could be limited by their pre-existing piping infrastructure and the availability of oxygen supply – an issue that South Africa will undoubtedly face in the future.

“[HFNO] consumes a lot of oxygen so you have to be sure you know what your total oxygen capacity is in the hospital. I think most hospitals are not designed with the anticipation that every bed that can potentially give oxygen will be giving oxygen, I think that’s been one of the logistical challenges of the pandemic for many hospitals,” he says.

“The main limitation is that it requires very high oxygen flow rates, we have fortunately been able to deliver it in most of the wards we’ve wanted to at GSH, but you can’t really undertake this without some kind of assessment as to what the global supply is and what delivered flow at every bed head is – determined by the calibre of the pipes. It’s very difficult to retrofit any of that, so it depends on existing infrastructure.”

Calligaro says that there are two considerations for hospitals giving this treatment, how much oxygen they have and if there is sufficient piping to support multiple HFNO machines.

He also warns that hospital oxygen supply must also service theatre, maternity and other ICU wards and COVID-19 patients. “You’ve got to speak closely with engineers to ensure the amount of oxygen you’re providing through these machines is not going to jeopardise the [hospital’s] oxygen supply.”

SUBSCRIBE to our newsletter.