Karl Friston: How should we respond to an upsurge in covid-19 cases? 

There is a third way beyond lockdown or herd immunity, says Karl Friston

Arguments about how to respond to the current upsurge in coronavirus cases in the UK and Europe appear to have taken an adversarial turn—pitting lockdown against herd immunity. Even if many consider this a false dichotomy, there is a third way: namely, directing resources to local contact tracing. Together, lockdown and herd immunity furnish a window of opportunity during which contact tracing and support will make a difference—a window that will only be open for the next week or two.

From my viewpoint—as a panellist on the Independent SAGE with a special responsibility for epidemiological modelling—this week has seen two developments and a passing window of opportunity. Following the news briefing on Monday morning—by Chris Whitty, chief medical officer for England and Patrick Vallance, government chief scientific adviser—I received e-mails from friends who were genuinely upset by what they construed as predictions of impending fatalities in the hundreds per day. [1] I reassured them that these fantastical numbers were not predictions, but—as Vallance emphasised—illustrations of what could happen under simplistic assumptions about viral spread, based on exponential growth.

Exponential growth is something that the public needs to understand, but is not apt to describe viral transmission: we are dealing with a pandemic, not a nuclear chain reaction. When one models what is likely to happen—in terms of viral spread and our responses to it—a plausible worst-case scenario is a peak in daily deaths in the tens (e.g., 50 to 60) not hundreds, in November. This may sound rather precise; however, this kind of modelling has already proved to have predictive validity to within days. [2]

Why worst-case? This brings us to the second development; namely, an adversarial debate about the role of population (a.k.a., herd) immunity and reducing contact rates via physical distancing and lockdowns. The adversarial premise is that there are three mutually exclusive things that work in our favour: (i) reducing contact rates, (ii) establishing a sufficient level of population immunity (via exposure or vaccination) and (iii) suppression of community transmission, through tracing the contacts of infected people and supporting them in isolation. The adversarial premise can be challenged because all the available evidence points to a synergetic interaction between these factors, meaning they work hand-in-hand to promote each other. For example, population immunity augments the efficacy of physical distancing—and both make contact tracing easier. 

We have already developed a substantial population immunity (around 8% in the UK) and our physical distancing policies remain adaptive and effective. [3] But contact tracing appears to be missing from the debate (or perhaps conflated with testing). Quantitative modelling—and proof of principle in other countries—suggests that contact tracing is the way to “turn off the tap” and suppress community transmission. In short, it is the most efficient way to preclude worst-case scenarios. So, why is this important now?

We have heard many metaphors for the current situation, including, “knife edge”, “tipping point”, “critical point” and so on. This is the case and works in our favour. Because the situation is (mathematically) unstable, it is transiently susceptible to perturbations, such as an increase in the efficacy of contact tracing. This passing window of opportunity will not last long (about another week or so). According to quantitative modelling, even a moderate increase in the efficacy of current contact tracing could suppress viral transmission and elude the fatality rates above—and all the morbidity associated with (strategic planning for) surges in cases. [4]

The kind of contact tracing we are talking about here can only be—I am told—implemented on the ground with appropriate detective work, local knowledge, and the ability to support and monitor people in self-isolation. In brief, it requires the “shoe leather” epidemiology so successfully employed in Germany and other countries—not the “call centre” epidemiology that we appear to be committed to in the UK. Furthermore, it does not rest upon enhanced PCR testing capacity. In principle, one could implement (forward and backward) contact tracing based upon clinical diagnosis (that may or may not be confirmed posthoc with PCR testing).

So why has the government not redeployed resources to local public health teams—or integrated testing and tracing with primary health and social care? I do not know; however, they may have discounted this “third way” after being told contact tracing has to reach unattainable levels of 80% before it will make a difference. [5] This is quantitative nonsense; exactly because of the knife edge on which we currently find ourselves. An efficacy of 25% would be sufficient to suppress viral transmission, if implemented now. [6]

One might argue that “Two out of three ain’t bad.” This would be a dangerous argument. Population immunity and physical distancing (with partial lockdowns) create a context in which local contact tracing could prevent the virus from completing its tour of the UK.

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

Competing interests: none declared. 


1) https://www.gov.uk/government/speeches/chief-scientific-advisor-and-chief-medical-officer-briefing-on-coronavirus-covid-19-21-september-2020–2
2) https://wellcomeopenresearch.org/articles/5-89
3) https://www.gov.uk/government/speeches/chief-scientific-advisor-and-chief-medical-officer-briefing-on-coronavirus-covid-19-21-september-2020–2
4) https://www.medrxiv.org/content/10.1101/2020.09.01.20185876v1
5) https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/888807/S0402_Thirty-second_SAGE_meeting_on_Covid-19_.pdf
6) https://www.medrxiv.org/content/10.1101/2020.09.01.20185876v1