Optimising in-hospital oxygen utilisation in South Africa’s 2nd wave – learnings from one hospital

At the height of South Africa’s second wave in early January 2021, a hospital in the hard-hit Western Cape was using almost 3 tons of oxygen per day. This was a 10-fold increase in oxygen utilisation compared with ‘normal’ pre-Covid oxygen utilisation for normal operations, normal occupancies and with a normal incidence of patients with pneumonia needing oxygen therapy. The hospital was running above their tested limit and the fear was not running out of oxygen (hospitals have massive oxygen tanks that get re-filled as needed – with no supply concerns for these), but rather what would happen if the system’s structural capability was exceeded – resulting in a drop in the pressure of piped oxygen, or worse-case scenario – oxygen supply cut off to the entire hospital.

The hospital did not have a shortage of piped oxygen (if structural capabilities were not exceeded), but there was a genuine shortage of small portable oxygen cylinders. It had become almost impossible to re-fill small portable oxygen tanks because of the high demand; so, in addition to ensuring piped oxygen was not being wasted, the hospital also needed to secure the portable oxygen cylinders that they still had available.

Oxygen utilisation was measured on an hourly basis and when the trend showed a sustained increase above the tested limit, the hospital management decided to act. The team’s concern was not unfounded. A few days after prioritising a reduction in inappropriate oxygen utilisation, this tragic incident happened at a hospital in Durban – an oxygen supply interruption seemingly resulting in the death of six patients.

Magic is created when a cross-functional team comes together to solve a problem. This melting pot of individual and occupation-specific expertise, observations and contextual wisdom is truly a thing of beauty. Leads from the technical department, anaesthesiology, internal medicine, nursing, pharmacy, and training and the management team sat around one table discussing the problem, learning about its implications, analysing the contributing factors and brainstorming potential ideas.

Contexts differ – and I am not for one minute suggesting that other hospitals will have the same contributing factors, but these were some of the main contributing factors at this hospital:

• The patients with Covid-19 pneumonia had extremely low oxygen levels (it was not uncommon to see patients with oxygen saturations of 85% in the general wards).

• ICU demand had far exceeded capacity and all patients were triaged for ICU beds and for advanced oxygen therapy (including high-flow nasal oxygen and intubation/ventilation). Many patients that would have previously been admitted to ICU were unable to get a bed in ICU – these critically ill patients were in the general wards.

• Doctors and nurses were desperate to increase these patients’ oxygen levels – and ‘double therapy’ had crept into practice. It was not uncommon to see patients on a combination of nasal cannula and a rebreather mask – nasal cannula running at 15 litres per minute (LPM) and the rebreather mask running at 25LPM. This was not a practice seen pre-Covid. This practice was seemingly driven by desperation and the ‘belief’ that this ‘double therapy’ would deliver a form of high-flow oxygen.

• A bed in a general ward only has one oxygen flowmeter. To deliver ‘double therapy’, portable/mobile oxygen cylinders were being used.

The changes that had the biggest impact were the following:

• Creating awareness of the problem and its impact

• Developing, implementing, and iterating a standardised oxygen protocol for escalating oxygen use (prescribed by all disciplines)

• Addressing knowledge deficits with regards to appropriate oxygen therapy for the different oxygen delivery systems (i.e., flow requirements for nasal cannula, venturi masks, partial re-breathing masks and rebreathing masks).

• Providing guidance on rationalising the use of small oxygen cylinders and limiting their use to patient transport only.

I remain in awe of the energy created when a group comes together to address a problem worth addressing – and the willingness of people to go above and beyond to make a difference for the better. Protocols were quickly proposed and iterated as they were implemented, and training for nursing teams was conducted at night and early in the morning to reach as many nurses as possible.

Within 48hrs, there was a 30% reduction in piped oxygen utilisation – down to a more acceptable and ‘safer’ 2 tons a day; and the hospital did not run out of portable oxygen cylinders!

My message in essence is to use the wisdom that exists in your system right now; come together as a cross-functional team and analyse your situation and its contributing factors; and then try different ideas based on this analysis. It is not often that ideas result in immediate improvement – sometimes many different ideas need to be tried out. But the likelihood of coming up with changes that are an improvement starts with harnessing the wisdom in the system.