Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
COVID-19 Data: Survival Rates for Patients on Ventilators
From the UK’s Intensive Care National Audit and Research Centre (ICNARC): Report on 775 patients critically ill with COVID-19
The report looks at 165 COVID-19 patients whose status was resolved — they died or were discharged — out of a population of 775 COVID-19 patients who were admitted to intensive care by reporting hospitals.
Of these 165 resolved patients, 52% were discharged alive and 48% died in the hospital.
Of the 165 resolved patients, 60% of them required advanced respiratory care (ventilators or similar).
Of the 60% requiring advanced respiratory care, 34% survived and were discharged, 66% died.
Of the 40% not requiring advanced respiratory care, 80% survived and were discharged, 20% died.
Half of the patients who survived were in intensive care for between 2 and 7 days, with a median of 3 days.
Half of the patients who died were in intensive care for between 2 and 6 days, with a median of 2 days.
Since this report includes only resolved cases, it is biased toward cases that resolve quickly; the remaining cases in ICU may not resolve in the same proportions of favorable and negative outcomes.
Favorable outcome by age distribution for the 165 individuals whose cases were resolved are:
16-49 years 76% survived
50-69 years 60% survived
70+ years 27% survived
The report contains an interesting graph that appears to cover the larger population of 774 cases admitted to critical care, and to show the status of those cases over a span of approximately four weeks. I’m not sure how to interpret the assumptions implicit in the graphic, but it appears to show that: after one week, approximately 20% have recovered and ~ 12% have died; after two weeks, those numbers are roughly 40% and 30%, respectively; after three weeks, approximately 58% and 38%, respectively; and that those remaining in critical care past three weeks show slow change moving forward.
Published in General
Germany puts out some halfway decent data on this: https://www.divi.de/images/Dokumente/DIVI-IntensivRegister_Tagesreport_2020_04_03.pdf (in German obviously)
Basically, 83% of patients who land in the ICU need a respirator at some point. Of all patients (+/- respirator) with a known outcome, 33% died (the other 2/3rds recovered). So assuming that everyone who died was on a respirator, that still means that well over half of patients who go on a respirator at some point end up recovering. At least in Germany, which has not had overfilled ICUs so far.
The data are useless. Ventilator and ICU use depend on the availability of those items. The USA and Germany have many so healthy people get them. China and Italy have few so only the sickest get them. You can’t compare the dissimilar health systems.
I also heard that people in NYC are put on ventilators just to keep them from coughing in the hospital-effectively a muzzle. Serious people get intubated. If the medical procedures are not the same, you can’t compare the data except to prove the practices/populations are not the same.
I’m not comparing anything, Don. Just reporting numbers — and suggesting that the UK figures are probably reasonably representative of COVID-19 mortality for those who end up on a ventilator.
Henry, an important post. However you slice it though, using anyone’s data, once one gets to the ICU or worse on a ventilator the chances of recovery are not good.
The HCQ remedy appears to work very well at the early stages, but in the severe cases it appears we don’t have an effective treatment. Is that right?
DonG: “The data are useless. Ventilator and ICU use depend on the availability of those items. The USA and Germany have many so healthy people get them.”
A good question, which is difficult I know now to answer, is how do those who do not have pre-existing lung and other conditions that make them very vulnerable to the disease fare?
I certainly agree with the second half of that: we don’t have an effective treatment for those who have advanced to respiratory failure. The sad reality is that anyone who ends up on invasive ventilation is in a precarious position, and the peculiar pathology of this virus seem to amplify that precariousness substantially.
As for the hydroxychloroquine, let me speak carefully so as not to trigger anyone, as this has become a somewhat heated topic. ;)
I’m still waiting for better data. However, I find the anecdotal and limited study (and I use the term casually) data to be encouraging, and I’m in favor of the treatment being used widely and experimentally (that is, in a variety of treatment protocols as people learn more about it) while we wait for better data and/or a better alternative. I’m very eager for larger and better administered studies to be reported, and I’m optimistic about what I think we’ll hear.
Poorly.
The more general form of the question is: what are the correlations between various demographic factors (sex, age, assorted comorbidities) and critical events in disease progression (ICU admission, ventilator support, death)?
That’s going to require a lot more detail than has been made available. Presumably it will be out there, in hospital charts and case notes, when this is over. Until someone really starts analyzing it, we’ll probably only be able to see the grossest correlations, such as the pretty obvious mortality/age relationship.
Lots of gray in the numbers.
Facility/Doc A does not use a technology that is not likely to make any difference and because they are trying to conserve resources. Facility/Doc B keeps throwing resources at an identical case to the bitter end in part because they have the resources. A fatality in B gets reported as death with the technology used, a fatality in A does not.
The effects of this bug are weirdly variable. How big a viral dose at the outset seems to make an enormous difference in outcome. Death rates in different states, countries are impossible to reconcile. I suspect that even when it’s over, the numbers will still be somewhat mysterious.
Henry, the problem with the 66% figure is that most patients receiving mechanical ventilation do not yet have a reported outcome.
There are 775 ICU admissions reported. 518 received mechanical ventilation (Table 1), which is 78.7% of the 658 for which they had data about this. (So it appears that there are 117 patients whose ventilator use is unknown). There are 165 patients with a reported outcome, so apparently 610 are still in the ICU (or not reported — this is Table 2). Of those receiving advanced respiratory support (which may mean something in addition to mechanical ventilation), there are 98 with a reported outcome, of whom 66.3% (65) died and 33.7% (33) were discharged alive (Table 3).
But this means that there are at least 420 patients who received mechanical ventilation, whose outcome is not reported. There is no reason to assume that the 66% death rate will continue to apply to these people. They may all survive, in which case the death rate among those on ventilators would be about 12.5% (65/518).
I don’t expect them all to survive, but it is premature to conclude that the 66% death rate will apply to all 518 patients receiving mechanical ventilation in this sample.
By the way, the Physicians Weekly blog post that you linked is quite wrong about this study. It says: “Of 165 patients admitted to ICUs, 79 (48%) died.” This is a serious misunderstanding of the ICNARC report, which is a summary of 775 patients admitted to ICUs. The 165 patients are those with reported outcomes (about half of which were deaths). But presumably, the remaining 610 patients are still in the ICU, and we don’t know whether they will survive or not.
Jerry, good point. I followed the links to the Guardian article and see that you’re correct about what was omitted in the Physicians Weekly post.
I’ll revise the main post after I figure out which numbers remain valid, given that only resolved cases were considered.
Thanks for being observant.
Henry,
Let’s not try to extrapolate hard predictions but gain the inferences that are obvious. First, stay the hell out of the ICU!!! If you’re already in the ICU at best you’ve only got a 2 out of 3 chance of making it. Second, putting COVID-19 patients on a ventilator means that you’ve only got a 1 out of 3 chance of making it.
This screams for us to try the anti-viral drugs or the serum (from the recovered patients full of anti-bodies) before they go into the ICU!!! Yes, watch out for the well documented side effects of the drugs but, are you nuts, you’ve got to try something other than ventilators and keeping the whole world 6 feet apart. Please!!!!!
Regards,
Jim
Jim, hold on. You mean I want to stay out of the ICU? Shoot. I’ve been doing everything I can to get in. I’m glad I read this before i upped my game. ;)
But seriously, yes, it’s pretty obvious that much of the horror of a ventilator shortage has less to do with outcomes and more to do with our sense of helplessness in the face of need. If ventilators are a last desperate grasp for those with the least hope, it’s still something, and we find it intolerable that that something, however unlikely it is to work, might not be available to those who have no other options.
Henry,
I understand your explanation of the psychology but sometimes just understanding isn’t gonna make it. The ICU and a ventilator are already a bad gamble. It costs $20 for a 5-day course of anti-viral drugs. There is no reason that they must be administered at the hospital. I would prefer to have telemedicine monitoring of your vitals and that would be enough “care” for somebody to take a 5-day regime of these drugs.
We are backing ourselves into a corner. We don’t need to. We need to think outside the box or we’ll end up in a box.
Regards,
Jim
My friend, you are preaching to the choir. I wrote two weeks ago in this post that I think we’re going about this the wrong way.
However, I do believe that any path forward that unnecessarily risks — or that appears to unnecessarily risk — filling our hospital hallways with dying people for whom no ventilator support, no matter how ineffectual it may be, is available, will be deemed unacceptable by those making the decisions, because such an outcome will be seen as unacceptable by the American people. And I have a hard time blaming Americans for seeing it that way, no matter how mistaken they may be about the prospects for the most seriously ill.
Do you belong to the George W. Bush school of thought, that “If people hurt, we [the government] have to move?
No. I belong to the “regardless what I think, what the vast majority of other people think is likely to drive the political decision-making more than my outlier position” school of thought.
In other words, I’m being descriptive, not prescriptive. In a charged environment the difference between those two perspectives is routinely overlooked.
By Monday the projected horror story about unemployment, foreclosures and bankruptcies will begin to look worse to government leaders than the horror story about overloaded hospitals.
I am willing to bet that Pelosi and Schumer have a narrative in place to blame Trump for deaths from the moment he overrides the epidemiologists to prevent social breakdown. And then they will also try to concoct a mismanagement scenario for the economic harm.
If Trump suggests that we open in places where there is low incidence with use of masks with testing to more rapidly isolate and trace and get back to work in a time frame much shorter than the governors are currently putting in place, he will be leading in the way we want.
Orders to stay home until June 1 are absolutely nuts. We cannot sustain that.
I agree that that’s where we are. It just pains me. I’m glad I’m old.
As of today, the number of critical/serious cases in UK is 163 (worldometers).
That is shockingly low.
If this number is correct, the UK may be close to getting herd immunity or a transmission rate under 1.
I’m surprised there is no post on Cuomo’s executive order to take ventilators from other places in NY to be used in NYC.
https://buffalonews.com/2020/04/03/cuomo-moves-ventilators-downstate-death-toll-here-rises/
With the length of time people are in critical care, how can it be justified to move equipment from places that will surely need them before NYC patients either recover or die.
It’s terrible. Like your neighbor raiding your pantry in a famine.
Jules, I haven’t commented on it — even though it is precisely my area that will be tapped for resources — because I don’t have any objection to the dynamic reallocation of medical resources in an emergency. I want it to be done competently, but I don’t mind it being done. We’re all in this together, this thing flares up with different intensities at different times, and the needs will move around.
I have little faith in Cuomo’s wisdom. I don’t like the man or approve of his brand of politics, nor of a great many of his policies. But I’m ignorant about the specific needs of specific places, so I won’t offer an opinion about this… except to say that I hope and assume that the emergency support will flow both ways.
@henryracette yes, resources dynamically shared, but seizing is not sharing.
There has to be a way to figure some kind of balance for any community that built reserves to keep the reserves for what appears impending need.
Poor data quality, do not attempt analysis based on worldometers critical, discharged or test count numbers. They are inconsistently reported between countries and I strongly suspect not reported at all in many cases, since we’re dealing with healthcare staff who may be in extremis. They are going to work the case load, not the reporting. Dead is dead, so you can mostly count on that number, unless it’s China, Iran, North Korea (or maybe France).
Exactly.
Case counts are perhaps useful if we use them as a proxy for the number of people presenting at hospitals in some distress. Even then, it’s unclear how many are still being tested versus being evaluated and sent home with self-quarantine instructions. It’s also unclear how many are being tested elsewhere, with a similar prescription to just go home.
Recovered counts are, I think, essentially meaningless. I can find no evidence of a coherent follow-up policy or clear CDC instructions for healthcare providers to notify a central registry. I think we’ve no idea how many have recovered.
Death statics are the closest thing we have to a reliable number, and even there the guesstimates of the percentage of elderly patients with various comorbidities who die of COVID-19 rather than merely with COVID-19 range pretty broadly. I’ve seen estimates that as many as 30% of deaths in this demographic may not actually be attributable to the disease itself. (That seems too high to me, but I know little about the normal morbidity in this demographic.)
And death statistics, as you observe, are calculated in very different ways across national boundaries. I wonder how consistently they’re tabulated within the US, in different states. I’m going to assume we’re fairly consistent, but I don’t know.
For one thing, they won’t surely need them in Buffalo. They may need them, and they may not.
The alternatives are to leave localities to keep their own stuff and only yield it voluntarily — which seems unlikely — or for a higher authority to step in.
Like Henry, I don’t claim that it’s a good or bad policy at the present time, because I don’t know enough. But I don’t think that it’s unreasonable in an emergency for government to take such action.
It is easier to tolerate if you trust the state leadership. I haven’t been a fan of Cuomo. He does seem to be doing a decent job in the emergency, though, and I’m thankful for that.
I agree about that. If I had my druthers, there would be vastly less commanding going on, and a lot more imploring. I think the authoritarian impulse has been unleashed, the petty tyrants of our overweening bureaucracy are enjoying their moment of glory, and it’ll be good when we can tell them to go back to their dingy little caves and shuffle their paperwork.
No problem. It’s rather annoying that we can’t altogether trust online sources. The Physicians Weekly blog post actually seemed professional, and more credible than a newspaper story, but it made a significant mistake.
Perhaps the news has been getting it wrong for a very long time, and we simply haven’t had the information to know.
OB, I concur. On another post I said that civil disobedience is on the way. You can’t have so many people ordered by the government not to work, not to earn a living, not to go outside and forced to go broke, forced to use up their life savings/retirement funds, forced to wait in food/bread lines and not expect some push back at some point. That point will be reached when we begin to hear more stories of people losing their homes because they can’t pay the rent or pay the bank or people being arrested for being out side or holding a church service. When that time comes I pray it is civil disobedience and not uncivil disobedience.
One of the benefits of this strange period is that it prompted me to improve a few graphics-related skills. I’ve become more adept at creating my own graphs in Microsoft Excel, and at quickly extracting useful graphics from other sites for insertion into posts and comments.
Here’s the graph that I think you are referencing, from the ICNARC report.
I think that you have correctly interpreted this figure. The dotted lines are the confidence intervals, which implies that there is a statistical model behind the graph.
At the far right end of the graph, after about 4 weeks, it is projecting about 60% survival (of those admitted to ICU), about 37-38% fatalities, and 2-3% unresolved after 4 weeks. The confidence interval on the survival figure is around 54%-67%.