Sun. Sep 25th, 2022
Image of a restaurant with a large sign that says
enlarge In response to an increase in cases, Germany has ordered restaurants to switch to delivery/takeout only.

Nobody wants to go back under heavy social restrictions. But the rising number of cases is causing many countries to implement targeted lockdowns and other limits to try to get the pandemic back under control — a move that has sparked backlash in many places. So it seems worth asking what the optimal combination of constraints might be. How do you get the most pandemic containment for the least restrictive social environment?

That’s exactly what an international team of researchers tried to find out, as described in an article published today. And while the researchers come up with some possible answers, their paper ends with an additional message: This is a very difficult question to answer. So to some extent, many countries will have to trade with imperfect information and hope for the best.

How do you answer that?

In an ideal world, we would have some idea of ​​the impact of every possible social restriction: closing restaurants, starting contact tracing, closing schools, and so on. With that information, we can look at the number of infections and the trajectory of the virus, and then come up with the smallest possible set of constraints that could lower the infection rate. But the real world is currently very far from this idealized situation, which motivated the researchers to try to provide a little more certainty about the effectiveness of various restrictions.

But the real world also makes that a difficult question to answer. After all, no one has assumed even one limitation; In general, a series of limits were introduced in most countries, with additional restrictions gradually being added. In addition, different cultures may find it more difficult to comply with bar and restaurant limits than wearing a mask mandate – different limits will entail different levels of compliance.

To get around this, the researchers collected large collections of public health policies that had been put in place during the first wave of infections in March and April. They then analyzed them all using four different methods and looked for the ones that were consistently associated with better results in terms of the total number of infections.

This turned out to be a lot of information. The researchers’ first analysis was a list of more than 6,000 different policy changes made in 79 different countries, states or provinces. But then they used two additional disability data sources to validate the first analysis, which involved a total of 42,000 additional interventions.

The correlations between these interventions and infection rates were then tested using four different mathematical approaches. (These were a case-control matching algorithm, two different types of regression analysis, and transformer modeling.) Approaches that yielded significant results in three or four of these tests were considered likely to be effective.

Not so fast

Unfortunately, one of the limitations of this analysis became apparent quite early on. During the first wave of the pandemic, testing capacity was generally limited, particularly in the United States. As such, the number of cases identified often correlated more with testing capacity than with the presence of the virus. Not surprisingly, the algorithms typically chose policies like “improving testing capacity” and “increased surveillance” as having a negative effect on the management of the pandemic. The reality is that in most countries, these policies have likely led to more case identification.

Aside from that kind of limitation, there was a high degree of correlation between policies rated as effective by different algorithms, which should be a reason for confidence in other results. Some of these are exactly the things people don’t like: national lockdowns and stay at home. Banning small gatherings emerged as the best intervention. Another top move, increasing the supply of personal protective equipment, is not really a surprise. The same applies to the fact that information about the virus and the pandemic appeared to have a positive effect in three of the four analyses.

Other things were somewhat surprising. Border restrictions have been shown to be quite effective, although that may be biased by some island states, such as Taiwan and New Zealand, which have had great success in controlling the virus’s entry. Another consensus choice was school closure. Some previous studies had suggested the spread in schools was minimal, while others had supported the measure, so the evidence here was ambiguous.

However, when the research team turned to the larger database of policy interventions, there was no full agreement with the initial analysis. Of the seven that rated all four analyzes as effective, only three saw any degree of agreement when the same four analyzes were repeated in the larger database. But again, social restrictions like ending small or large gatherings and closing offices looked good, as did border closures. So, while a specific policy may not have been replicated, closely related policies often did, suggesting that we would get clearer results by pooling similar constraints in one group.

A further indication of this could be the fact that national lockdowns were only rated as inconsistent. The researchers suspect this was caused by lockdowns that came on the heels of some previous restrictions, and so lockdowns haven’t added as much as they could if implemented individually.

One policy that scored particularly poorly across all data sources was the widespread disinfection of the environment. It is not clear why that was the case. Another policy that didn’t seem to do much was social distancing measures specifically targeting public transport.

General messages

Another thing the research team checked is what they call the “entropy” of the results. That is, if a policy is universally effective, wherever it is attempted, it will have low entropy. From this we can deduce that cultural differences probably did not have a major influence on the success of the policy. It turns out that some of the more effective measures, such as social distancing and travel restrictions, were high entropy, meaning they worked in some countries but not so well in others. You just need to look at the different levels of restriction compliance in the United States to understand how this could happen.

Yet local culture seemed to influence the effectiveness of the policy, not whether it was effective at all. Different forms of social distancing and restrictions on meetings, businesses and restaurants scored consistently well in most analyses.

It was also significant that better communication about the rationale for policy often had just as strong an effect as the implementation of the policy itself. If you explain to people why you might have to close restaurants, chances are a lot of people will stop going to restaurants, whether the rules apply or not.

Yet the information is too complex for the ideal goal we would like to see: identifying the minimum level of restrictions that can keep the pandemic under control. As hospitals fill up, we will likely have to implement severe social restrictions to ensure the health care system is not overloaded. However, one thing the study does offer is another piece of evidence that those limitations will do the job.

Nature Human Behavior, 2020. DOI: 10.1038/s41562-020-01009-0 (About DOIs).

By akfire1

Leave a Reply

Your email address will not be published.