Local Covid-19 Report and Infection Modeling

Local Covid-19 Report and Infection Modeling 1

Even though Lambert and I are not keen about videos as a method of getting information (transcripts are faster!), we thought we’d pass along a reader’s video account as a way of eliciting more reports on what is happening on the ground, particularly near hot spots.

From Kurt:

As a recent (2 years) reader I much enjoy hearing everyone’s perspective, especially localized views like the readers in the UK, so I thought they might like to see how an average American town is handling things. Obviously, we’re worried here in the shadow of NYC, but in some ways, it’s not stopping things? Which is a bit odd. What I found most interesting was how many families I saw out and about (within reason, people did try and stick to some distancing, but it was hard at intersections or when the sidewalks got narrow).

Consider the popularity of the center of town (and in particular the grocery store) in light of this very informative toy modeling exercise (hat tip Monty). This is a terrific demonstration of how various interventions, and cheating around those interventions, affects disease propagation. Per Kurt’s concerns about people still going to his main street and grocery store, at around 16:56, the modeling exercise below shows how having people go to a central location, like a grocery story, significantly undercuts the benefit of social distancing.

Note that early on, the video also demonstrates that quarantining, as in sending the infected to a separate location, is very effective in tamping down the spread of infection. However, Bloomberg reports today that what most of the world is doing, quarantining at home, is another kettle of fish:

Italy needs to shift to mass quarantining of coronavirus patients with mild symptoms instead of letting them isolate at home, according to a group of Chinese experts who traveled to the European nation to advise officials there.

Doctors in Wuhan made the same error early on in the outbreak, said Liang Zong’An, head of the respiratory department at the West China Hospital at Sichuan University. While seriously ill patients were admitted to hospitals, doctors at the time recommended that those with mild symptoms isolate themselves at home, in part to reduce the strain on Wuhan’s overburdened health care system.

Back then, it was not well understood how infectious the virus can be even in those who don’t seem very sick. But researchers now know that those with mild symptoms who are told to stay at home usually risked passing the virus to family members, as well as to others outside their homes as some still moved around freely….

Liang said his team advised Italy to follow China’s lead to forcibly isolate patients with mild symptoms from their families. In China, a study of one province showed that 80% of cluster infections originated from people told to rest at home, according to Xiao Ning, a researcher from the Chinese Center for Disease Control and Prevention who was part of the team.

Finally, to round out this odd assortment of efforts to understand how behaviors affects disease spread, readers (and Lambert and yours truly) have been consistently lamenting that the data on the disease is terrible due to limited testing and inconsistencies in classification, most notably China not counting positive test results as confirmed cases unless the patient also presented symptoms. Even South Korea, widely viewed as having engaged in extensive testing, as of about a week ago had administered 338,000 tests in a country with a population of over 50 million, meaning fewer than 1% were tested. So in the absence of more random testing of the population, we are really flying blind.

Readers suggested working back from deaths. The wee problem is that even thought that data is probably less bad than the confirmed cases tally, that does not mean it is all that good. For instance, readers have said one reason the Covid-19 death rate in Germany is low is in part because if there is a co-morbidity, like COPD, hospitals are classifying that as the cause of death and not the virus.

Similarly, in the UK, deaths understated because people who died at home are not counted even though some almost certainly died of Covid-19:

At least 40 more people have died from coronavirus in the UK – as hidden deaths at home were today revealed for the first time.

New figures reveal 210 people died in England and Wales from the killer bug up to March 20 – 23 per cent more than official NHS numbers have shown.

Yesterday the Department of Health reported 1,408 deaths in the UK but these numbers only include those who have died in NHS hospitals.

If the 23 per cent increase was applied to yesterday’s hospital-only total of 1,408, it would result in a total of 1,732 deaths.

The new ONS figures which include non-hospital deaths only go up to March 20 – three days before strict lockdown measures were imposed by Boris Johnson to curb the spread of the disease.

Unlike the NHS figures, which are limited to those who died in hospital after testing positive for the disease, these wrap in deaths where Covid-19 is mentioned as a suspected cause of death where someone has not necessarily tested positive for the disease.


In New York City, about 1 million people are over 65, and roughly 1/3 of them live alone. Particularly with the hospitals so overwhelmed, any who died at home would not have an autopsy done to determine if Covid-19 were the cause. Readers can pipe up, but it would seem unlike that any investigation would be made of seniors who had died outside a hospital. Unless they had tested positive for Covid-19, their deaths would be unlikely to be attributed to it even if that were a reasonable surmise.

Print Friendly, PDF & Email

Leave a Reply

Your email address will not be published.