COVID: superiority of contact-tracing, epidemiol musings

A new article confirmed reasonably conclusively what was pretty obvious from the epidemiology we know so far about SARS-CoV-2: contact tracing is superior to identifying cases through symptomatic surveillance in terms of decreasing transmission (see covid contact tracing vs sx LancetID2020 in dropbox, or doi.org/10.1016/S1473-3099(20)30287-5)

Details:

-- after the outbreak in Wuhan, China, the Shenzen Center for Disease Control and Prevention, in a retrospective analysis, identified 391 SARS-CoV-2 cases and 1286 close contacts, from January 14 to February 12 (this is a secondary site, away from Wuhan, Hubei province). 1^st person identified of symptomatic Covid-19 in Shenzen was January 8

-- compared cases identified through symptomatic surveillance vs contact-tracing, estimating the time from symptom onset to confirmation of Covid-19, isolation of case, and admission to the hospital

-- contact-based surveillance vs symptomatic surveillance: mean age 45, 72% vs 45% female, overall younger age (15% being 0-9 years old, vs 2%; though fewer 20-69 years old, 75% vs 91%), mild cases 21% vs 28%/moderate 76% vs 62%/severe 3% vs 10%, asymptomatic 20% vs 3%, fever 71% vs 88%

-- close contact was defined as those who lived in the same apartment, shared a meal, traveled, or socially interacted with the index case 2 days before symptom onset

    -- contact-based surveillance included the identification of cases for monitoring and testing of close contacts of confirmed cases, independent of their symptoms

-- symptomatic cases were isolated and treated at designated hospitals; close contacts who tested negative were quarantined at home or a central facility and monitored for 14 days

    -- symptom-based surveillance: symptomatic screening at airport and train stations, community fever monitoring, home observation of recent travelers from Hubei, and testing of patients admitted to the hospital

-- the results below represents confirmed cases of SARS-CoV-2, by PCR

Results:

-- multivariate assessment showed that males had more severe symptoms (OR 2.5), and the probability of severe symptoms increased slightly with age (though those age 60-69 vs 50-59 did have an OR 3.4 of increased severity risk)

-- symptomatic cases: 356 (91%) had mild or moderate clinical severity initially, and as of February 22:

    -- cases were isolated an average of 4.6 days after developing symptoms

    -- 3 cases had died, 225 had recovered (median time to recovery 21 days)

-- contact tracing:

    -- cases were isolated after an average of 2.7 days (1.9 days less than in symptomatic cases)

    -- there was a gradient of contact risk for "close contacts"; as compared to other close contacts:

         -- household contacts were at a higher risk of infection, OR 6.27 (1.49-26.33)

         -- those traveling with the case were also at a higher risk, OR 7.06 (1.43-34.91)

    -- household secondary attacked rate was 11.2% (9.1-13.8)

        -- children were as likely to be infected as adults, with infection rate of 7.4% in children <10yo vs 6.6% in the population at large

        -- and there was no significant association between the probability of infection and the age of the index case

    -- the observed reproductive number for the virus, R₀, was 0.4 (0.3-0.5), with a mean serial interval of 6.3 days

-- in 183 cases with a well-defined period of exposure and symptom onset, mean incubation period:

    -- 4.8 days, and 95% develop symptoms within 14 days

        -- they estimated that 5% of people who develop symptoms would not show should symptoms until >14 days after infection. This is higher than a prior model suggesting that only 1% developed symptoms after 14 days

-- in 228 cases with known outcomes:

    -- median time to recovery was 20.8 days, longer in those 50-59 years old at 22.4 days vs those 20-29 years old at 19.2 days, and this median time to recovery tracked with severity of illness (those with severe symptoms had a 41% longer time to recovery than those with mild symptoms)

    -- symptom-based surveillance patients had virus confirmed 5.5 days after symptom onset (vs 3.2 days after symptom onset in those with contact-based surveillance)

-- among those isolated after developing symptoms, the symptom-based surveillance group were isolated 4.6 days after symptom onset, whereas the contact-based group was isolated 2.7 days after symptom onset

-- In the contact-based group: 20% of cases were asymptomatic at the time of 1^st clinical assessment and nearly 30% did not have a fever, suggesting a relatively high rate of asymptomatic carriage, though lower than some modeling studies project

Commentary:

--this study found several important things:

    -- that contact-tracing and isolation reduced the time during which cases were infectious in the community, and thereby considerbly reducing the R₀ to 0.4 (prior projections were >2, and <1 is likely to lead to viral eradication over time)

    -- that there was no difference between the age of infected people or the probability of being infectious by people of different ages: children were at similar risk of developing infection as adults, though less likely to be sick, but still seem to be important vectors of SARS-CoV-2

    -- current 14-day quarantine may miss more people than anticipated from prior studies (more like 5% than 1%)

-- one advantage of the study is that they started pretty much a day zero of Covid-19, comparing the effects of tracking patients by symptomatic presentation vs close contacts. This was still not a randomized controlled trial, but a retrospective analysis. But the data seemed to be pretty meticulous and augments our understanding of the disease, especially since it basically reinforces and substantiates our increasing knowledge-base from prior smaller or less well-done studies. Though, again, the caveat that pretty much all of the human studies are methodologically pretty flawed

---------------------------------------------------------------------------------

At this point, I think we are developing a reasonable sense of the epidemiology of this disease, albeit based on suboptimal studies. However I do think there is a convergence of information which serves as a very useful guide:

-- Most cases of Covid-19 on average have an incubation period of about 4-5 days from contact. This interval however has wide bounds, and the 14 day period of isolation is probably good for 95% of people. However there are a few cases extending out even to 30+ days. So a reasonable guess may be that after 14 days people should utilize more precautions for the next 2 weeks, with increased social distancing, use of masks, increased hand cleaning, and higher levels of isolation than otherwise

    --this study also reinforces that Covid-19 is a prolonged illness, with about 3 weeks from symptom onset to resolution in most cases, increasing with disease severity

-- asymptomatic patients are able to spread the disease. We do not know well how often this happens (minimal data), the infectivity rates of these patients, or how long asymptomatic people remain potentially contagious. However, identification of these patients is important in terms of preventing continual spread of the virus, especially since the highest viral loads are very early in the disease

-- Contact-tracing is extremely important, independent of symptoms. These people are at high risk of having the infection, even in the asymptomatic or early symptomatic timeframe. And isolation of these people who are SARS-CoV-2 positive is extremely important to prevent continuing spread of disease.

-- Using fixed criteria of fever and other specific symptoms may be useful in a time of very limited testing ability, since the positive predictive value for having the virus is much higher. However, since we know that the highest level of infectivity is in the 1^st 4-5 days after symptom onset, waiting until this constellation of symptoms happens will miss very large numbers of patients who are highly communicable of the virus. And, as result, will severely limit our ability to stop it spread...

    -- for example, an Italian study found that fatigue and altered sense of smell or taste were the 2 most commoon presenting symptoms: http://gmodestmedblogs.blogspot.com/2020/04/covid-nonpharm-interventions-work.html

-- children may well be a very important part of the viral reservoir and communicability, especially since they tend to be much less symptomatic and much less likely to be diagnosed at an early time. And, younger ones get about 5 different URIs/year, making it even more difficult to differentiate their mild symptoms from Covid-19. Another important issue is the difficulty of social distancing and quarantining kids. But, testing and isolating children is likely to be an important part of preventing viral dissemination

-- and, Wuhan did show us a very important lesson: with very aggressive contact-tracing, and enforced quarantine/isolation, they were able to get rid of the virus to the point that there are now no patients in the hospital with Covid-19. However, this only just happened after about 5 months of very aggressive controls. In the US, our controls are sporadic and inconsistent, making it hard to imagine that we will control this virus any time soon. But, as all of the studies suggest, the sooner we moved to a more aggressive posture, the sooner we are likely to be free of the virus. Unfortunately, our country seems to be moving in the other direction, which is likely to make this virus linger with 2^nd and 3^rd waves. And there are dire warnings that another Covid-19 outbreak might coexist with the next influenza outbreak

    -- we know that Covid-19 can present as a mild flu (see http://gmodestmedblogs.blogspot.com/2020/04/covid-presenting-as-mild-flu.html ), but will the combination be much more serious/lethal??

    -- and we may have limited hospital resources (ventilators, etc) given the added potential stress of severe flu plus Covid (see comment by CDC director Robert Redford: http://gmodestmedblogs.blogspot.com/2020/04/covid-nonpharm-interventions-work.html

----------------------------------------------------------------------

So, the bottom line is that if we want to really get rid of this virus, we should be doing testing, testing, testing, followed by isolation of those who are positive.

-- contact-tracing does lead to earlier diagnosis and earlier quarantine, at an especially important time given the high transmissability of SARS-CoV-2 early on.  And the outcomes in this study suggest that contact-tracing leads to a decrease in the communicability of the virus

-- contact infection tracks with the intensity of the contact, suggesting that family members who go out into the world to shop etc and get infected may decrease their transmission of the virus at home by minimizing contact/social-distancing in the house (studies in Wuhan, with their aggressive infection containment approach, ultimately found that 90% of cases were  from "family clusters")

-- young people seem to get the infection as often as older people, their symptoms tend to be milder, but they still seem to be disease vectors

-- the 14 day quarantine period is probably reasonable, though given that 5% of people may well have a longer infectious period, it certainly makes sense to continue with high-level personal hygiene/handwashing, use of a mask, as well as social distancing for some time thereafter, perhaps another 2 weeks.

--and, the current study simply formalizes what i think most of us already thought, but allows for a fuller and coherent picture....

geoff

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

to get access to all of the blogs:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just click on the magnifying glass on top right, then  type in a name in the search box and get all the blogs with that name in them

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list