Dr. Cameron Kyle-Sidell is an ED-ICU Doc at Maimonides in New York, currently under the COVID fire. He put up an inflammatory tweet yesterday which caused a lot of different reactions, because it clearly challenged the widespread even if only recent belief that one should intubate the COVID patients quite early, usually much earlier than one would in typical respiratory failure:
So here you go:
Please share your thoughts and experiences – my comments in bold!
Very interesting. My experience is zero cases so far with COVID-19. Just some thoughts for what it’s worth from someone with no experience with COVID: Hypoxia is due to V/Q mismatch, diffusion, hypoventilation, low pO2 (altitude), shunt, low cardiac output (as more O2 is extracted off HgB and mixed venous O2 is low), and finally hemoglobin problems (whether by poisoning or defective HgB).
It was pointed out to me by Brett Berliner tonight in a paper (via twitter) that SARS CoV-2 appears to disrupt the beta chain of HgB to take up O2. I’d never heard of this, but if this is the case, exchange transfusion makes sense as a therapy for hypoxia as long as one is past the stage of illness where the viral levels are low. If this is a hemoglobinopathy, no amount of increase in mean airway pressure or FiO2 will fix the hypoxia.
Perhaps people out there with actual experience with COVID-19 can shed additional light on how the hypoxia tends to respond to positive pressure or increases in FiO2, or more importantly a lack of expected response, that would indicate the mechanism of hypoxia (such as a hemoglobinopathy).
It is very interesting that the hypoxic patients described in the post don’t appear to be very symptomatic in terms of SOB, nor do they attempt to compensate for the ding to DO2 from low O2 sats (O2 content) with an increase in HR to drive up CO. Normally DO2/VO2 ratios are around 4-5, so either the illness suppresses the normal physiological response to a decrease in the ratio, or the VO2 is suppressed somehow (seems unlikely). In any event, perhaps we should revisit the idea to intubate early if hypoxic, unless the patient actually looks really bad.
I’m Pulm/cc. This was fascinating and sparked some ideas. But, I need someone smarter than me to take it forward. The first thought is this is not what we know. It’s a different beast. Stop trying to make it fit what we know.
Why are people not short of breath or tachycardic? I suspect most of this is that compliance is high and there is no issue with expiratory flow limitation, hence WOB is more or less normal… And once hypoxia is corrected with O2, they feel fine.
Does the virus hide hypoxia from the body? Maybe it turns off normal feedback that would make us short of breath and tachycardic. The body doesn’t compensate because it doesn’t know it’s hypoxic. That’s why people are being found dead at home because they didn’t know they were slowly suffocating. Maybe it’s not affecting kids because their feedback mechanisms are more heightened already because of growth. Trying to think outside of the box after listening to this. I definitely agree that treating these patients like normal pneumonia with vent/pressors/fluids isn’t the answer.
Looks like these pts have severe V/Q mismatch without abnormalities in respiratory mechanics. Almost like a cyanotic heart condition. I agree if the mechanics are not worsening we may not achieve a lot by early intubation except to prevent sudden deterioration, if anything. HFNC with 100% will be the right choice. Has anyone tried nitric oxide before intubation? Might work to improve V/Q mismatch and reduce RV strain.
I am definitely planning to look at RV function and use inhaled vasodilators both for the VQ and for RV function.
Great talk and very fascinating points. Short of hyperbaric chambers for everyone, I think keeping everyone on high flow for as long as possible seems very reasonable. Given the ongoing reports I’m seeing regarding the pathophysiology regarding COVID-19 related lung disease, including your shared observations, it seems like using stress index on intubated patients would be the best measure of appropriate PEEP in this setting without “trashing” the lungs. If our understanding is correct regarding the atelectatic, high compliance Sars-cov-2 lung is correct, then the PEEP table is not going to accurately predict the proper settings and the stress index measured once a day would solve the problem of too much PEEP.