Karl Friston: New covid-19 restrictions can only work as a prelude to a restructured find, test, trace, isolate and support system

Last week it was “herd immunity.” This week it is “circuit breakers.” Everyone—from SAGE to Her Majesty’s Government, from Independent SAGE to the All-Party Parliamentary Group, from the John Snow Memorandum to the Great Barrington declaration—appears to have a cunning plan. [1-4] The consensus—largely endorsed by Keir Starmer—appears to rest on three interventions that may or may not be tenable. [5,6] These include local lockdowns and social distancing measures (scaffolded on a three-tier approach), a restructured and enhanced find, test, trace, isolate and support (FTTIS) system, and now, a “circuit breaker”. So, what is a circuit breaker?

I interpret a circuit breaker to be a two-week national lockdown—of the sort imposed in March for a longer period—that is, on this occasion, accompanied by an extended (and possibly early) half-term for schools. Natural questions then arise: is two weeks long enough—and for what? To answer this, it is necessary to specify what we hope to achieve with a circuit breaker. If the aim is suppression of community transmission, or at least, an innocuous endemic equilibrium (a state where the disease cannot be totally eliminated but remains in the population), then there are clear answers to this question. 

A circuit breaker will move us towards an endemic phase of the outbreak, but it will not be innocuous. [7] Dynamic causal modelling, which includes the effects of local lockdown and social distancing measures, suggests a circuit breaker, implemented in the next few days, will allow us to elude peak fatalities of around 100 deaths per day (averaged over seven days)—and break the (pandemic) cycling between low-prevalence, high-transmission states and high-prevalence, low-transmission states. [8] Under this model, a circuit breaker should partially quench the secondary wave, after which the virus will slowly burn through communities who have not yet established an effective level of population immunity. [9] Quantitatively, this will have the look and feel of an endemic infection for a few weeks or months, but with daily deaths in the dozens per day that slowly escalate

To achieve an innocuous endemic or suppression, we will have to turn to the intervention we know works; namely, “shoe-leather” epidemiology, informed by clinical surveillance and timely access to local testing data. In short, the only raison d’être for a circuit breaker is as a prelude to recover that window of opportunity that was lost over the summer—a window in which FTTIS is the final player.

A circuit breaker of two weeks now may create the right environment for an enhanced FTTIS system, even if it takes four weeks to be transformed from its current state. As noted by nearly everyone I speak to, this restructuring will require a dismantling of the current test and trace system to integrate it within existing public health and NHS infrastructures, with a decentralisation of contact tracing. [10] If this restructuring can improve the efficacy of contact tracing and isolation to even moderate levels (e.g., 25%), quantitative modelling suggests we could return to summer levels of prevalence by Christmas. [11] With a more efficient—if challenging—efficacy of 80%, it may be possible to suppress viral transmission by the New Year.

One outstanding question is why a circuit breaker should entail a national lockdown? From an epidemiological stance, this speaks to the distinction between a secondary and second wave. Here, a secondary wave implies the virus is spreading through communities who have not previously been exposed or, indeed, did not exist in the primary wave (e.g. communities of students in accommodation halls). A second wave refers to reinfection of communities that have already experienced a first wave (much like seasonal influenza). The distinction is crucial because the secondary wave rests upon transmission between communities, as opposed to between individuals within a community. In turn, this suggests that the kind of transmission necessary to attenuate a secondary wave requires both the reduction of contact between people and communities or regions. The latter is assured to a greater extent by a national lockdown, in relation to local or regional measures.

In sum, the best elements of all cunning plans harness the three interventions at hand—in a carefully orchestrated way—to suppress viral transmission with the least effect on education and the economy. The most coherent strategy, at the present time, appears to be the use of a circuit breaker to create a window of opportunity in which an enhanced and restructured FTTIS programme will take us through to the end game.

Karl J. FristonScientific Director: Wellcome Centre for Human Neuroimaging, Professor, Queen Square Institute of Neurology, University College London, and Honorary Consultant: The National Hospital for Neurology and Neurosurgery, and a memeber of Independent SAGE. 

Competing interests: none declared.

Note: The quantitative predictions in this opinion piece are based upon dynamic causal modelling. Although this model has been optimised using Bayesian model comparison over the course of the pandemic, it is only one model. As such, any predictions or assertions should be qualified by the fact that they are entirely conditioned upon the model used, which may or may not be the best model.


1. https://www.independentsage.org/
2. https://www.marchforchange.uk/covid_secure_uk
3. https://www.thelancet.com/lancet/article/s0140-6736(20)32153-x
4. https://en.wikipedia.org/wiki/Great_Barrington_Declaration
5. One should probably exclude the UK government and the Great Barrington declaration from this consensus
6. https://labour.org.uk/press/keir-starmer-responds-to-the-prime-ministers-statement-in-the-house-about-new-covid-19-restrictions/
7. i.e., a level of endemic infection whose associated morbidity is mild in relation to other endemic coronaviruses (e.g., those that cause the common cold).
8. https://wellcomeopenresearch.org/articles/5-103
9. https://www.medrxiv.org/content/10.1101/2020.09.01.20185876v1
10. https://www.theguardian.com/commentisfree/2020/oct/12/englands-three-tier-system-is-an-admission-of-failure
11. Efficacy is defined here (and in the model) as the probability of supported isolation if infected but asymptomatic.
12. https://www.fil.ion.ucl.ac.uk/spm/covid-19/TR5_Second_Wave.pdf