Breathe Easy I: Calculation and Conservation of Mission Oxygen Volume

Evacuating or retrieving a patient from a remote environment with a high oxygen requirement can mean your supplies are rapidly consumed. Our saving grace is that because oxygen is a gas it can be stored under pressure, and because it behaves just like any other gas, it will obey the gas laws. This means we can store a large supply in a (relatively) small amount of space.

Part One: How much oxygen even is there in a cylinder?

Take a standard 'E' cylinder, which is the largest portable cylinder: it’s got a physical volume of just 4.7L, but is pressurised to 137 bar (13,700 kPa). Using Boyle’s Law (P1·V1 = P2·V2), where P1·V1 represents the initial conditions inside the cylinder and P2·V2 represents the final conditions outside the cylinder, we can quite easily work out what volume of oxygen we will get from the cylinder (V2). We know P1 (the pressure in the cylinder) we know V1 (the volume of the cylinder), and we know ambient air pressure is ~101 kPa (P2).

Substituting in these values for the whole 'E' cylinder: (13,700 kPa x 4.7L) / 101 kPa = 637L of available oxygen.

That sounds like a phenomenal amount until you realise that a single patient on 15 L/min will use it all in just 42 minutes. One study of ventilated patients with moderate COVID ARDS needed an average of 654L/h during interhospital transfer by air, and 11% of their cohort needed >900L/h. These oxygen requirements vastly exceeded expectations, even for a seasoned team of retrieval physicians [1].

In a mechanically ventilated patient, requirements are slightly harder to calculate. In broad terms, it's the fraction of your minute ventilation (MVe) that is O2 - so if you're delivering 400mL tidal volumes at 12 breaths per minute at an FiO2 of 0.5, that's roughly 2.4L/min O2 needing to be supplied. If only it were this simple, but in reality there are other factors that need to be taken into consideration, such as leak from the circuit.

If you're using a Hamilton T1 ventilator, base flow will be 3L/min for adults and children, and 4L/min for neonates [2]. Hamilton T1 ventilators also show oxygen consumption; if you click on 'System' and then 'Info' it will be displayed as a measured parameter.

High oxygen requirements can burn through supplies rapidly. Calculating the quantity needed for transporting patients is imperative for transporting patients requiring supplementary oxygen – whether mechanically ventilated or not. This assumes even greater significance when transporting patients long distances, especially when this occurs by air.

The website Oxygen Calculator has a handy online application that allows you to calculate how long your cylinder supply will last based on the flow rate and current cylinder pressure:

https://www.oxygencalculator.com/cylinder/cylinderduration

Part Two: How to compensate for altitude

The additional issue of hypobaric hypoxia needs to be considered when transporting patients by air. This is less relevant when transporting mechanically ventilated patients, as they are breathing via a closed circuit with guaranteed airway pressures. Even in the pressurised cabin of a fixed-wing aircraft the additional hypoxia may place further strain on oxygen supplies.

Consider a healthy patient at sea level breathing room air; their PAO2 can be roughly approximated by using the Alveolar Gas Equation:

Assuming a normal PCO2 of 40mmHg, an FiO2 of 0.21, and a sea-level barometric pressure of 713mmHg (this figure excludes the 47mmHg of water vapour pressure), PAO2 will equal:

But what happens once they are in the cabin of a plane pressurised to ~8,000 ft / 2,400m?

Looking at this, you can see that cabin pressure – while higher than sea level – still isn’t particularly high altitude in the scheme of things. At 8,000 ft, barometric pressure is 75kPa (560mmHg). When we remove water vapour pressure (47mmHg) from this - it means that the air we breathe has a pressure of 513mmHg. When we run the numbers to see what our PAO2 would be at this altitude:

The pressure of the oxygen reaching our alveoli is now 58% of what it was. Assuming no A-a gradient or changes to the oxyhaemoglobin dissociation curve - at a PAO2 of 58mmHg we can expect even a healthy individual to have an SpO2 of 90-91%.

Now - this is where things become clinically relevant - we can actually work out how much FiO2 we need to deliver to return this patient to a normal PAO2 (100mmHg) by simply rearranging that equation.

An FiO2 of 0.29 is required to offset the additional effects of hypobaric hypoxia. In a healthy individual, this doesn’t need offsetting – you can quite happily live with an SpO2 of 90%. But in a patient with a pre-existing O2 requirement, this is additional hypoxia that will increase that further. If you consult the table below, you can see that generating an FiO2 of 0.29 only requires 2L O2 via nasal cannulae – but remember – this 2L/min is an additional 120L/hour.

Part Three: Remember to include an oxygen reserve

Once the flow rate required to maintain an adequate SpO2 is known, and the estimated travel time is known, oxygen carriage requirements can be calculated. Within this, it's standard practice to include an additional 50% reserve – this is important in the event of the 3 deadly D's of air retrieval: delay, diversion, or deterioration.

Part Four: Putting it all together with some worked examples

I know that’s been a lot of mathematical equations in quick succession – but this is an important concept to grasp to avoid running short on oxygen part way through a retrieval. Using all the concepts above, here are two worked examples:

Patient 1: You are retrieving a patient who has been evacuated from a mining town in the central highlands of West Papua, and is now in a local hospital. From your preliminary phone calls with the treating centre – you have established that the patient is mechanically ventilated (settings SIMV, VT 500mL, RR 12) and has an FiO2 requirement of 0.4, which has been stable for several hours. Assuming a negligible circuit leak, you calculate that this equates to an O2 requirement of 2.17L/min. Pre-departure planning suggests a door-to-door transfer from the local hospital to the receiving facility in Malaysia is likely to be 8 hours. This means your patient will require 1,043L O2 with an additional 50% reserve, this brings the total mission oxygen volume to 1,565L.

Patient 2: is in the same situation but this time is not ventilated, and has an O2 requirement of 7L/min via non-rebreather, and this has been stable for several hours. Pre-departure planning suggests a door-to-door transfer from the local hospital to the receiving facility in Malaysia is likely to be 8 hours, including an estimated flight time of around 5 hours. This means your patient will require 3,360L O2 for the duration of the trip, and a further 600L O2 to achieve altitude equivalence. Combined this is 4000L, and with an additional 50% reserve, this brings the total mission oxygen volume to 6000L.

Many dedicated air ambulances will have piped oxygen integrated into the airframe. If using oxygen cylinders for the duration of the ground transfer between hospital and aircraft, consider bringing your own cylinders with you from your port of origin, because pin indexing on local cylinders may not be compatible with your regulators. These are often country specific, with many different mutually incompatible types – for example a regulator from Germany will not fit a cylinder from New Zealand and vice versa.

Part Five: In-flight oxygen conservation strategies

• If spontaneously ventilating, use the oxygen delivery that most efficiently improves SpO2. As you can see from the table above, different oxygen delivery devices will provide the same FiO2 at different flow rates. 7L/min could generate an FiO2 of 0.5 or 0.7 depending on whether it is delivered by simple face mask or non-rebreather mask. Consider dispensing with simple face masks and just utilising either nasal cannulae for low flow delivery or non-rebreather masks for higher flow rates.
  

• Aim for an SpO2 ~94% - there is increasing evidence that hyperoxia is not beneficial and may even be harmful in a range of pathologies (3). Normalising SpO2 to >98% is rarely necessary (with some important exceptions). Correcting SpO2 to a target of 94% will allow you to titrate FiO2 to the lowest necessary figure. Some masks will even do this for you! (4)
  

• Utilise PEEP – this is the other tool at your disposal for improving oxygenation. PEEP may improve recruitment of atelectatic lung units (and therefore reduce the burden of lung units with a low V/Q ratio and reduce the shunt fraction). Keep in mind that in excess it can act to compress capillaries (and therefore create units with a high V/Q ratio and increase physiologic dead space fraction).
  

• Utilise alternate means of supplying oxygen [5] - we will discuss supplementing your supplies with portable oxygen concentrators in a future article.

References:

[1] Beaussac M, Boutonnet M, Koch L, et al. Oxygen management during collective aeromedical evacuation of 36 COVID-19 patients with ARDS. Mil Med. 2020: doi:10.1093/milmed/usaa512

[2] Calculating oxygen consmption for Hamilton Medical ventilators. Hamilton Medical. 2020. Available from: https://www.hamilton-medical.com/en_US/Resource-center/Article-page~knowledge-base~c1b09f7f-3224-45b9-aa12-4cfd37e6d5ff~.html

[3] Chu DK, Kim LHY, Young PJ, et al. Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. Lancet. 2018;291(10131):1693-705.

[4] Hinkelbein J, Glaser E. Evaluation of two oxygen face masks with special regard to inspiratory oxygen fraction (FiO2) for emergency use in rescue helicopters. Air Med J. 2008;27(2):86-90.

[5] Nowadly CD, Portillo DJ, Davis ML, et al. The use of portable oxygen concentrators in low-resource settings: a systematic review. Prehosp Disaster Med. 2022:1-8.