Thread by @peripatetical

SARS-CoV-2 infection in patients that don't have symptoms. It's complicated. A thread.

Data suggest that viral load does start to increase and is maximal slightly before the onset of symptoms, although there is clearly a lot of variation between patients.
nature.com/articles/s4159…

Although it can be hard to pinpoint the exact time of transmission, it also appears that transmission can occur from people with infection before the onset of symptoms.

This was suggested by early studies that found that the serial interval (the time between the onset of symptoms in cases in a transmission chain) is shorter that the incubation period (the time between exposure and the onset of symptoms) papers.ssrn.com/sol3/papers.cf…

Cluster investigations provided some direct evidence of this. There appeared to be some doubt about when symptoms developed in an index case in a early German cluster
nejm.org/doi/full/10.10…

Many studies have now more convincingly documented pre-symptomatic transmission
ncbi.nlm.nih.gov/pmc/articles/P…
jamanetwork.com/journals/jamai…
cdc.gov/mmwr/volumes/6…

It's important to note that public health interventions are mostly targeted at symptomatic people, so symptomatic people may have a higher transmission potential, but public health actions may reduce their infectivity.

One issue is the definition of "asymptomatic". Some of the patients I saw from the Diamond Princess reported that they had trouble communicating with some of the Japanese doctors, and their assessments were mostly a temperature check.

In practice, it can be difficult to tell when symptoms develop. We've had patients with a chronic cough that just happened to get a test for unclear reasons. Some patients have described gastroenteritis, then a fever a week later.

Loss of smell is now fairly well established as a symptom, but may not have been recognised earlier on
jamanetwork.com/journals/jama/…

The other issue is about asymptomatic vs pre-symptomatic infection. This is important, because if completely asymptomatic people transmit, then there would be missing links in transmission chains.

Pre-symptomatic transmission is still important, but these cases are more likely to be detected eventually, so the transmission chain will be eventually filled in.

Cross sectional studies that had no follow up will therefore conflate cases who never develop symptoms (a minority) and those who go on to develop symptoms.

The cases in the King County outbreak provide good data on this - of the 76 residents tested, 23 had SARS2 detected. Of these, 13 had no symptoms at the time of testing, but 10 developed symptoms subsequently.
ncbi.nlm.nih.gov/pmc/articles/P…

We are also seeing patients that may have had infection some time ago with positive tests, who probably are no longer infectious but may be shedding viral fragments. And at the moment with so few true cases, we may be seeing a few false positives

@PaulGlasziou and colleagues showed that studies that followed up patients after diagnosis had a lower proportion that were asymptomatic. They estimated that around 16% of infections were asymptomatic, and this was lower in aged care settings.
papers.ssrn.com/sol3/papers.cf…

Asymptomatic cases may have a lower viral load, or may not spread virus as far if they are not coughing. But asymptomatic cases may continue with their usual activities and have more contacts.

Complicating this further is the possibility that "negative serial intervals" may occur - that is, the person infected may become symptomatic before the person that infected them.

And, if you detect a close contact who has asymptomatic infection, were they the upstream source, or the downstream case?

So, what are the implications of all this? First, quarantine of close contacts and travellers from high risk settings is important - it is really a form of pre-emptive isolation

Second, we should test close contacts to ensure they aren't asymptomatic or pre-symptomatic cases, particularly in sensitive settings like aged care and healthcare facilities.

Finally, when we see positive PCRs in asymptomatic patients, we need to assess whether they are asymptomatic, pre-symptomatic, post-symptomatic or false positive tests.