There is a problem with modeling behind the blockade of the extended city of Melbourne

I fully support the goal of eliminating the coronavirus from Victoria and at the same time hope to eliminate it from across Australia.

I wrote A book Making a case is the best way to bring Australia back to normal, given the uncertainty of the vaccine timeline and the ongoing difficulty of managing the pandemic.

However, a modeling study used by the Government of Victoria to justify extending the extension of Melbourne’s Stage 4 blockade for another two weeks suggests that the flaw may have resulted. ..

Different, more traditional models offer to relax more detailed location-based (eg, local government or postal code groups) restrictions and achieve the same result with less economic and social costs.

There’s no doubt that the Stage 4 lockdown, which was introduced at 6 pm on August 2nd, did a great job.

In the case of facing upwards, it usually takes two weeks or more for the measures to take effect. But the decline in Victoria was dramatic.

That doesn’t mean that costs aren’t an issue. Delays specific to Extension and resumption plans It’s quite It works like this:

step 1 The one announced on Sunday extends the Stage 4 limit by an additional two weeks, but with some additional freedom. These include:

• Move start of curfew from 8pm to 9pm
• Exercise for 1 to 2 hours
• Allow outdoor gatherings for two or one household
• Creating a “social bubble” for single people living alone and for single parents with children under the age of 18
• Resume playground

Steps 2, 3, and 4 are all subject to health advice and rely entirely on a decreasing number of new cases each day.

If the new case meets the required threshold, Step two Starting September 28th, we allow the following:

• A public gathering of up to 5 people from a maximum of 2 households (up to 2 hours, or within 5 km of your home if you are in the metropolitan area of ​​Melbourne)
• Return some students to school gradually and resume childcare
• More workplaces can be resumed
• Outdoor pool can resume personal training sessions with up to 2 clients allowed
• An outdoor religious gathering where 5 people and leaders can participate.

Step 3 After October 26th it will be as follows.

• The curfew has been abolished and there are no restrictions on reasons or distance to leave the house
• Up to 10 people can gather outdoors and one named household can create a “household bubble” that allows up to 5 visitors from that household at one time.
• Further progress in grades 3-10
• Hairdressing, retail and hospitality can be conditionally resumed
• A gradual return of adults to outdoor contactless sports (outdoor contact sports under 18 are permitted).

And from November 23rd, according to all the necessary requirements, Step 4 It contains:

• Up to 50 people can gather in public and up to 20 visitors can gather at home.
• Limited reopening hospitality, reopening retail and real estate
• 50 at weddings and funerals (20 at private residences)
• And then back to community sports.

The Step 2 trigger is an average of less than 50 new daily cases over 2 weeks, while the Step 3 trigger is an average of 5 new daily cases over 2 weeks and zero mystery cases over the entire 2 weeks. For step 4, the trigger is 14 days statewide and there are no new cases.

There are some good progressives in this plan. The school is relatively early and opens before the pub. The playground is quickly opened and the allowance for social bubbles is being created. And the big gathering is the last.

The question is, why is the pace of resumption slow?

One of the reasons is that the government is trying to avoid disappointment for the extended ones. But playing those games seems like a second order to provide clarity. The plan seems to be slow as it relies on some modeling results.

Let’s take a look at that modeling.

Victorian model

The model used by the Victorian government is Australian Medical Journal.. It has been peer reviewed. However, peer reviews only tell us that the model is accurate for what it claims to do.

The model is an “agent-based” epidemiological model.

So, unlike the standard,Customer“Models that use numbers as inputs Easily infected, infected, recovered Explicitly lists the equations that describe individuals and behaviors and the flow of information. This is a computer simulation based on agent interaction.

When the agents run randomly against each other, the simulation runs repeatedly, observing the progress of the pandemic. This is a convenient approach, but it relies heavily on important assumptions. The agents spread the virus by interacting with their neighbors, but the geographic distribution of the agents is fairly smooth (to make these interactions computable).

This means that such models do not divide the population into populations, and as a result, if any virus is present anywhere, we predict that it will eventually occur anywhere. They conclude that the best way to prevent the virus from reaching everywhere is to eliminate the source of infection, the movement of people.

Not surprisingly, Victoria decided to do so.

It used a well-tuned model, but was based on moving around people, and decided to stop moving around people.

Is it the right model?

Recall that the premise of the agent-based model is that people interact with their neighbors and are probabilistic, but fairly smoothly linked.

This is a map of outbreak patterns across Melbourne, where the strongest blockades are taking place.

This is the current pattern, which I’ve seen all the time, but the same all the time.

This pattern suggests that people interact more intensively within their local area than they do to generate the same probability of transmission throughout the city.

It also shows that if you have a strict lockdown and you need the resources to do that work, there are places where it is more important to place resources than others.

It’s worth observing some other things.

Checking Google Trends data for common COVID-19 symptoms such as: Anosmia, It shows that people have been gurgling this semester at a fairly steady rate since April.

Hopefully, that means the government doesn’t have many people missing in the test. (A surge in searches for Google’s common symptoms when new cases didn’t seem to have surged could indicate an undiscovered infection).

Summer has finally arrived. If allowed, people go out more, and because we know, Coronavirus, It Significantly reduce spread..

What Should Victoria Do?

The government has published other modeling they have done so far model I’m using.

Not doing this is too economical for the additional months of lockdown.

It is important to consider network patterns, that is, how people move in cities and social groups.

Second, the government can reopen either on a zip code basis or on a local government area basis. In this way, the government can monitor for outbreaks in the less populated areas that initially resume and use them to signal the pace of resumption.

You can update the limit every two weeks with real-time information.

I live in Toronto, Canada, with a footprint similar to Melbourne, and when I reopened I didn’t treat the whole area as one. The goal pushed in Toronto is smaller than the elimination, but it worked pretty well.

Making Melbourne the whole trigger unit seems incompatible with the nature of its occurrence.

The most defensive cases are based on the idea that people regularly travel long distances throughout Greater Melbourne. In the middle case, monitoring location-specific lockdowns is more difficult than monitoring city-wide lockdowns.

The least defensive case is based on the notion of fairness.

Third, the government should encourage people to go out as much as possible. There is no obligation to mask outdoors. A more relaxed approach to outdoor gatherings makes the task of implementing important directives much easier.

Google Mobility Report Go out anyway..

Finally, I can’t stress this enough and test and trace. This is an updated theme for My book Covers how to do it in a pandemic Newsletter..

Lockdown alone does not reduce infection to zero. However, when the number of cases is small, active identification and isolation of infected persons is performed.

Victoria is in a remarkable distance of zero infection while avoiding financial distress.

It doubles and you have to do it.

---

---

This article is reissued from conversation Under Creative Commons license. Read Original work..