COVID-19: Keep calm and wash your hands.
A friend sent me a few questions on COVID-19. I transcribed the resulting Q/A and hope that others find this information helpful. I’ve done my best to cite sources and clearly state where I am speculating.
tl;dr don’t panic, don’t buy masks [note from Jan 2022: I’ve left this here as an honest record of the things I said and felt we knew at the time, but we now know how critically important high-filtration masks are at stopping transmission. Nevertheless, we were faced with a severe shortage at the time, which is partly what motivated this comment.], do prepare, and do wash your hands.
- What percentage of the population is expected to be infected?
The average number of new cases “caused” by each person infected with SARS-CoV-2 (i.e., the R0*) is around 2.25, but, despite what you may read or see in the movies, that number doesn’t really tell us much about the likely percentage of the population that could be infected. Prior to interventions and behavior change, the biggest factor in determining the expected size of an outbreak is actually the variability in new cases, i.e., secondary infections, (see our pre-print for more information).
If individuals infected with SARS-CoV-2 cause a highly variable number of secondary infections (like SARS ’03 and Ebola ‘15), then we’d estimate around 5% of the population could be at risk (assuming no interventions and that everyone is equally susceptible, obvisouly neither of these are realistic and the real numbers will be much lower). If it’s flu-like, i.e., less variability in secondary infections and less reliant on super-spreading events, then it could be 40% (again, there’s no chance this is a reasonable number given the interventions and social distance we know are occurring).
Critically, the outbreak is winding down in China (see figure), so we can just ask the data. The percent infected (accounting for 2 out of 3 cases going unreported [R1]) in Hubei province was 0.4%, even assuming 5x under-reporting the percent infected in Hubei is still under 1%. That’s around 100,000 cases. The rest of the provinces in China are between 0.05–0.1% (again these figures account for published estimates of under-reporting).
Now, the big issue here is how quickly and effectively the rest of the world ramps up testing, surveillance, and an effective public health response. For more information on the role of contact tracing in controlling SARS-CoV-2, see this paper by Hellewell et al.; if you’re curious about how to incorporate behavior change into these estimates see this paper by Eksin et al.; and to learn more on super spreading and epidemics start here.
2. Is there evidence for protective immunity post-infection? Estimates on protective effect, duration?
This is one of the big open questions, and, unfortunately I’m very far from an expert on the human immune system. From what I’ve read, it appears as though individuals do have protective immunity [R2] and — as a result — the efficacy of treating individuals with blood-serum therapy from convalescent patients is being investigated. The anecdotal reports of individuals being “re-infected” can probably be explained by patients initially having mild symptoms [R3], which quickly worsen, and/or the false negative rates of tests and/or individuals being PCR positive after no longer meeting the clinical definition for COVID-19 [R4].
Of course, given the complexity of the human immune system, I’d also suspect that some individuals probably can get re-infected. The question is whether it’s common. To that end, all the evidence points to re-infection being extremely rare/non-existent. For more information, Prof. Jesse Bloom at the Fred Hutchinson Cancer Research Center has compiled an excellent resource on our current knowledge of SARS-CoV-2 immunology.
3. What is the current estimated average duration of illness? Severity?
Median hospital duration is estimated to be 12 days [R5]. Mean symptom onset-to-recovery is 22 days (18–83 95% CI) [R6]. The incubation period is around 4 days [R5]. The case fatality rate is somewhere between 0.7 and 2% see the WHO joint mission report, a recent editorial in NEJM, and [R5]. This may sound low, but influenza is one of leading causes of death worldwide each year and has a case fatality rate of around 0.1%. Important caveat: estimating the case fatality rate — especially early on in an outbreak — is notoriously difficult, see this paper by Jewell et al. for more information.
That being said, the elderly and individuals with chronic medical conditions are at much higher risk for severe disease, but appropriate and timely treatment in a hospital improves an individual’s prognosis dramatically. What this means is that a calm, effective public health response and a prepared medical system can have a large, positive effect on reducing mortality.
4. Is anyone doing community level modeling on hospital capacity for severe cases, and integrating that with models estimating the pandemic dynamics? (Many hospitals are at capacity or over and divert during regular flu season. )
This is a big concern. Lombardy in Italy is reporting that hospitals are near capacity, but a lot of that is likely due to being close to capacity already from an ongoing influenza outbreak. I’m not aware of any large-scale, public data sets on hospital capacity, but I suspect this is something a number of research groups are actively investigating. I’ll update this section as we learn more.
5. How long is the virus expected to circulate for? Months? Years? Forever?
I think this depends a lot of on how the world responds and all I can do now is speculate. The outbreak in China is winding down, so it lasted about three months there. My intuition is that most countries will probably take at least that long to bring their outbreaks under control. However, I’d guess that we’ll be dealing with SARS-CoV-2 for quite some time as it persists in places with weaker public health systems and/or slower responses. The rumors that SARS-CoV-2 will “go away” during the summer are unsubstantiated. See this post by Prof. Marc Lipsitch, director the Center for Communicable Dynamics at Harvard School of Public Health.
6. Any evidence of adaptation?
None. Coronaviruses have a pretty slow rate of mutation accumulation (relatively speaking) and all the evidence to date suggests the virus is just picking up mutations as expected. The best source for information on SARS-CoV-2 genomic surveillance is next strain and for scientific analysis of viral genomic data see http://virological.org/.
7. What should we do to prepare?
Most importantly, don’t panic. As Prof. Caitlin Rivers at the Johns Hopkins Center for Health Security reminds us, now that testing is ramping up we are going to see more cases. This is normal and exactly what we need to do public-health-wise to control the outbreak. And also, don’t run out and buy masks, see current CDC recommendations. They don’t really protect the public [note from Jan 2022: I’ve left this here as an honest record of the things I said and felt we knew at the time, but we now know how critically important high-filtration masks are at stopping transmission. Nevertheless, we were faced with a severe shortage at the time, which is partly what motivated this comment.] and are desperately needed by health care workers. For example, news reports suggest that stores in Colorado are already sold out of masks (and that state doesn’t have any reported cases yet).
What can you do?
First, follow travel guidelines from the CDC and be sure you’ve checked out out ready.gov. For example, make sure you (and anyone you’re responsible for) have prescriptions filled and enough food on hand for at least a week. Families should also plan for school closures and work-related disruptions. They might not happen, but that doesn’t mean you can’t be ready.
Second, stay informed and don’t amplify fake news and/or conspiracy theories. As the excellent resource Calling Bullshit says, ask yourself three questions: 1) who is telling me this? 2) how do they know it? and 3) what are they trying to sell me? Here’s a great Twitter thread from Prof. Brandon Ogbunu on who he looks to for COVID/SARS-CoV-2 information.
Third, there has been an unprecedented scientific response and a lot of the data you see are being collected/maintained by volunteers. SARS-CoV-2 has already wiped trillions off the books. We have to change the way we fund outbreaks responses now! As Bill Gates recently said in his New England Journal of Medicine Op-Ed, “We also need to invest in disease surveillance, including a case database that is instantly accessible to relevant organizations, and rules requiring countries to share information.” You can read about the public database we’re building in our recent correspondence to The Lancet Infectious Diseases.
Last, and tied with not panicking for importance, WASH YOUR HANDS! It works to stop SARS-CoV-2 from spreading. I’ll leave you with this thread from Prof. Karen Fleming at Johns Hopkins on why hand washing is so effective and so vital during this outbreak.
References
R1 — Bhatia, S., Imai, N., Cuomo-Dannenburg, G., Baguelin, M., Boonyasiri, A., Cori, A., … & Gaythorpe, K. Report 6: Relative sensitivity of international surveillance. https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College---COVID-19---Relative-Sensitivity-International-Cases.pdf
R2 — Thevarajan, I., Nguyen, T. H., Koutsakos, M., Druce, J., Caly, L., van de Sandt, C. E., … & Tong, S. (2020). Breadth of concomitant immune responses underpinning viral clearance and patient recovery in a non-severe case of COVID-19. medRxiv. https://www.medrxiv.org/content/10.1101/2020.02.20.20025841v1
R3 — Li, Q., Guan, X., Wu, P., Wang, X., Zhou, L., Tong, Y., … & Xing, X. (2020). Early transmission dynamics in Wuhan, China, of novel coronavirus–infected pneumonia. New England Journal of Medicine. https://www.nejm.org/doi/full/10.1056/NEJMoa2001316
R4 — Lan L, Xu D, Ye G, et al. Positive RT-PCR Test Results in Patients Recovered From COVID-19. JAMA. Published online February 27, 2020. doi:10.1001/jama.2020.2783 https://jamanetwork.com/journals/jama/fullarticle/2762452
R5 — Guan, W. J., Ni, Z. Y., Hu, Y., Liang, W. H., Ou, C. Q., He, J. X., … & Du, B. (2020). Clinical characteristics of 2019 novel coronavirus infection in China. New England Journal of Medicine https://www.nejm.org/doi/full/10.1056/NEJMoa2002032
R6 — Dorigatti, I., Okell, L., Cori, A., Imai, N., Baguelin, M., Bhatia, S., … & FitzJohn, R. (2020). Report 4: severity of 2019-novel coronavirus (nCoV). https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-2019-nCoV-severity-10-02-2020.pdf
R7 — Lei, J., Li, J., Li, X., & Qi, X. (2020). CT imaging of the 2019 novel coronavirus (2019-nCoV) pneumonia. Radiology, 200236.
https://pubs.rsna.org/doi/10.1148/radiol.2020200236
R8 — Jun Chen, Lianlian Wu, Jun Zhang, Liang Zhang, Dexin Gong, Yilin Zhao, Shan Hu, Yonggui Wang, Xiao Hu, Biqing Zheng, Kuo Zhang, Huiling Wu, Zehua Dong, Youming Xu, Yijie Zhu, Xi Chen, Lilei Yu, Honggang Yu medRxiv 2020.02.25.20021568 https://www.medrxiv.org/content/10.1101/2020.02.25.20021568v1
*R0, or R nought or R zero, is the expected number of secondary infections resulting from a single infected individual in an entirely susceptible population.