What explains differences in severity?

One of the unknowns in the virus crisis is what explains differences in severity. Of the people who have been infected, it seems that more than 95 percent experience low severity. Also, we see wide differences in severity across countries. Is Taiwan doing better than Spain because fewer people have been infected in Taiwan, or the infections are less severe in Taiwan, or both?

It seems to me that the possible explanations for variations in severity include:

1. How you are attacked–how much of the virus you get and how far it goes initially into your respiratory system.
2. How well your individual body defends.
3. How you are treated by the health care system.

The conventional wisdom, as I understand it, is that (2) matters, and I believe this conventional wisdom. That is, we think that young people without underlying conditions defend better once infected than do old people or people with underlying conditions. Of course, it would be better to have knowledge of which underlying conditions affect the ability to defend.

The conventional wisdom, as I understand it, is that (3) matters, but I am skeptical about it. The conventional wisdom is that we need to keep the number of hospital beds and ventilators ahead of the spread of the virus, or otherwise people will die unnecessarily. The conventional wisdom seems consistent with the high death rates in Northern Italy, Spain, and New York City. But there could be other explanations. Perhaps the rate of infection was higher in those areas. Perhaps how you are attacked matters, and people in these areas were more likely to be attacked more severely.

Suppose that more ventilators and hospital beds had been available in these dire regions. Would that have produced more cures, or merely kept some people alive a few more weeks? I am getting the impression that a shortage of ventilators means that victims who are beyond hope might have to be denied a ventilator, but it is less clear that people who could survive if given a ventilator must be denied one. I am by no means committed to this point of view. It is just a guess. Any evidence to the contrary would be sufficient to get me to change my mind.

The conventional wisdom is relatively silent about (1). But I wish we knew more. For example, suppose that strong attacks only come from symptomatic spreaders, while getting the virus from an asymptomatic spreader means that you face a weak attack. That would imply that fears of asymptomatic spreaders are exaggerated, which would have some significant policy implications. It would imply that a focus on identifying and isolating the symptomatic individuals is the key to preventing deaths. It might mean that universal masks and scarves, while not preventing all infections, might do well at preventing severe infections, particularly if symptomatic individuals are identified and isolated.

16 thoughts on “What explains differences in severity?

  1. One pontential piece of evidence for channel number one (how you are attacked–how much of the virus you get and how far it goes initially into your respiratory system) is the relatively large number of seriously ill health care workers. They get exposed to a heavy viral load as part of their work. If it weren’t, in part, a viral load issue, they should have a serious illness distribution (conditional on infection) much like general population. In fact, because they have better access to testing, they should actually have lower conditional serious illness probability but it seems to be higher.

    • “the relatively large number of seriously ill health care workers”

      Given the hours and stress of the jobs, I could also see this being potential evidence for #2. For example, if you had an earthquake that caused healthcare workers (some of whom already work 24-hour shifts) to strain beyond normal, maybe you’d see lots of them get sick, too.

    • “the relatively large number of seriously ill health care workers”

      We just don’t have enough information to connect health care worker illness rates to support the contention that the viral load of initial exposure causes proportionally more severe symptoms.

      We have other good reasons to believe that to be true, from experience and studies with similar viral infections. But we just haven’t done enough tests to know what is really going on.

  2. I was under the impression that obesity was the most predictive underlying condition for a severe response isn’t it?… w/ diabetes, pulmonary disease, cardiac issues bundled in.

    Obesity and COVID-19 severity in a designated hospital in Shenzhen, China – The Lancet
    https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3556658

    Covid-19 in Critically Ill Patients in the Seattle Region — Case Series NEJM
    https://www.nejm.org/doi/full/10.1056/NEJMoa2004500

    Characteristics of COVID-19 patients dying in Italy
    https://www.epicentro.iss.it/coronavirus/bollettino/Report-COVID-2019_26_marzo_eng.pdf

    Telegraph UK – Obese or overweight coronavirus patients most in need of critical care
    https://www.telegraph.co.uk/news/2020/03/23/obese-overweight-coronavirus-patients-need-critical-care/

    CDC
    https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/groups-at-higher-risk.html

    -it’s certainly not access to ventilators.
    https://www.nytimes.com/2020/04/04/opinion/coronavirus-ventilators.html

  3. It will eventually be understood that New York was particularly bad because of the use of mass transportation and cabs. It could also eventually be shown to be a consequence of of high density housing unit use. I write this because it is what New York has in common with Europe and less so with almost all the rest of the US.

    • There are lots of nyc webcams and recent social media images available. If you look above ground, some of the busiest spots on the nicest days of the year, are shockingly empty. Wall Street and Times Square, whoa.

      But if you look below ground, there apparently was not much attempt to prevent the subways from running crowded with people without masks, some very sick and literally on their way to seek medical attention. If it really were true that those were all ‘essential’ workers, then everybody essential had a very high chance of getting infected in a narrow slice of time. An approach not entirely without risk.

  4. The disease in the northeast spread out of NYC along the commuter railways into and CT and NJ.

  5. There may also be differences in how well the virus binds to the ACE2 and/or how well in replicates one inside the cell.

  6. If a key factor in the virus is high density housing, then the vast rural and low density suburban areas of America should not be at high risk at all.

    Let’s watch and see if their current disease load (which is tiny) actually grows.

  7. 1 matters a lot. 2 matters a lot – a weakened immune system can not mount a good defense.

    3 only treats symptoms for the most part. There are some drugs, but not good ones

  8. Somewhat on topic, where does one look now for the freshest and most reliable estimates of the following: Once a healthy person of age X is infected, 1) likelihood of symptoms; 2) likelihood of hospitalization; 3) likelihood of a stay in the ICU; 4) likelihood of RIP. Obviously there are variables (e.g. how well a hospital can accommodate your needs, as suggested by ASK); and the numbers will change (treatments). I think healthy people should know their risks once the lockdown loosens.

  9. “That would imply that fears of asymptomatic spreaders are exaggerated,”

    I don’t think that necessarily follows. Even if the people that the asymptomatic spreader infects only experience a low severity, they could in turn spread a high dose to the people at home with them if you don’t notice the infection quickly enough.

Comments are closed.