As much as doctors and hospitals are raising alarms about a shortage of ventilators for Covid-19 patients, some doctors have begun to call for entirely reassessing the standard paradigm for their use–according to a cluster of articles to appear in the last week. “What’s driving this reassessment is a baffling observation about Covid-19: Many patients have blood oxygen levels so low they should be dead. But they’re not gasping for air, their hearts aren’t racing, and their brains show no signs of blinking off from lack of oxygen.” Within that group of patients, some doctors wonder if the standard use of mechanical ventilators does more harm than good. The issue is controversial; I’ll just report what I find in the articles over the past week. Please share ongoing updates in the comments.
I. Gattinoni: “COVID-19 pneumonia: different respiratory treatment for different phenotypes?”
Luciano Gattinoni, one of the world’s experts in mechanical ventilation, “says more than half the patients he and his colleagues have treated in Northern Italy have had this unusual symptom. They seem to be able to breathe just fine, but their oxygen is very low”. …
He says these patients with more normal-looking lungs, but low blood oxygen, may also be especially vulnerable to ventilator-associated lung injury, where pressure from the air that’s being forced into the lungs damages the thin air sacs that exchange oxygen with the blood. 
Gattinoni labels these patients (more normal-looking lungs, but low blood oxygen) as Type L, and urges they be treated differently than the type of acute respiratory [ARDS] patients seen prior to Covid-19. This second type he calls Type H. (His editorial is in ). I found a picture of Type L and Type H lungs at this link on p. 12.
Patients with respiratory failure who can still breathe OK, but have still have very low oxygen, may improve on oxygen alone, or on oxygen delivered through a lower pressure setting on a ventilator.
Gattinoni thinks the trouble for these patients may not be swelling and stiffening of their lung tissue, which is what happens when an infection causes pneumonia. Instead, he thinks the problem may lie in the intricate web of blood vessels in the lungs.[a]
Gattinoni says putting a patient like this on a ventilator under too high a pressure may cause lung damage that ultimately looks like ARDS.
In other words, the high pressure of the ventilator may turn a Type L patient into a more serious Type H patient. “If you start with the wrong protocol, at the end they become similar,” Gattinoni said. Oy! He recommends the two types (which can be determined in a number of ways) be treated differently: Type L patients receive greater benefit from less invasive oxygen support, via breathing masks, such as those used for patients with sleep apnea, nasal cannulas, or via a non-invasive high flow device.
Gattinoni said one center in central Europe that had begun using different treatments for different types of COVID-19 patients had not seen any deaths among those patients in its intensive care unit. He said a nearby hospital that was treating all COVID-19 patients based on the same set of instructions had a 60% death rate in its ICU. [He did not give the names of the hospitals.]
“This is a kind of disease in which you don’t have to follow the protocol — you have to follow the physiology,” Gattinoni said. “Unfortunately, many, many doctors around the world cannot think outside the protocol.” 
But there are some doctors who may want to think outside the protocol, yet face pressure against doing so–according to Cameron Kyle-Sidell, an emergency room and critical care doctor at Maimonides Medical Center in Brooklyn.
Whyte: You’ve been talking on social media; you say you’ve seen things that you’ve never seen before. What are some of those things that you’re seeing?
Kyle-Sidell: When I initially started treating patients, I was under the impression, as most people were, that I was going to be treating acute respiratory distress syndrome (ARDS)… And as I start to treat these patients, I witnessed things that are just unusual. …In the past, we haven’t seen patients who are talking in full sentences and not complaining of overt shortness of breath, with saturations [blood oxygen levels] in the high 70s [normal is said to be between 95 and 100].[b].
This originally came to me when we had a patient who had hit what we call our trigger to put in a breathing tube, … Most of the time, when patients hit that level of hypoxia, they’re in distress and they can barely talk; they can’t say complete sentences. She could do all of those and she did not want a breathing tube. So she asked that we put it in at the last minute possible. It was this perplexing clinical condition: When was I supposed to put the breathing tube in?…
We ran into an impasse where I could not morally, in a patient-doctor relationship, continue the current protocols which, again, are the protocols of the top hospitals in the country. … So I had to step down from my position in the ICU, and now I’m back in the ER where we are setting up slightly different ventilation strategies. Fortunately, we’ve been boosted by recent work by Gattinoni.
Whyte: Do you feel that somewhere the world made a wrong turn in treating COVID-19?
Kyle-Sidell: I don’t know that they made a wrong turn. I mean, it came so fast. … It’s hard to switch tracks when the train is going a million miles an hour. …But I do think that it starts out with knowing, or at least accepting the idea, that this may be an entirely new disease. Because once you do that, then you can accept the idea that perhaps all the studies on ARDS in the 2000s and 2010s, which were large, randomized, well-performed, well-funded studies, perhaps none of those patients in those studies had COVID-19 or something resembling it. It allows you to move away from a paradigm in which this disease may fit and, unfortunately, walk somewhat into the unknown.
…One of the reasons I speak up, and I hope people at the bedside speak up, is that I think there may be a disconnect between those who are seeing these patients directly, who are sensing that something is not quite right, and those brilliant people and researchers and administrators who are writing the protocols and working on finding answers. The first thing to do is see if we can admit that this is something new. I think it all starts from there.
Gattinoni’s paper and Kyle-Sidell’s on-line discussions are having an impact in the popular press. Yesterday, the Telegraph reported that “British and American intensive care doctors at the front line of the coronavirus crisis are starting to question the aggressive use of ventilators for the treatment of patients”.
In many cases, they say the machines– which are highly invasive and require the patient to be rendered unconscious– are being used too early and may cause more harm than good. Instead they are finding that less invasive forms of oxygen treatment through face masks or nasal cannulas work better for patients, even those with very low blood oxygen readings….This is the sort of treatment Boris Johnson, the Prime Minister, is said to have received in an intensive care unit at St Thomas’ Hospital in London.
…Increasingly, doctors in the UK, America and Europe are using these less invasive measures and holding back on the use of mechanical ventilation for as long as possible…Invasive ventilation is never a good option for any patient if it can be avoided. It can result in muscle wastage around the lungs and makes secondary infections more likely. It also requires a cocktail of drugs which themselves can prove toxic and lead to organ failure.
“Instead of asking how do we ration a scarce resource, we should be asking how do we best treat this disease?” says physician Muriel Gillick of Harvard Medical School.
III. Need Non-invasive Ventilation Risk Health Care Workers?
Yet there’s an important reason the standard protocol is to bypass non-invasive ventilation in Covid-19 patients (in the U.S.), and I don’t know if Gattinoni or Kyle-Sidell address it: they are thought to pose risks for heath care providers, at least without adequate protective devices.[c]:
One problem, though, is that CPAP [continuous positive airway pressure] and other positive-pressure machines pose a risk to health care workers…The devices push aerosolized virus particles into the air, where anyone entering the patient’s room can inhale them [spillage]. The intubation required for mechanical ventilators can also aerosolize virus particles, but the machine is a contained system after that.
“If we had unlimited supply of protective equipment and if we had a better understanding of what this virus actually does in terms of aerosolizing, and if we had more negative pressure rooms, then we would be able to use more” of the noninvasive breathing support devices, said [Lakshman] Swamy [an ICU physician and pulmonologist of Boston Medical Center].
But surely it would be easier to procure adequate protective equipment than obtain more ventilators, especially if it’s a way to beat the grim statistics for a significant group of Covid-19 sufferers. Italy has special plastic helmets that cordon off the patient’s head from the shoulder up, redolent of Victorian diving helmets. A virus filter prevents the aerosolization risk that is behind the common protocol. The Italian helmet, however, hasn’t been approved by the FDA, and anyway, Italy has banned its export given its own COVID-19 crisis. Fortunately, at least one group in the U. S is building its own coronavirus helmets.
Please share your thoughts, updates, and errors.
Normally, when lungs become damaged, the vessels that carry blood through the lungs so it can be re-oxygenated constrict, or close down, so blood can be shunted away from the area that’s damaged to an area that’s still working properly. This protects the body from a drop in oxygen. Gattinoni thinks some COVID-19 patients can’t do this anymore. So blood is still flowing to damaged parts of the lungs. People still feel like they’re taking good breaths, but their blood oxygen is dropping all the same.
One doctor treating COVID-19 patients in New York [Cameron Kyle-Sidell] says it was like altitude sickness. It was “as if tens of thousands of my fellow New Yorkers are stuck on a plane at 30,000 feet and the cabin pressure is slowly being let out. These patients are slowly being starved of oxygen”. 
Lung scans show the same “ground glass” appearance in both covid-19 and high altitude pulmonary edema (HAPE).
[b] An oximeter I recently bought, of not very good quality, has me at 97.
[c] Except perhaps when mechanical ventilators are in too short supply.( I am not up on the current regulations). Of course, another reason is the danger in delaying intubation that might be necessary.
 March 17, 2020 around 30 cases down.