Here comes another bit, expanding on the study linked:
I don’t know if the links below are helpful for you, in particular. However, there is a particular expert I listen to (podcast) regularly, and who is really an expert on this type of virus.
The scripts of this podcast are available, in German. You can search them for keywords, and use a translation tool to get an idea. (However, please be aware that context is really, really important.)
The particular study linked above was explained, in laypeople’s terms, here:
Google Translate of some relevant bits on the study, I only very quickly corrected some obvious problems:
[T]his is a study from the group of Christophe Fraser, certainly one of the best epidemiological modelers. It’s a very interesting study, I think. It is published in “Science”. The point here is to first calculate a much better, more accurate epidemiological model that is simply much more fine-grained, where more information goes in than was previously known. Scientific literature provides now more and more data that can also be evaluated and then fed into such models. The start of this study is actually the observation that we have more and more descriptions of transmission pairs in the literature and can therefore actually better determine the generation time of this infection. So: how long does it take from symptom to symptom or from infection to infection? Regarding “symptom to symptom” we speak of generation time, regarding the other - “infection to infection” - we speak about series length. And what you actually need is the series length. But it’s all relatively difficult to pinpoint that. Therefore, we can at least try to make a good approximation of the generation time. We can derive this, from literature reports, and that’s how it starts here. 40 transmission pairs from the literature are evaluated, so that a mathematical model that already exists is fed in order to derive certain parameters and certain proportions of the entire transmission activity. [We get] the transmission number R0, which has been recalculated in this study with two. This is a relatively low value when you look at what other analyzes have found before. [Other calculations were] sometimes around two and a half.
Interviewer: So, one person would infect two [and a half] other people.
Exactly. Now of course we have the option to break down these transfers to shares. And the question that is first asked of this mathematical model is: What is the proportion of presymptomatic transmissions, of symptomatic transmissions, but also of environmental transmissions and asymptomatic transmissions? So asymptomatic means a carrier who never gets symptoms himself. And presymptomatic means, of course, that it is transmitted before the transmitter has symptoms; But you can still find this transmitter later, because [the infection] will cause symptoms. Of course, you can still identify the contact patients later. This is a consideration that will be discussed later in the release. But first these values that come out: Presymptomatic 0.9, i.e. a share of 0.9 in the two (i.e. in the R0 value of two), symptomatic transmission has a share of 0.8, and then environmental transmission has 0.2, and asymptomatic transmission 0.1. If you add these four values together, you get two again. If you now visualize the numbers, you will come to the conclusion that the presymptomatic transmission share is 46 percent of the total transmission activity. It is a value that we discussed a few days ago from another working group, from another paper.
Interviewer: So the infected person transmits the virus before he is even sick and does not notice it.
Exactly right. So apparently almost half of the transmission activity is carried out before the [onset of the] symptoms. These are average values, that is: averaged over many transmitters and then analyzed in a mathematical model. This is noteworthy, because we have two things to think about now. On the one hand, the value R0 of two actually seems to be good news. Because if we have an R0 out of two, then we have less of what we have to reduce in the number of transmissions in order to lower the R0 below one and thus also bring the epidemic to a standstill.
However, if you now realise that 46 percent of all this transmissions take place before the symptoms, it will of course be very difficult to reduce these transfers again: [b]ecause you can actually only isolate symptomatic patients. These considerations are now fed into an interesting calculation that wants to find out: What can you actually do with certain interventions to recognise an infected person? How long does it take to recognise it? And how many has the infected infected during this time anyway, because 46 percent of the transmission happens before the onset of symptoms? And because it also takes a while before a diagnosis is made after the onset of symptoms and then the contacts can also be identified. A very important requirement plays a role here, namely the generation time of the infection, which is now recalculated here, which actually tells us: Even if you isolate immediately at the onset of symptoms, i.e. immediately remove a symptomatic [patient] from the transmission situation, then they not only will already have infected others, but these infected people are already infectious again [themselves] at the time when the symptoms begin in the first patient.
Interviewer: And maybe [other] people have been infected.
[They] are just starting with it, so they arrive at the beginning of their infectivity. And they’re just starting to infect people. We actually have observed something like this in the Munich case tracking study and wondered about it. But now there is in principle a quantitative proof that really backs the whole thing up with numbers and rates that this actually happens. Now this study makes a very interesting calculation about the possibilities of the intervention. What actually comes out of the study is that we are [already] late [when] simply identifying cases and tracking contacts, because the whole thing simply depends on recognising symptomatic patients. So it really comes down to the last day. It really matters that a symptomatic patient should not be symptomatic for a longer time before being tested. [Only after] the test the reporting chain runs. Then the health department has to come and ask: Who were you in contact with? Then these contacts will be identified at some point. This all takes so long that according to the latest calculations that most of the time it would take to get the patient out of the transmission events has long been lost. In other words, here it is, calculated in a formally very correct manner, very robustly, and based on the very latest figures: given a certain point in time of the epidemic onwards, targeted diagnostics plus case tracking plus isolation of the contacts [cannot stop this epidemic]. That’s not possible anymore. What you can do to stop an epidemic like this is to just do a lockdown. Then you no longer have to follow any cases, then everyone is at home. You can, of course, take a combination of measures where you [put in place] a lockdown that is a bit milder. Which includes something like a ban on meetings.
Interviewer: What we have at the moment in Germany.
Exactly, plus case tracking measures [that are taken in Germany]. But there are also calculations in here that say there are certain efficiencies [to be considered]. You can reduce transmission rates by a certain factor. But [all] the measures always have only limited efficiency, and [due to] this limitation of efficiencies, [they] can also [be accounted for] in [their model]. [However, e]ven then one comes to the conclusion that this combination of measures cannot actually stop [the epidemic]. Then something else is done [here], something else is included. And that would be [the development of the epidemic] you get when you use such an app [for contact tracing which is suggested as a hypothetical model in this paper].
I really suggest you have a go at the rest, sorry that I can’t do further gTranslate proofreading right now.
Bottom line: epidemiologists have been informing British and US gouvernment responses. It is argued that the UK is in lockdown because of this particular study. And the US followed, in part, their example.
If you want to push R0 below 1, a lockdown is effective. The lockdown, in combination with other measures, can vary in intensity (see: Germany), and still reach the goal. But given the proportion of pre-symptomatic infections to this R0, we should stay socially distanced as much as possible not to have rising numbers again, which *will turn into logistic growth if no hard measures are taken.
So better not start discussing “hard” numbers like “R0 below 1 for a week”. Just don’t.
I, for one, am waiting for science to find out more. And then I am willing to listen to the experts. As a biologist, I tend towards listening to scientists more than to, say, economists. But this is a political decision. And if you can say:
then you can surely appreciate that right now, information we have indicates that we do not have enough information, or options, to get out of this mess without collapsing your societies health system. Which will kill a very high number of people.
That’s not about a categorical “we fight until COVID-19 is gone”.
One more thing: science informs us that SARS-CoV-2 is very likely indeed here to stay. It is a human Coronavirus now. The question with that is: when will we have an effective vaccine?
