A previous question on SE has asked about whether testing, a part of the strategy of containment, can help reduce transmission enough such that it reduces the need for aspects of mitigation (Could increased testing for COVID-19 slow its spread enough to allow the relaxation of social distancing measures?)

Here, I would like to ask a different question, which assumes containment measures as effective.

Epidemiologists seem to hold contradicting views on whether to conduct containment measures (testing and contact tracing) in regards stopping the Covid19 pandemic, in the context that it has indeed become one (with widespread and sustained community transmission in many countries).

The first is to continue with it. As seen in:
*WHO approaches [1,2]
*Countries, including China, South Korea, Singapore, and Germany [3,4]
*countries who formerly were not practising this but now do aim for it: UK, Spain [5,6]
*and some major journals [7].

The second is to not do so. In other words, to proceed with only or mostly mitigation measures. As seen in:
*Countries: notably Sweden [8]. (The US has low testing rates, but there does not seem to be an explicit strategy to not perform containment even if they could. In fact, the calls for more testing is increasing, though whether this results into a full fledged containment strategy (with rigorous contact tracing) remains uncertain)
*and some epidemiologists [9, 10].

I would like to know what are the epidemiological reasons, per se (and so not focusing on explicit economic, political, social or other aims or reasons), of not choosing the former and opting only for the latter approach?

In the case of the former, we can use epidemiological principles to illustrate its epidemiological bases. For instance, the formula for R0, that is, the basic reproduction number, R0 = BxKxD [11], we can see that containment's testing and contact tracing would allow us to ascertain the denominator for B, the attack rate. R0, would allow us to know the effectiveness of our interventions over time (aiming for R0<1). Also, knowing B allows us to trace and isolate those in contact with those who are infected, and so reducing transmission. This, when coupled with mitigation (which targets the K variable in the R0 formula), is thus a double-pronged attack against disease transmission.

With this in mind, a mitigation-only approach would deprive or deny a country (or another type of authority), the use of that other main plank in the R0 formula.

I add here as a postscript some common causes I have found, but the most common one, that is, that the infection is already too widespread, is not actually an epidemiological reason but rather a consideration of economic feasibility.

A semi-epidemiological one (i.e. deferring to epidemiological orthodoxy) seems to be instances where countries have opted to follow WHO's usual epidemic control phases model (which is actually designed with pandemic flu in mind as opposed to Covid19's epidemic behaviour).

Would you have further ideas as to explicit epidemiological reasons why mitigation might be preferred over containment, whether in general or specifically in the Covid19 pandemic?

Thank you.

1) That traditionally, the categorisations of actions during epidemic control phases are containment when the outbreak has not become a full-blown epidemic (ie in the early stages of the outbreak), and to focus on mitigation when it has become one (ie past the early part of the outbreak). A simple lay summary here is reported here [12], and this categorisation is recorded formally in WHO's usual categorisation (ie in general and not for the Covid19 outbreak specifically [13]). Note however, that WHO themselves reiterated the testing and tracing message (counter to their usual categorisation), since the declaration of PHEIC itself [14], highlighting that the Covid19 pandemic is not like that of pandemic flu, which the pandemic phase model fits better [15]. The WHO has criticised countries that have stopped containment on account of following the prior orthodoxy when they declared a pandemic on 11th March 2020: " describing this as a pandemic does not mean that countries should give up. The idea that countries should shift from containment to mitigation is wrong and dangerous" [16,17].

2) That containment is no longer feasible. This is related to the assumption, e.g. via statistical models, that the proportion of asymptomatic infected population is already too high. Hence this relates to the increasing practical, logistical and economical resources needed to meet needs: e.g. "crush of patients and shortage of tests" [18]. This has been the case in the case of Sweden. In the case of the UK (before 16th March 2020), it appears that a combination of considerations 1) and 2) were at play : " after the WHO declared coronavirus a pandemic, the government moved the UK from the “containment” phase into “delay”, accepting the inevitability of millions of infections " [19,20].

3) Depending on other countries' containment measures.


1] https://www.pscp.tv/WHO/1BdGYQVaPrgGX?t=3m52s
2] http://www.euro.who.int/en/about-us/regional-director/statements/statement-every-country-needs-to-take-boldest-actions-to-stop-covid-19
3] https://www.theguardian.com/world/2020/mar/21/coronavirus-asia-acted-west-dithered-hong-kong-taiwan-europe
4] https://www.ft.com/content/6a8d66a4-5862-4937-8d53-b2d10794e795
5] https://www.gov.uk/government/news/health-secretary-sets-out-plan-to-carry-out-100000-coronavirus-tests-a-day
6] https://www.irishtimes.com/news/world/coronavirus-spain-plans-to-ramp-up-testing-as-death-toll-surpasses-1-300-1.4209000
7] https://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(20)30679-6.pdf
8] https://www.weforum.org/agenda/2020/03/sweden-under-fire-for-relaxed-coronavirus-approach-here-s-the-science-behind-it/
9] https://www.wbur.org/commonhealth/2020/03/10/coronavirus-epidemic-mitigation-strategy-tufts-infectious-disease
10] https://www.wbur.org/commonhealth/2020/03/10/coronavirus-epidemic-mitigation-strategy-tufts-infectious-disease
11] β is the risk of transmission per contact (i.e. basically the attack rate (that is, the proportion of cases over those infected)); κ is the number of such contacts that an average person in the population normally would have per time unit (in the absence of any disease); and D is the duration of infectivity of an infected person, measured in the same time unit as κ was (Giesecke, Johan. Modern infectious disease epidemiology. Third edition. CRC Press, 2017. pg 111)
12] https://www.newyorker.com/news/news-desk/what-it-means-to-contain-and-mitigate-the-coronavirus 13] https://apps.who.int/iris/handle/10665/272442
14] https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports
15] https://www.newscientist.com/article/2235342-covid-19-why-wont-the-who-officially-declare-a-coronavirus-pandemic/
16] https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-mission-briefing-on-covid-19---12-march-2020
17] https://qz.com/1813587/is-coronavirus-technically-a-pandemic-does-that-matter/ 18] https://www.nature.com/articles/d41586-020-00823-w
19] https://www.theguardian.com/world/2020/mar/13/who-urges-countries-to-track-and-trace-every-covid-19-case
20] https://www.theguardian.com/world/2020/mar/12/uk-moves-to-delay-phase-of-coronavirus-plan

  • 1
    "to proceed with only or mostly mitigation measures. As seen in: … Sweden". Note that in terms of deaths-per-million (a much more reliable indicator than the popular figures stated in the news), Sweden is now (April 16), the 8th highest in the world (118.4), while nearby Denmark and Norway are a half and a quarter of that at 53.3 and 28.23. Sweden is currently just ahead of rapidly climbing USA (94.54), while nearby Canada is less than a third that at 27.25. statista.com/statistics/1104709/… Apr 16, 2020 at 20:55


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.