Why the U.S. is out of ventilators (and why it might not matter)

In “State-by-state model of hospital bed and ICU demand” I wrote:

The biggest question mark for me is their forecast of ventilator usage (19,000 best-estimate for peak; 40,000 worst-case). If the data nerds are right, we have way more ventilators in the U.S. than we need to handle this challenge. We had at least 62,000 ventilators in service back in 2009 (source), plus another 100,000 older models in storage. Presumably a lot of the in-service ones are tied up with people who are sick with chronic conditions, but all of the older models should be available and that is a more than adequate resource. If these folks are right, everyone who is freaking out about ventilator supply is wrong.

A friend sent me a link from a critical care nurse that explains why the U.S. might well run out of ventilators. From April 1:

Yes our threshhold to intubate and ventilate is much lower than normal. We now have an abundance of patients (that turn out to be negative a day or two later) that are now on a ventilator and occupying an icu bed who would have probably been fine with a less invasive respiratory intervention.

These patients usually have a boatload of comorbidities and severe existing lung disease already so it isnt a snap to just say “oh youre negative so lets just pull the tube, send you to the floor, and youll be fine”.

Once intubated it can be a long road to extubate with these patients if we even can at all. Under normal circumstances we try everything else first with them before tubing them because we know that once tubed there will be very little chance that many will be able to be extubated without going the trach, PEG, rehab route.

This is why the rapid test will be a Godsend once widely available. We can find out who is positive and needs a tube vs just a routine COPD exacerbation that can get BiPap and not risk infecting staff. Although the sensitivity and specificity would have to be very high on the rapid test or staff would still be at risk.

If so it could free up many ICU beds that are currently being occupied by covid negative patients.

In other words, if you want a near-guarantee that no nurse or physician will get infected you have to intubate and ventilate anyone who might conceivably have COVID-19.

Some background from the same guy, March 31:

Updates at my 200 bed community hospital in NH:

Our ICU is 90% full right now. They are tubing anyone that has a high likelyhood of having the ‘rona if they go into rapid resp failure since using BiPap or Hiflo vapotherm is an exposure risk to staff and will aerosolize the virus and blow it all over the room. Out of these patients, only one has tested positive. The others have all come back negative. The positive is a vented 70+ yo lady with comorbidities who is also fighting sepsis from a bowel perforation unrelated to the virus. We have her on hydroxychloroquine and every big gun IV abx we have at our disposal including azithromycin. For the last three days she has been holding her own and her vital signs have remained stable on low-moderate pressor support.

All of our rooms have now been converted to Neg pressure rooms in the ICU as well as many rooms on the med-surg and tele floors.

As of yesterday, hospital wide, we only had 4 positive inpatients including my patient in the ICU. Only one staff member has come up positive as far as I know (our hospital is rather quiet in reporting staff positives and this info has to be come by through the grapevine). We now have to fill out a questionnaire before work attesting to no covid19 symptoms in order to get a cleared for work pass. All staff members, from contractors, to admin, to clinical must wear surgical masks at all times when in the hospital, unless treating a covid or suspected covid patient. For them, N95 or Papr must be worn with protective gown, headgear, faceshield, and double gloves.

No visitors of any kind or for any reason are allowed in the hospital and have not been for the last 2 weeks.

Its been busy, and we only have a handful of patients so far. Im looking at at least 80 hours this week. While its nice to have job security I could do without the extra dash of impending doom sitting in the pit of my stomach every day I get dressed for work.

From March 21:

In regards to the N95 shortage, I dont blame anyone with the foresight to have prepared for this fiasco and purchased them. I certainly did and had bought a several N95s in the size Ive been fitted for (along with additional food and necessity preps) once it was obvious this was more than the 2003 SARS nothingburger.

I was dumbfounded however upon learning that my hospital was completely unprepared for this 2 weeks ago when they told us they had only 40 disposable N95 masks on hand and 10 papr helmets. For the entire fu**ing hospital! There was a near mutiny in our ICU for a few days once this word got out.

I lay all the blame at the feet of our infection prevention department and their incompetent director, who amazingly still has her 6 figure salary and who seems to do little more than clog up our inboxes with reminders about handwashing and take photo-ops in front of the triage tent outside the ED.

What if God sends a ventilator and critical care team to each of the 330 million residents of the U.S. (including the 22 million undocumented, of course! According to our local church, God loves the undocumented more than anyone else.)? How much would that cut the death rate from COVID-19?

“Respiratory support for patients with COVID-19 infection” (The Lancet):

The ICU mortality rate among those who required non-invasive ventilation was 23 (79%) of 29 and among those who required invasive mechanical ventilation was 19 (86%) of 22.

Maybe 15 percent then? My physician friends say ventilation implies a 90 percent risk of death, so the ventilator will delay death and result in a 10 percent reduction in the near-term death rate. The only doc who is optimistic about the ICU and ventilators is a guy who… works in the ICU running ventilators (“critical care”). But he is unwilling to quantify the advantages conferred on patients other than to say that it may help some.

“Mortality rate of COVID-19 patients on ventilators” (Physician’s Weekly):

Probably the best published information we have so far is from the Intensive Care National Audit and Research Center (ICNARC) in the UK. Of 165 patients admitted to ICUs, 79 (48%) died. Of the 98 patients who received advanced respiratory support—defined as invasive ventilation, BPAP or CPAP via endotracheal tube, or tracheostomy, or extracorporeal respiratory support—66% died.

An article in The Guardian said this about the ICNARC study, “The high death rate raises questions about how effective critical care will be in saving the lives of people struck down by the disease.”

A medical school professor, when I asked him whether it was true that the ventilators we read about in the media every day seldom make the difference between life and death: “It’s true, but don’t let facts interfere with my panic.”

So… the epidemiologists currently trying to act as prophets are forecasting ventilator use on the assumption that only patients who need ventilators will get one. However, COVID has caused our health care system to use ventilators at a prodigious rate, just as the system is using masks and other PPE at a prodigious rate.


  • “Texas ‘mom and pop’ business flooded with orders for helmet ventilators amid coronavirus crisis” (NBC), an inexpensive device that might work better than intubation and full-scale ventilation
  • while journalists in the NYT imply that you’ll go on the ventilator, watch some cartoons, then go off and hit Starbucks and the gym, the same newspaper got a physician to write “What You Should Know Before You Need a Ventilator”: [COVID-19] causes a gummy yellow fluid, called exudate, to fill the air sacs, stopping the free flow of oxygen. If only a few air sacs are filled, the rest of the lung takes over. When more and more alveoli are filled, the lung texture changes, beginning to feel more like a marshmallow than whipped cream. … These machines can’t fix the terrible damage the virus is causing, and if the virus erupts, the lungs will get even stiffer, as hard as a stale marshmallow. … The heart begins to struggle, begins to fail. Blood pressure readings plummet, a condition called shock. For some, the kidneys fail completely, which means a dialysis machine is also needed to survive. … Eventually, all the efforts of health care workers may not be enough, and the body begins to collapse. No matter how loved, how vital or how needed a person is, even the most modern technology isn’t always enough. Death, while typically painless, is no less final. Even among the Covid-19 patients who are ventilated and then discharged from the intensive care unit, some have died within days from heart damage.

18 thoughts on “Why the U.S. is out of ventilators (and why it might not matter)

  1. What prevented amassing warehouses full of masks, ventilators, prefabricated hospitals etc, in the past few decades between pandemics?

    A previous pandemic “killed … around 100,000 people in the U.S.” That wasn’t a huge national crisis then, was it? In proportion to population, that’s in the same range as the 100,000 – 200,000 range predicted for Coronaplague, but now that is a huge national crisis. Why the difference?

    • Covid is now estimated to kill up to 240,000 in the US this season, assuming a lengthy near-lockdown. If we went about business as usual, as we did in the 1968 pandemic, the estimate goes above two million. The 1968 pandemic killed 100,000 in the US over four seasons. So in the absence of extreme containment efforts, we’d be looking at a catastrophe perhaps 40 times worse than the 1968 pandemic. Hence the different reaction.

    • PhilG has linked to this story that describes the one to two hundred thousand estimate as “grim”, “stark”, and “dire”. I don’t think it’s generally been accepted that millions of deaths are unthinkable but about 200,000 are OK, has it?

      The point about the ’68 pandemic being of longer duration suggest another comparison. The Vietnam war killed about 60,000 US personnel, fewer than the ’68 pandemic and over a longer period (1965 – 1975, with some more before and after). Yet that war was a huge national crisis. So neither numbers of casualties nor duration seem to account for why this pandemic is considered so dire.

    • The Vietnam war was a large national crisis because nobody sane could understand why it was necessary in the first place.

    • Gary: Sweden and Brazil aren’t doing anything official in terms of containment, at least at the national level, and certainly not “extreme containment efforts.” Are you actually expecting those countries to experience a 0.6% death rate among their respective total populations?

      Also, with coronavirus now thought to be spreadable just by breathing do you actually think our porous “lockdowns” are going to cut the death rate by 10X? https://www.cnn.com/2020/04/02/health/aerosol-coronavirus-spread-white-house-letter/index.html

  2. My local hospital was tranquil from the outside. A dozen ambulances were idling by the emergency room entrance.

    NYC had almost 300 coronavirus fatalities yesterday. 150 is daily death rate for entire city in normal times.

  3. I was surprised to learn how much those little ventilators cost: about $30-40k each! Meanwhile, a dose of the “Trump pills” cocktail costs about $20, and (anecdotally at least) alleviates the need for a ventilator. Hmmm…

  4. If mom has trouble breathing, she’s going on the ventilator regardless of the risk of death, thresholds & statistics.

  5. Phil would you please ask some of your investor friends what stocks they like here, lets ask about Abbott Labs for example.

    • Joe: When they want to play in the market, they do much more exotic stuff than pick stocks. Like they’ll short closed-end mutual funds, which apparently do badly all of the time compared to the market as a whole, but even worse when things are volatile. What if the market goes up? They’ll be wiped out, right? No. They also short a closed-fund fund that itself short the market (some sort of “bear fund”?). Or they’ll sell out-of-the-money options (not a great strategy over the last couple of months!). Sometimes they’ll do a trade that I can understand, e.g., short Tesla when it goes above 800, but that’s rare.

  6. Obama always says if you like your giant stock pile of ventilators you can keep them!

  7. Thanks Phil, I get it, I was just looking for clues on companies, like Claude Shannon (he borrowed a card table from my Dad when I was a kid) looked for at Bell Labs, he spotted Hewlett Packard way before the market. He was quite the stock picker.

    “Stay Safe” is code for F off so I don’t use it.

    • Fun fact about Shannon, for any EE’s in the house: the Shannon-Hartley theorem that dictates how much information can be packed into a slice of bandwidth was disproven, and hardly anybody noticed. So, what is the limit? It’s infinity, but only if you rely on interference from multiple sources being used to simultaneously construct multiple local signals, rather than trying to avoid it. It’s just a big linear algebra problem to synthesize the waveforms and coordinate their transmission. The wizards at Artemis Communications proved this, and demonstrated 5G communications using their software-defined modems, negating the need for any changes in infrastructure to roll out 5G and/or infinite bandwidth to the masses. But after an investment by Google a few years ago, they’ve gone silent. Anybody know what they are up to?

  8. Very cool-thanks!

    All I know is the “pioneers” from Bell Labs always say to me “what did Google invent besides the search engine”.


  9. Maybe it’s worth some effort to figure out how to prevent BiPap and Hiflo vapotherm from spreading the virus. Is there a reliable filtration system that is known to filter coronavirus?

Comments are closed.