2

Toward the end of the widely read Medium article The Hammer and the Dance, the author talks about "the ROI [return on investment] of social distancing". He provides a chart and a table showing the costs and benefits of various social distancing measures, such as banning gatherings of a certain size, closing restaurants, closing schools, etc. But he notes that these illustrations are fake and "all the data is made up" because "nobody has done enough research about this or put together all these measures in a way that can compare them."

Is that really true? It would seem strange if these social distancing techniques are (to epidemiologists) apparently a go-to tool for epidemic suppression, and yet there is no real research on their effectiveness.

What I'm specifically curious about is if there are data or models that try to measure the effects of individual social distancing measures. This could be either for diseases in general, for specific classes of diseases (e.g., viruses spread via respiratory droplets), or for specific diseases (e.g., influenza). Obviously there won't be info for this specific virus due to its novelty but is there anything that would serve as a reasonable starting point?

I'm curious because most places around the world that have implemented some sort of lockdown appear to have taken a "do it all" approach, activating as many kinds of social distancing as they can. This makes sense as an initial emergency response to hit the brakes as hard as possible. However, as different jurisdictions consider lifting some of these restrictions, I don't see a lot in the press about decisions being made based on their relative efficacy. The discussion I see is mostly about things like increasing testing and monitoring the curve of cases to see when it's okay to lift restrictions, but not about how to decide which restrictions to lift. There are some general statements like "large sports events probably won't come back for a long time" but even those aren't really explained. I haven't seen anything like "banning gatherings of more than 10 people will slow the spread X% relative to banning gatherings of more than 50 people" or "reopening schools will be less damaging than reopening restaurants". Is there anything like that?

2

Modelling has certainly been done but they operate under assumptions that may not be true.

For example the one done for Singapore:

Findings For the baseline scenario, when R0 was 1·5, the median cumulative number of infections at day 80 was 279 000 (IQR 245 000–320 000), corresponding to 7·4% (IQR 6·5–8·5) of the resident population of Singapore. The median number of infections increased with higher infectivity: 727 000 cases (670 000–776 000) when R0 was 2·0, corresponding to 19·3% (17·8–20·6) of the Singaporean population, and 1 207 000 cases (1 164 000–1 249 000) when R0 was 2·5, corresponding to 32% (30·9–33·1) of the Singaporean population. Compared with the baseline scenario, the combined intervention was the most effective, reducing the estimated median number of infections by 99·3% (IQR 92·6–99·9) when R0 was 1·5, by 93·0% (81·5–99·7) when R0 was 2·0, and by 78·2% (59·0 −94·4) when R0 was 2·5. Assuming increasing asymptomatic fractions up to 50·0%, up to 277 000 infections were estimated to occur at day 80 with the combined intervention relative to 1800 for the baseline at R0 of 1·5.

Interpretation Implementing the combined intervention of quarantining infected individuals and their family members, workplace distancing, and school closure once community transmission has been detected could substantially reduce the number of SARS-CoV-2 infections. We therefore recommend immediate deployment of this strategy if local secondary transmission is confirmed within Singapore. However, quarantine and workplace distancing should be prioritised over school closure because at this early stage, symptomatic children have higher withdrawal rates from school than do symptomatic adults from work. At higher asymptomatic proportions, intervention effectiveness might be substantially reduced requiring the need for effective case management and treatments, and preventive measures such as vaccines.

but this model is based on an influenza model which is a droplet spread infection but there's increasing suspicion that SARS-CoV-2 can also spread by nuclear droplets (like measles and tuberculosis). And their value of R0 goes up to 2.5 but others are now saying that the true R0 is closer to 5.7. If these are the case, then the social distancing model falls apart as we can't get apart far enough.

Interventions to mitigate early spread of SARS-CoV-2 in Singapore: a modelling study https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30162-6/fulltext

High Contagiousness and Rapid Spread of Severe Acute Respiratory Syndrome Coronavirus 2 https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article

1

3Blue1Brown made a great video (Simulating an Epidemic) showcasing the effects of tweaking various parameters on outbreaks using simulations.

Background:

His simulation is based on the SIR model, which categorizes a population into people Susceptible to the disease, people Infected with the disease, and people Recovered from the disease. For every unit of time a susceptible person spends within the infectious radius of an infected person, the suspectible person will have some probability of contracting the disease. After a certain number of time, an infected person will recover (or be removed) and will not be able to spread the disease.

Findings: The 5 main takeaways from the video are

  1. The growth rate of new cases is sensitive to # of daily interactions, probability of infection (e.g. better hygiene) and duration of infection [4:41]
  2. Changes in how many people are tested (and quarantined) cause disproportionately large changes to the total number of people infected [9:06]
  3. Social distancing slows the spread of infection, but even small imperfections prolongs it [12:54]
  4. Reducing contact between communities, late in an outbreak, has a limited effect [14:00]
  5. Shared central locations (e.g. grocery store) dramatically speed up the spread [17:42]

Disclaimer:

3Blue1Brown notes that his simulations are toy models and uses a small population-size, so his findings are not necessarily generalizable to real-world examples.

Here is a fan-made interactive simulation based on the same video: https://prajwalsouza.github.io/Experiments/Epidemic-Simulation.html

4
  • Giving us a summary of the youtube video would greatly improve your answer. Like many here (probably most), I'm not willing to sit through a 23 minute video to see your evidence. – Carey Gregory Apr 24 '20 at 0:42
  • 1
    @CareyGregory Thanks for the feedback. I don't think I can do his simulations justice on a short written answer, but I've added a better summary of what the video covers. – D.Tan Apr 24 '20 at 1:44
  • Thanks, I've seen this and other similar stuff like this. However, as you say these are self-acknowledged "toy" simulations. I'm interested in actual research. It could be a simulation, but it has to at least be a simulation that the authors are going to stand behind as something that's relevant to real-world epidemics. – BrenBarn Apr 24 '20 at 3:54
  • I agree that this isn’t peer-reviewed research. I figured this answer would be valuable because we can still learn important lessons from simple simulations and can be a good starting ground for finding more indepth work on the topic. – D.Tan Apr 25 '20 at 2:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.