Substack cometh, and lo it is good. (Pricing)

COVID-19, the springtime of 2020

A lot of my Covid-19 commentary is on Twitter, but since I delete my tweets every 2 weeks it’s ephemeral. So I’ll post about once a week about a “status update” of sorts of my perceptions, predictions, and general sense.

First, I’m more optimistic than I was a few weeks ago. The main reason is that most of the nation is shut down, and social distancing is happening, broadly, albeit to various degrees. My mid-February back-of-the-envelope estimate of ~500,000 excess deaths is unlikely. I think the range of 100,000 to 200,000, given by the American government, is reasonable. But I’m hoping we can do a little better than that. Let’s say ~85,000.

The reason I’ll go on the low side is that I think there is still much to learn. And, a lot of the unknowns are “positive” for us in the medium term. If regional heterogeneity persists, that means we can learn from the successes (e.g., test-and-trace as in South Korea). Perhaps there is a “miracle drug” in the offing. We don’t know. I think the “worst-case” scenario of untrammeled infection is pretty well understood, so I don’t actually see as many downside unknowns.

A lot of people are asking me about the IHME model, and the fact that hospital bed shortfalls don’t seem to be as bad as the model predicted. I think this isn’t an issue with the model, it’s an issue with how accurate people think models are going to be. Model-building is important for fine-grained decision making but at the high level, the metrics are going to be more coarse. Currently, the COVID-19 epidemic is highly regional in the USA. A friend whose best friend is a medical resident at a New York City hospital confirms it is as bad as the media represents.

That being said, the western United States seems to be doing OK. The catastrophe in Washington state never really materialized. People are dying and have died, but the early actions of the government and populace of Washington seem to have forestalled a major outbreak. Similarly, the early actions of the Bay area leadership seem to have made a difference.

With that optimism put out there, we need to note the difference between mortality and morbidity. COVID-19 patients who have to go to the hospital use many resources and even mild cases have a longer course than the flu. More importantly, like SARS it is likely that for the severe cases, who outnumber the deaths by several folds, there will be lifetime problems.

Some parts of the USA will come out of lockdown on May 1st. Others will not. We will have a better sense of the variables that impact R and virulence as the summer progresses.

Speaking of variables, Obesity and COVID-19 Severity in a Designated Hospital in Shenzhen, China:

Of them, 53·1% were normal weight, 4·2% were underweight, 32·0% were overweight, and 10·7% were obese. Patients with obesity, versus without, were tended to have cough (P=0·03) and fever (P=0·06). After adjusting for potential confounders, compared to normal weight, overweight showed 86% higher, and obesity group showed 2·42-fold higher odds of developing severe pneumonia. Despite a non-significant sex interaction was found (P=0·09), the association appeared to be more pronounced in men than in women. The odds ratios (95% confidence intervals) for severe pneumonia in overweight and obesity was 1·96 (0·78-4·98) and 5·70 (1·83-17·76) in men, and 1·51 (0·57-4·01) and 0·71 (0·07-7·3) in women, respectively.

The average American adult has a BMI of ~26.5. We define “obese” as BMI > 30.

14 thoughts on “COVID-19, the springtime of 2020

  1. Apropos the biggest potential risk upsides which await the data:

    1) Antibody-enhanced disease (which may already be responsible for excess mortality in older people / larger outbreak sizes / large hospital settings through exposure to, and reinfection with, serologically divergent strains of the current virus and/or preexisting non-protective immunity to other coronaviruses). AED may play both against the natural immunity and against vaccine-induced immunity.

    2) spread in the insular, uniquely predisposed communities (either due to cultural practices such a religious conservatism, or to genetic make-up as in the tribes)

  2. If the disease is highly regional we are we locking down most of the country?

    In the case of obesity, I’d like to say the CDC is trying not to panic the population. The best place I’ve seen on BMI is the NHANES study, and something like 40% of older people are obese and another 30% overweight.

    But very clearly marked as a risk factor in Europe, but not here.

    https://wwwn.cdc.gov/Nchs/Nhanes/2011-2012/BMX_G.htm

    https://dqydj.com/bmi-distribution-by-age-calculator-for-the-united-states/

    Another factor is low Vitamin D levels, which might be explaining by black and brown people are dying more in Europe and north America.

    To reduce it down this is disease that attacks unhealthy epithelial tissue, and that is a hallmark of “metabolic syndrome”.

    It looks as is the UWASh model was significally redone yesterday. All models are wrong, and I’m not sure this one turned out to be useful. And let’s define useful — $10 trillion.

  3. I would be interested to know how smoking / vaping plays in as a risk factor for pneumonia or lung damage from the coronavirus. Haven’t seen anything about that in the popular press.

  4. Thanks for all the info you have been providing. If it wasn’t for you, I’d be way behind the curve.

    Appreciate it.

  5. The biggest downside I see is in hard-hit places like NYC (& Bergamo & Madrid etc). I don’t see how you ever get cases low enough that test & trace is feasible. I see no alternative to it continuing to burn through the population, even if it is at a “manageable” level (by which I mean nobody dying in the streets, while hospital ICUs are basically filled with COVID patients for months on end). This might result in ~75K excess deaths in NYC alone (8.4M*0.60[herd immunity target]*0.015[CFR]).

    The second downside is seen in the fact that Singapore has lost control of the track & trace strategy, and now is resorting to lockdown. This despite their renowned efficiency and equatorial environment.

    I think 85,000 deaths would be possible iff nationwide lockdown could be maintained for months. Since it can’t/won’t, there will be a seesaw battle between behavioral economics and the inexorable logic of the virus.

  6. the herd immunity is probably higher that 0.6. someone who dug deep think that first-pass estimates are low-balls. so closer to 0.7?

  7. Herd immunity has been estimated as a simple function of Ro:
    threshold(HI)=(Ro-1)/Ro

    So if Ro=3, for example, the threshold to achieve herd immunity will be 67% of the population.

    It stands to reason that this should also be influenced by the dispersion factor k, but I can’t remember exactly if this was discussed in the references I previously provided.

  8. @Joe Q – COPD can be taken as a proxy for smoking in long term smokers, although people can get it other ways, e.g. working for a long time in very dusty environments (particularly fine silica dust), and it is a subset of chronic respiratory disease, which is one of the high risk co-morbidities; there are other subsets, e.g. chronic asthma, but not too many I think (maybe AmBer can correct me on that). I have seen no specific figures for smokers and Covid-19 outcomes, but ~ 50% of males in Mainland China smoke (c.f. 10% in Hong Kong). No idea about vaping, which is illegal in Hong Kong (don’t know its legal status in the Mainland, but it doesn’t seem to be a thing). But note that chronic asthma is also likely to be pretty high in the Mainland (as it is in Hong Kong) due to poor air quality. Oxides of nitrogen, sulphur dioxide and ozone do your lungs no favours. Neither do the decay products of radon, which is found in high concentrations in buildings in Hong Kong because we have very ‘hot’ granite bedrocks (high uranium content). The effects of smoking + exposure to high concentrations of radon are synergistic.

  9. Sorry, should have added fine respirable particulates as one of the particularly damaging factors in poor air quality.

  10. Looks like I might be wrong about chronic asthmatics being significantly more at risk – that doesn’t seem to be supported by data. That would make ‘chronic respiratory disease’ a closer proxy for COPD, and so a closer proxy for smoking.

  11. @John Massey Maybe an off tangent question, but one I had wanted to ask a few days ago in the thread regarding China’s response to COIVD19. I’ve heard news from Taiwan claiming, of all things, they tried to warn the WHO as early as Dec. 30th that there was human to human transmission to Wuhan (I suspect some of their businessmen-disguised-intel operatives in Hubei overheard the rumors early on):

    https://www.ft.com/content/2a70a02a-644a-11ea-a6cd-df28cc3c6a68

    Do you think it’s possible the central CCP regime also knew of human to human transmission much earlier (~mid to late December) than they’re willing to admit?

  12. Riordan – No.

    I don’t know how anyone in Taiwan could know that this was an infectious disease with human to human transmission by then when doctors in Wuhan did not know that by then.

    I know what doctors in Wuhan were saying in December because my doctor was talking to them, and all they knew was that they had a puzzling cluster of cases of ‘atypical pneumonia’ which was worrying them, because of the severity of the pneumonia – it was ‘atypical’, something they hadn’t seen before. I went to see my doctor on December 31 and she told me that she was worried about it, and that I should “be careful”, and those were the words she used – a cluster of cases of ‘atypical pneumonia’. Nothing about a new virus, nothing about human to human transmission. She must have been worried about that possibility, because she was around during the 2003 SARS epidemic, as I was, and Wuhan is only 4.5 hours by high speed rail from Hong Kong, but she certainly didn’t know that at that point, and doctors in Wuhan didn’t know either, otherwise they would have told her.

    December 31 was the day that the Chinese health authorities (necessarily with the prior approval of the central government) notified the World Health Organisation of 41 cases of ‘atypical pneumonia’ in Wuhan. That was entirely consistent with what my doctor told me the same day.

    Footnote: Please don’t ask me to respond to every bullshit story coming out of Taiwan or circulating on the Internet on this subject. I don’t mind at all in this case, because you seem like a reasonable person, and purely by chance my doctor talked to me about it and I remember exactly what she said.

  13. The monthly death rate in the U.S. is around 230.000. If there will be 85.000 Corona deaths, we should count that to be a success. 230.000 Corona deaths spread over 2 months, let’s say, would be an increase above normal of 50%. Maybe still a success. (The rate of 230.000 is my calculation, namely 7.700 daily (in 2017) times 30).

    By comparison, the most afflicted areas in Italy had a three to four times increase in death rate (as far as I am aware).

Comments are closed.