Bacon's Rebellion

Democracy Thrives in Sunlight

Biocomplexity Institute Unveils COVID-19 Model

—

by

James A. Bacon
in ,

Under one Biocomplexity Institute scenario, COVID-19 demand could start exceeding Virginia hospital capacity for most of the state by mid-August.

by James A. Bacon

John Butcher and I have been cautiously optimistic that the COVID-19 epidemic is near its peak in Virginia, and that Virginia hospitals seem to be in little danger of being overwhelmed. Now comes the University of Virginia’s Biocomplexity Institute, which has developed a sophisticated, Virginia-specific epidemiological model. The Institute agrees that social-distancing restrictions are working, but warn that the epidemic could heat back up if they are relaxed.

Back in January, the administration asked the Institute to design an epidemiological model as a decision-support tool for Virginia public health officials. The Institute has done so, and yesterday it released its first publication, “Estimation of COVID-19 Impact in Virginia.” Institute leaders also elaborated on their model Monday in a press conference with Governor Ralph Northam Monday.

Under a scenario in which current restrictions are maintained through June 10th in line with Northam’s stay-at-home order and in which public activity changes “halfway” back to normal afterwards, the epidemic could peak in August-September, reports the Richmond Times-Dispatch. In that case, hospitals could be overwhelmed across the state, with the most severe problems in Northern Virginia.

The Institute factored some interesting data into its model, including Google mobility data showing a 44% decrease in retail & recreation activity, 18% reduction in grocery stores, and 39% decrease in workplaces. Cuebiq data shows a 50% reduction in average individual mobility. Meanwhile, the UVa quant jocks hope to improve the model by such techniques as incorporating age structure into its transmission dynamics and assessing evidence for the role of seasonality.

It’s reassuring to see that the Northam administration has its own Virginia-specific epidemiological model and that it’s maintained by people who know what they’re doing. It’s likewise reassuring to know that the Northam team had the foresight to ask the Biocomplexity Institute as early as January to work on the project. Now the administration can make better-informed decisions about when, and to what degree, it can start rolling back its social-distancing restrictions.

We can still debate the trade-offs between public safety, the economy, and individual liberties. But we can do with a clearer idea of what those trade-offs are. Well, maybe “clearer” is overdoing it. But in the land of the blind, the one-eyed man is king — even if he has cataracts and macular degerneration.

Update: John Butcher is less impressed. “Lots of grant $ and computer time to tell us what we already knew.”

Share this article

ADVERTISEMENT

(comments below)

ADVERTISEMENT

(comments below)

Comments

24 responses to “Biocomplexity Institute Unveils COVID-19 Model”

  1. LarrytheG Avatar
    LarrytheG

    All of these models – are modelling what happens when we are socially distancing.

    Is there a model that shows what happens if we do not do social distancing?

    It’s hard to understand that many are looking at these models – that are actually showing the effects of social distancing… looking at the “peak” and deciding the down slope is pointing to where we can end social distancing.

    Virtually all of the epidemiologists are warning that’s a wrong way of looking at the graphs.

    How about a model that shows what happens if we end social distancing?

    1. virginiagal2 Avatar
      virginiagal2

      There are model runs that show that. Basically, it bounces back up. You have to change things to change the curves. That means things like getting better treatments out, better supply of PPEs, more testing, better identification of contacts, and almost certainly some longer term behavioral changes.

      It’s really important to understand how damaging this disease is. It’s not influenza. Riding it out is not a reasonable option.

      See https://www.washingtonpost.com/health/coronavirus-destroys-lungs-but-doctors-are-finding-its-damage-in-kidneys-hearts-and-elsewhere/2020/04/14/7ff71ee0-7db1-11ea-a3ee-13e1ae0a3571_story.html

      “ The new coronavirus kills by inflaming and clogging the tiny air sacs in the lungs, choking off the body’s oxygen supply until it shuts down the organs essential for life.

      But clinicians around the world are seeing evidence that suggests the virus also may be causing heart inflammation, acute kidney disease, neurological malfunction, blood clots, intestinal damage and liver problems. That development has complicated treatment for the most severe cases of covid-19, the illness caused by the virus, and makes the course of recovery less certain, they said.

      The prevalence of these effects is too great to attribute them solely to the “cytokine storm,” a powerful immune-system response that attacks the body, causing severe damage, doctors and researchers said.“

      1. LarrytheG Avatar
        LarrytheG

        and yet there are those in BR that are advocating “riding it out” OR “hiding under the bed”.

        Geeze..

  2. Ben Slone Avatar
    Ben Slone

    See https://fivethirtyeight.com/features/a-comic-strip-tour-of-the-wild-world-of-pandemic-modeling/…

    1. LarrytheG Avatar
      LarrytheG

      talk about a conundrum! 😉 thx for posting!

    2. Acbar Avatar
      Acbar

      This “comic” strip format is well done and well-worth the slog to read all the way through it.

      1. Reed Fawell 3rd Avatar
        Reed Fawell 3rd

        Regarding:
        See https://fivethirtyeight.com/features/a-comic-strip-tour-of-the-wild-world-of-pandemic-modeling/…

        Yes, Acbar, I agree. Indeed, I’d suggest that cartoon gets closer to the modeling problems as anything that I have read. Likely, I feel that way because it goes into detail on an important limitation that I touched on earlier regarding modelling generally. Most modelers refuse to know and/or appreciate what they do not know. This is a very human failing, comprised of several interactive parts. We all desperately seek certainty in a grossly uncertain world. This gives us all a false sense of security. Many of us also want to make our own unique mark, to gain acceptance, fame, power, and riches. So we too often are blind to facts in order to make that mark. Hence folly, blindness, and hubris are the default positions of ambitious people, and each failing (sin) feeds upon the other, among highly ambitious people, to point of obsession and compulsion.

        Hence, for example, the World Health Organization (WHO), charged with protecting the health of the world’s peoples, in mid-January, declared with certainty (without qualification) that Coved-19 was NOT contagious, that the virus could not be passed from one human being to another. Pause a moment, and try to wrap your head around that fact, then ask why?

      2. Reed Fawell 3rd Avatar
        Reed Fawell 3rd

        To begin to fully appreciate the steps taken by other successful nations in stemming the Covid-19 tide, I suggest this fine article I found in MIT Technology Review. It shows how slow Virginia has been in protecting Virginians. It also shows the incredible recklessness of the World Health Organization (WHO).

        Here is one quote on the background behind WHO debacle, though WHO is not mentioned:

        “It was also thanks in part to SARS that Asian countries understood the need for rapid action, as Leo Yee Sin, head of the NCID, noted back in early January. At that point, covid-19 was still being referred to as a “mystery pneumonia.” Around the region, passengers on flights from affected parts of China were given mandatory temperature checks. As the crisis deepened, those flights were canceled, and then borders were closed entirely. Not every country followed quite the same model of response: Hong Kong and Japan shut their schools early, while Singapore kept its open. But all acted quickly, in coordinated responses led by experts.”

        Here is another comment reflecting on Virginia irresponsible conduct during this crisis:

        “Nonetheless, open communication from governments has been a consistent pattern in Asia’s more successful responses.”

        The article is titled

        “How to manage a pandemic
        Coronavirus was a test, and many of the world’s most advanced nations have all too visibly failed. What can we do better? by James Crabtree, April 14, 2020

        The article in MIT Technology Review is at:

        https://www.technologyreview.com/2020/04/14/999239/how-to-manage-a-pandemic-covid-asia-vs-west/?truid=a39bb7a8441b9bca5e48857d62d0351d&utm_source=coronavirus&utm_medium=email&utm_campaign=coronavirus.unpaid.engagement&utm_content=04-15-2020

        1. LarrytheG Avatar
          LarrytheG

          interesting article:

          Asia’s democracies often took more basic routes, monitoring and managing the outbreak with tools no more advanced than phones, maps, and databases. Singapore in particular rolled out an admired contact tracing system, in which centralized teams of civil servants tracked down and contacted those who might have been affected. Their calls could be shocking. One minute you were oblivious at work; the next minute the Ministry of Health was on the phone, politely informing you that a few days before you had been in a taxi with a driver who subsequently fell ill, or sitting next to an infected diner at a restaurant. Anyone getting such a call was sternly instructed to sprint home and self-isolate.

          What made this possible was that anyone infected could be grilled for hours. “They sat me down and interrogated me about my travel: every day, minute by minute,” my friend told me. “Where did I go? Which taxi did I take? Who was I with? For how long?” The process of tracking and tracing was laborious but produced impressive results. Nearly half of the roughly 250 people infected in Singapore by mid-March first learned that they were at risk when someone from the government called and told them.

          Just as efficient was South Korea’s testing regime, which in January forced local medical companies to work together to develop new kits and then rolled them out aggressively, allowing planners to keep track of the pandemic’s spread. South Korea had tested about 300,000 people by late March, roughly as many as the United States had managed by then, but in a country with a population one-sixth as large.”

      3. Reed Fawell 3rd Avatar
        Reed Fawell 3rd

        To keep abreast of this crisis I highly recommend The Coronavirus Tech Report from MIT Technology Review. It comes out daily and free to all who wish to subscribe.

        To get it go to:
        https://www.technologyreview.com/tag/coronavirus-covid-19-news/

  3. Nancy_Naive Avatar
    Nancy_Naive

    What are your credentials, and John Butcher’s, that gives credence to your skepticism? Scientists? Mathematicians? Epidemiologists? Or, just Bloggers Extraordinaire?

    Gut level? Or, just science denying because you can, or must.

    1. Ben Slone Avatar
      Ben Slone

      Do you mean me – or someone else? I can answer for me…

      1. Nancy_Naive Avatar
        Nancy_Naive

        No, not you, John Butcher. Apparently, he’s prescient.

  4. Reed Fawell 3rd Avatar
    Reed Fawell 3rd

    First, thank you Ben Sloan whose commentary in view is far more on target than Jim Bacon’s.

    Second, what really impacted me, and made such a difference in all of this mess, were the beautiful and perfectly aligned teeth, and the amazingly sexy smile of the late middle aged, but aging, male who claimed to be Rector of the University of Virginia, then engaged on Bacon’s Rebellion in a Madison Avenue sales show during what then was then thought to be crisis pandemic. Bacon’s Rebellion should be careful here, not carry water for nonsense, like the Washington Post does daily. We have enough prostitutes walking the streets.

    1. James A. Bacon Avatar
      James A. Bacon

      Reed, sometimes I think you forget that Bacon’s Rebellion (a) encourages people with diverse viewpoints to contribute to the ongoing dialogue on the blog, and (b) that we all need to show humility about the limits of our knowledge. When I wade into a topic like epidemiology, where I have no expertise whatsoever, I try to be careful about making grand pronouncements.

      I also thought UVa Rector Jim Murray made a logical set of arguments about how to go forward. We need not agree with every conclusion in order to take his arguments seriously.

      1. Reed Fawell 3rd Avatar
        Reed Fawell 3rd

        Jim, you didn’t get thrust of my comment. It was not directed at The Rector, but at UVa, and its long sorry history over the past decade. To me, a long time student of that history, the Rector was the metaphor for that sordid history, hence how I “read” the photo and the message, hawking a product.

  5. LarrytheG Avatar
    LarrytheG

    re: ” “Lots of grant $ and computer time to tell us what we already knew.”

    Is that really true? Isn’t the UVA model showing something different than other models like the one Butcher is providing and if this is true then it certainly deserves discussion especially since many folks are looking at these models “peaks”as if once they start down – we can estimate when the outbreak is “over” and that’s not what most epidemiologists are saying and it’s exactly what the UVA model IS showing.

    In a way, it’s almost as if some folks are “shopping” for the models that predict a quick end to the epidemic as opposed to being truly informed by the width and breadth of all the modelling.

    Personal rants against UVA as an institution are not relevant to this discussion IMHO but I have to say – criticism of UVA, even piling on, as an institution is not exactly unheard of here in BR, so to actually give them credit for something is a bit out of character for BR. Usually their alleged sins far exceed their goodt “stuff”! 😉

  6. Reed Fawell 3rd Avatar
    Reed Fawell 3rd

    “That means things like getting better treatments out, better supply of PPEs, more testing, better identification of contacts, and almost certainly some longer term behavioral changes.”

    Again, Virginia stubbornly brings up the rear.

  7. sbostian Avatar
    sbostian

    I have a question, not an assertion. What is the track record of the Biocomplexity Institute? Why should they be believed any more than the discredited Imperial University model?

    1. LarrytheG Avatar
      LarrytheG

      How would you know for anyone providing such info?

  8. sbostian Avatar
    sbostian

    I was just asking a question. If we are to accept the forecasts (which Governor is using to make policy), their track record in providing similar models and forecasts should be disclosed to the public. We have seen that almost all the models used by governments in the US vastly overestimate hospitalizations due to coronavirus and all of those modes have already been adjusted to take into account “social distancing”. If the Biocomplexity Institute has no better track record than Imperial and University of Washington models, then my question is “why should we grant them any credibility?” Otherwise, people who are being injured by the shutdown are being asked to buy another pig in a poke.

    1. LarrytheG Avatar
      LarrytheG

      Are all of them wrong and there are none that are correct?

      My understanding is that most all of them were based on an assumption of 50% compliance with social distancing.

      not true?

    2. virginiagal2 Avatar
      virginiagal2

      The Imperial and Washington models have reasonably good track records. The problem is understanding what they are and are not showing you.

      The analogy I like is cash flow analysis. If I come in and show you that at your current rate, you will run out of cash in two months, and you lay off people and are able to reduce your burn rate, that doesn’t mean the analysis was wrong. It means that the underlying assumptions have changed.

      If you actually read some of the Imperial papers that becomes obvious. The caveats are pretty clearly noted. See for example https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf

  9. sbostian Avatar
    sbostian

    Read the papers. The one you cite is a preliminary release to substantiate their dramatic reductions in lethality a few weeks later. If you look at Ferguson’s track record in past epidemics, he systematically overestimates the morbidity and mortality of almost every epidemic he has modeled. Additionally, he allowed his “worst case” forecasts to be flogged by media and politicians as if they were “expected case” forecasts. I have been a statistician for almost 59 years and would accuse any scientist of an ethical lapse if they allow worst case scenarios to be presented as most likely outcome.

Leave a Reply

Read More