Air Collective Forum

Establishing timelines for shortages in our countries

Knowing the timeline and degree of the ventilator shortage in your country is going to be very useful in deciding what type of device will help best.

For example, if the leadtime to the shortage is short, then simpler devices with lower production time are more urgent. If strong lockdowns or other mitigation countermeasures will need to be in place for a long time, focusing on treatment devices that help patients recover before developing ARDS, like CPAP and sleep apnea machines, might be more appropriate.

Here is an overview by country, based on the Neherlab model:

** Keep in mind the projections are changing a lot these days, since this week is around 2 weeks after countries have gone into lockdown, so the trajectories are changing. Double check these yourself to get the latest.**

Spain is already beyond its ICU capacity. The overflow period is projected until the end of July, and the peak overflow is estimated between 80,000 and 200,000 ICUs.

Bulgaria is estimated to start ICU shortage between end-of-May 2020 and early-Febuary 2021. (It seems to have been caught early, so the large range is due to the effectiveness of long-term mitigation.) The peak overflow will be between 17,000 ICUs (if earlier) and 4,000 (if later.) The overflow period will last 5-6 months.

The Netherlands has reached ICU overflow, which is projected to last until July. It is possible another overflow period returns bwteen February 2021 and July 2021. The peak overflow will be between 58,000 ICUS and 8,000 ICUS.

The UK is projected to hit ICU capacity by mid to late April, With strongly effective mitigation, The first overflow will be only 2 months and 4,000 ICUs overcapacity, but could return from December to June 2021, which would be a more severe shortage of 37,000 ICUs. If the mitigation is weak, one overflow period of is projected. It will last until August and there would be a 200,000 ICU shortage at peak.

For other countries, and more specific areas, you can run the Neherlab projections yourself with the instructions below.


If you know any experts on this, please reach out to them and ask to give us guidance here. I am just sharing from the sources I’ve found to be reputable, but I’m not a statistician.

There are two sources I’ve found helpful, reputable and level-headed when it comes to the statistical models:

  1. Our World In Data, which focuses on explaining the important factors in determining the rate of spread. I recommend this for a foundational understanding of the projections - why the infection rates are predicted to grow like they are.
  2. The University Of Basel’s Neherlab, who have created an interactive model for seeing ICU shortages in different scenarios.

(If you know of any other sources, or your national ministries are able to release their projections, please look at those too, and share them here. Until then, this is the best way I know for us to inform ourselves to make better decisions.)

Get a sense of how long you have until the shortage, how long it’ll last, and how big the shortage is

The Neherlab interactive model gives us a clue about the time factors that affect what types of device we choose:

  • lead time: time left until there are respirator shortages
  • duration: how long it will last
  • degree: how many extra ICU units (or ventilators) will be needed at the peak

If the degree is low, for example, devices like ventilator multipliers may be adequate.

The lead time might help you decide if you want to focus on short-term wins. For example, with a short lead-time and a long duration, you might choose to focus on simpler ambubag devices, or producing smaller volumes quickly, and count on industrialized responses to succeed later in the the overflow period.

Knowing the timeline and degree of the ventilator shortage in your country s going to be very useful in deciding what type of device will help best.

Here we can see the projections for Holland. There is a very short time left until there is an ICU shortage. If their mitigation efforts have a weak effect, the overflow is projected to last until July. The peak overflow would be around 17,000 needed ICU beds at the beginning of May.

Getting oriented with the interactive model

  • When you load the Neherlab model, first choose your area on the left where it says “population”, then below that you’ll see “simulation time range”. Make that from today to September 2021. (Unfortunately it extends that long in many cases.)

  • Then press “run” on the right, and wait a few seconds for the results. (You have to press run every time you change parameters.)

  • The red line will show you predicted ICU cases over time. If that line has a flat part, it shows when you’ll run out of ICU beds. There you’ll see a purple line which shows the amount of ICU overflow patients at that time. That’s the line we’re all working towards reducing.

  • The little bubbles on the left, those are actual numbers, not the mathematical projection. So you can see how the model compares to the current situation.

  • When you notice a difference, consider that confirmed cases are not actual cases, and that there are also time delays in reporting.

  • Scroll down to see the final number of cumulative deaths, and how they are distributed by age.

Look at a range of scenarios

This is not a predicable situation, and many factors will change over time. For example, whether or not lockdowns work, or how long the economy can survive on lockdown. So, I suggest this tool is best used to look at a few different scenarios. Then, you’ll have an idea of the range of possibilities.

There are two variables I am changing to understand different scenarios:

  1. The mitigation factor (on the bottom left, which is the rate at which the responses, such as lockdowns or finding and isolating the infectious, is able to slow transmission rates.

I am looking at a range of mitigation factors, because I think there are many (unpredictable and volatile) factors that will make strong mitigation difficult for more than a few months. I look at strong and weak mitigation.

  1. The second is what extra ICU capacity can do (that’s on the left, above time range.) If Elon Musk just delivered respirators, or you see hospitals using ventilator multipliers, factor that in by adding those numbers to the ICU capacity and re-run the projection.

On the positive side, you can see what extra ICU capacity can do to save lives. Again, consider cases where ventilator multipliers may increase capacity. This is more complicated than just saying 4X, since ICUs require a lot more equipment and people than just more ventilators, but you will get some idea.

Keep in mind that the default parameters of this model are unlikely to pan out, and that there are also other factors like staffing issues that will cause limits. But as an engineering community we can do what we can so that number of available breathing support devices are not the restricting factor.

In Bulgaria, strong mitigation will push the overflow until next year, whereas with weak mitigation, the overflow problem will start sooner. This raises the question of what happens if the mitigation weakens over time, or if Bulgaria’s economy can withstand the mitigation for so long. The further out we look, the more unpredicable this gets. There maybe medical advances that help with mitigation, or technology projects that help find and isolate at-risk people. But at least, now you have a sense of the timeline.

What did you decide? What factors did you consider?

1 Like