Replies

New post by Sebastian Rushworth M.D.

Can asymptomatic people spread covid-19?
by Sebastian Rushworth, M.D.

If anyone answers your question, it might be from this
medical stat genius: Sebastian Rushworth M.D.

Can asymptomatic people spread covid-19?
by Sebastian Rushworth, M.D.

One of the most contentious issues when it comes to covid-19 is whether or not people who are asymptomatic can spread the disease. The reason is that there are major ramifications depending on the answer. If only those with symptoms spread the disease, then telling people who are completely symptom free to stay at home and avoid social contacts makes zero sense. But if even those without symptoms can spread the disease, then there is at least some theoretical basis for quarantining asymptomatic people, a.k.a. lockdowns.

My personal intuition has been that covid probably does spread asymptomatically. This intuition is based on the fact that the virus has spread so rapidly and successfully around the world, and also on the complete and utter uselessness of lockdowns (I’ll explain my thinking on this further along in the text). This is very diferent from MERS and SARS, which were both rapidly and easily contained, probably because there was no asymptomatic spread, thus making it easy to identify and quarantine those who were infectious.

Back in November I wrote an article about a study conducted on US marine corps recruits. These recruits were forced to quarantine at home for two weeks before beginning their training, and were only allowed to begin the training if they had been symptom free during that entire two week period. In spite of this, the virus was able to sneak in to the training camp and spread among the recruits. Of the recruits that were determined to have brought the virus in to the camp, most remained symptom free throughout. To me, this in itself is pretty powerful evidence that asymptomatic spread does happen. But of course, 18 year olds are not known for being reliable - maybe they were so keen to start their military careers that they went to the training camp in spite of a sore throat and head ache, and just pretended to be asymptomatic.

Since then, however, more data has been published, which can perhaps help to answer the question more definitively. And to be fair, we don’t just want to know if asymptomatic spread can happen, we also want to know to what extent it happens in reality, and what the difference in infectiousness is between people who are symptomatic or pre-symptomatic and those who stay asymptomatic.

If people with symptoms are 100 times more infectious than people without symptoms, then that is for all practical intents and purposes the same as if people without symptoms are not infectious, because it means that it should be enough to contain only those with symptoms in order to effectively stop the spread of the disease. If on the other hand they’re only slightly more infectious, then efforts to contain the disease by quarantining only those with symptoms are unlikely to be successful.

An article was published in The Lancet in January that sought to establish the relative infectiousness of people with symptomatic as opposed to asymptomatic disease. It used data from Singapore’s screening and track and trace programs. Singapore screens large segments of the population for covid-19 on a regular basis. When someone tests positive for covid through the screening program, all recent contacts that have spent more than half an hour within two meters of that person are also tested. Both the “index case” and the contacts are then placed in quarantine until they test negative.

This has allowed the Singaporean government to catch a lot of people with asymptomatic infection, and to see the extent to which they are able to pass the infection on to others. This data can then be compared with the data from people who have symptomatic infection and their close contacts. So, what the authors of the study did was to take the data on everyone who spent time in quarantine as a result of being a close contact of someone with covid between August 1st and October 11th 2020, and compare the relative rates of secondary infection among those exposed to someone with symptomatic covid and those exposed to someone with asymptomatic covid, as diagnosed by PCR.

A total of 628 index cases and 3,790 close contacts were identifed (roughly six close contacts per index case). 5.2% of the contacts of symptomatic index cases developed covid-19 (i.e. switched from having a negative to a positive PCR test) during their quarantine period, as compared with 1.6% of asymptomatic index cases. On the face of it, this would suggest that people with symptomatic disease are roughly three times more infectious than those with asymptomatic disease.

Of course, there is a factor that makes this result more difficult to interpret, and that is the false positive rate of the PCR test. Data from real world testing suggests that the false positive rate of the PCR test is well below 1%. In Australia, as few as 0.2% of tests have been positive for much of the pandemic. If we assume that this number is in the ballpark, and subtract 0.2% from the two numbers provided above from the Singaporean data, that would give us 5.0% infected among contacts of symptomatic people, and 1.4% infected among contacts of asymptomatic people, which would still suggest that people with symptomatic infection are roughly three to four times more infectious than people with asymptomatic infection.

One major issue with this study, however, is the fact that all of this is observational data. The contacts haven’t only been in close contact with the known index cases. They’ve interacted with lots of other people too. How do we know that the supposed index case is the person that actually infected them? We don’t. So technically, it could actually be the case that every single one of the people that became infected after spending time with an asymptomatic person was actually infected by someone else who was symptomatic. If that’s the case, then people who are symptomatic are infinitely more infectious than people who are asymptomatic. It’s impossible to know for certain.

In order to be able to figure that part out, you would need a randomized trial, where you keep your study objects in a cage and cut them off from all contact with other people for a month or so, then introduce them to someone with either symptomatic or asymptomatic covid-19 and see if they become infected. Well... I guess it doesn’t have to be a cage...

Since that trial design would be difficult to get approved by an ethical review board, we have to look around for an observational study that accomplishes something similar. Luckily, a Japanese study, recently published in the International Journal of Infectious Diseases, does just that. Like Singapore, Japan has had an extensive track and trace system in place since the beginning of the pandemic. The study describes the first cluster of cases identified in Japan, back in January 2020. At that time there was virtually zero societal spread in Japan, so for anyone who was a close contact of a known index case, and who subsequently developed covid-19, we can be reasonably certain that it was the known index case that infected them.

It began with a tourist from Wuhan. The tourist infected a person who subsequently went to a party in Tokyo on January 18th, at which ten people were infected, four with symptomatic covid, and six with asymptomatic disease (it must have been quite a party!). Of the six asympomatic people, one went on to infect two people. One of these developed symptomatic disease while the other remained asymptomatic. The asymptomatic person then went on to infect two more people, both of whom developed symptomatic disease, one of whom died.

What does this prove? Well, it’s observational data, so it’s hard to say that it proves anything. But considering how little societal spread there was at the time, it seems highly likely that the people identified as having spread the infection to others further down the chain actually were the spreaders. This study therefore provides pretty strong support to the notion that people with asymptomatic disease, who remain asymptomatic (i.e. who aren’t just pre-symptomatic), can spread the disease on to others, who then go on to develop symptoms.

Overall, 36 people were identified in the cluster of infections arising from the tourist from Wuhan. 23 of these were symptomatic, while 12 remained asymptomatic (one had unknown symptom status). Overall, people with symptomatic infection were estimated to be 3.7 times more infectious than people with asymptomatic infection. However, the confidence interval was wide due to the small number of cases, with the lower bound estimate being that people with symptomatic disease are over thirty times more infectious than those with asymptomatic disease, while the upper bound estimate is that people with asymptomatic disease are almost as infectious as those with symptomatic disease.

When taken together, I think these two studies show that people with asymptomatic covid-19 can and do spread the disease, and the best estimate from both studies is that those with symptomatic disease are about three to four times more infectious than those with asymptomatic disease. That is not a huge difference. Even though people with symptoms are a couple of times more infectious than people without symptoms, those without symptoms could still be causing more infections overall.

The reason is that people with symptomatic disease will for the most part stay at home and avoid contact with other people, at least from the moment they start to have symptoms (by which time they could, of course, already have infected plenty of other people). People who are asymptomatic, on the other hand, will continue to lead their lives as normal, for example going to work at a care home or hospital, where they will come in to close physical contact with frail elderly people for extended periods. Considering that the evidence from randomized trials shows that face masks provide little to no benefit, they will then likely spread the infection on regardless of how rigorously they follow mandates to wear face masks at work. And that’s probably why lockdowns haven’t had any noticeable effect on the number of people dying of covid. Telling 30 year old office workers to work from home won’t stop the virus circulating in hospitals and care homes. Those at risk of serious disease will therefore still be exposed and infected.