People are asking whether patients with COVID19 who don’t have symptoms, e.g. cough or even just fever, can transmit the disease to others. The answer seems to be yes. Alas, this is not good news, but we can still take rational steps. Let’s talk about this. #QuarantineLoophole 1/
Should you avoid getting near symptomatic relatives? Yes, alas. What about college students or travelers returning home who aren't showing symptoms? That’s harder. If they could have been exposed, you should practice social distancing for a while, even if they're asymptomatic. 2/
If people are asymptomatic but have credibly been exposed to COVID19, they should be in home quarantine for 14 days, to be safe. But for how long can someone be asymptomatic and still transmit? It appears that this is on the order of 2-4 days. 3/
Let’s do some epidemiology. 4/
The period between exposure and showing signs or symptoms of the infection is called the *incubation period*. The period between exposure and becoming infectious (able to spread the disease to others) is called the *latent period*. | Image modified from @marcelsalathe 5/ 
Let’s call difference between latent period & incubation period the *mismatch* period (latent minus incubation), which can be negative (eg, HIV: asymptomatic carriers can be infectious) or positive (eg, smallpox: patients are usualy symptomatic before they're infectious). 6/
This difference (mismatch period) between the incubation and latent periods has also been called the “omega” period in some veterinary studies, or it can be called the "period of subclinical infectiousness.” 7/
From point of view of a pathogen, it might be good to evolve to be transmissible *without* incapacitating the host or to evolve to elicit caring behavior by other humans, so as to infect them (a subject for another day, discussed in my book Blueprint amazon.com/Blueprint-Evol… ). 8/
The latent period is partly related to disease dynamics (incl social network interactions & even diurnal rhythms, since they determine the time point from which the pathogen can be transmitted from infected to healthy hosts). Regarding diurnal rhythms: bmcinfectdis.biomedcentral.com/track/pdf/10.1… 9/
In contrast, the incubation period of a pathogen is primarily related to the biology of the interactions between hosts and pathogens (for instance how quickly the host mounts an immune response to clear the pathogen). 10/
When the latent period is *shorter* than the incubation period (as shown in the big red arrow in the figure), this can present a challenge to epidemic management since asymptomatic patients can spread the disease. 11/
If most transmission occurs before disease is apparent (eg, HIV), *reactive* control measures will be ineffective. Conversely, successful disease eradication (eg, smallpox & SARS in 2002 — cdc.gov/about/history/… ) is facilitated by low transmission by asymptomatic people. 12/
That is, for diseases with an incubation period shorter than the latent period, detection and thus quarantine of symptomatic hosts is possible before they become a source of pathogen, leading to better disease control. 13/
The success of reactive disease control strategies has been shown to depend on the timing of the onset of infectiousness relative to the onset of detectable clinical symptoms. How long is the mismatch? 14/
This is why TESTING asymptomatic or minimally symptomatic people is so important, and why the lag in implementing widespread testing the USA (compared to Korea, say) is so unfortunate. 15/
Now, what do we know about COVID19 incubation and latency periods? Like everything else with COVID19, which has scientists worldwide scrambling to understand from genetic, clinical, epidemiological, and sociological points of view, we cannot yet be 100% sure. However… 16/
A careful study of 468 infector-infected pairs found that 12% involved transmission *before* the infector was symptomatic. #mismatch #quarantineloophole medrxiv.org/content/10.110… 17/
Examination of 124 Wuhan cases with clear contact history showed incubation period of 5 days (range: 1-11 days). It found that 73% of secondary cases were infected *before* onset of symptoms in first case. It estimated the mismatch period to be 3 days. medrxiv.org/content/10.110… 18/
A fine analysis of cases from China (N=135) & Singapore (N=93) found incubation was ~7-9 days. And there was pre-symptomatic transmission, with infection occurring on average 2.9 days (China) & 2.6 days (Singapore) before symptom onset in the infector medrxiv.org/content/10.110… 19/
Another analysis looking at 28 pairs of individuals suggests that "a substantial proportion of secondary transmission may occur prior to illness onset.” medrxiv.org/content/10.110… 20/
Study of 52 Chinese patients found incubation period to average 5 days, w typical range of 2-14 days. Such analyses yield the widespread recommendation of 14 day quarantine period (so as to be sure person is not infected, whether symptomatic or not). medrxiv.org/content/10.110… 21/
A further wrinkle is that these numbers are averages. Even if the latent period were not shorter than the incubation period on average, it could still be so for some patients. Indeed, this may be one of the factors that can make some individuals super-spreaders. | @salathe 22/
The likely existence of asymptomatic transmission of SARS-CoV-2 causing COVID19 does NOT mean we should not bother to test patients (e.g., to see if they are febrile) and ask them to self-isolate. Isolating carriers and patients is still essential! 23/
It is wise to detect & sequester patients, as is widely done (China, Korea, Singapore) by offering widespread free testing using real-time (rRT-PCR) detection of the SARS-CoV-2 virus (detecting viral RNA) cdc.gov/coronavirus/20… and (much easier for screening) fever checks. 24/
In sum, on average, COVID19 patients take about 7 days from exposure to SARS-CoV-2 to show symptoms (typical range 2-14 days), and a meaningful percentage of carriers can spread the disease for 2-4 days before they are symptomatic. We have our work cut out. 25/
Incidentally the @CDCgov released new guidelines for RNA testing on March 9: cdc.gov/coronavirus/20… 26/
And here is good advice about how to prepare your household for COVID19 from our amazing @CDCgov cdc.gov/coronavirus/20… 27/
A note from February 18 in @NEJM regarding two cases in Germany suggestive of possible asymptomatic spread of COVID19 that I had unaccountably missed. nejm.org/doi/full/10.10… 28/
And here is new work, as of March 16 in @ScienceMagazine, using a modeling approach to show that "undocumented infections" (which are numerically more common than documented infections) were the original source of 79% of documented COVID-19 cases. science.sciencemag.org/content/early/… 29/
